Late-Winter Habitat of the Little Smoky Boreal Woodland Caribou

CWBM 2015: Volume 4, Number 2 ISSN: 1929-3100

Original Research & Management

Late-winter Habitat of The Little Smoky Boreal Woodland Caribou (Rangifer tarandus caribou) Population, Alberta, Canada: Vegetation Composition and Structural Characteristics, Management Implications, and Habitat Conservation Plan Gilbert PROULX1 1 Alpha Wildlife Research & Management Ltd., 229 Lilac Terrace, Sherwood Park, Alberta, T8H 1W3, Canada. Email: [email protected]

Abstract The recovery of a species at risk such as the boreal woodland caribou (Rangifer tarandus caribou) must start with the conservation of valuable habitats. Little is known, however, about habitats used by the Little Smoky caribou population inhabiting the Alberta Newsprint Company (ANC)’s Forest Management Agreement (FMA) area (143,932 ha), in northwest Alberta. This study was conducted in late winters (January-February) of 2013 and 2014. The objectives were to 1) test a draft query (list of parameters from a vegetation database that are used to identify specific areas) largely based on a Government of Alberta’s Habitat Suitability Index (HSI); 2) verify habitat use by woodland caribou; 3) in the event that the draft query proves to be unsuccessful to predict late-winter caribou distribution, identify stands that should be used in the development of an effective query; and 4) test the new query. I hypothesized that caribou tracks would be found mostly in mixed tamarack (Larix laricina) and black spruce (Picea mariana) muskegs with ≤20-m-high trees and >30% canopy closure, and in lodgepole pine (Pinus contorta) – tamarack – black spruce mixed stands. In 2013, the distribution of 190 caribou tracks during the snowshoe survey of 26 transects (56,025 m) suggested that the draft query was inadequate to predict the late-winter distribution of the Little Smoky woodland caribou. Tracks were absent from pure black spruce and lodgepole pine stands, and from 60-79-years-old pure tamarack stands. Tracks were significantly more frequent than expected P( < 0.05), or were present as expected (P>0.05) based on habitat availability, in muskegs with mixed black spruce and tamarack stands. In 2014, a new query based on caribou habitat use in 2013 successfully predicted the late-winter distribution of caribou. Seventy- five (96%) of 78 tracks recorded during the survey of 13 transects (31,130 m) were found in medium- and high- quality

Correspondence: Gilbert Proulx, Alpha Wildlife Research & Management Ltd., 229 Lilac Terrace, Sherwood Park Alberta, T8H 1W3, Canada. Email: [email protected] Page 82 Proulx

polygons. Tracks were significantly more frequent than expected P( <0.05) in muskegs of ≥60%-<90% black spruce and ≥60%-<90% tamarack (P<0.01). Caribou used ≥90% tamarack stands, and ≥60%-<90% lodgepole pine stands with black spruce, closed canopy trees and large trees, as per availability (P>0.05). Otherwise, caribou avoided all upland lodgepole pine stands. There were no caribou tracks in ≥90% black spruce stands. Caribou used seismic lines as they encountered them, i.e., without attraction or avoidance. This study confirmed my hypothesis that caribou prefer mixed tamarack and black spruce muskegs with ≤20-m-high trees and >30% canopy closure, and lodgepole pine-tamarack-black spruce mixed stands. It also showed that the new query was adequate to predict late-winter habitat use by the Little Smoky boreal woodland caribou inhabiting the ANC’s FMA area. Late-winter habitats used by caribou represent only 28,183 ha (20%) of the FMA area. The caribou range within the ANC’s FMA area is highly fragmented, and adequate habitats are scattered across the landscape. A habitat conservation plan is proposed to ensure the persistence of landscapes that would provide the Little Smoky caribou population with interconnected late-winter habitats that would meet their seasonal requirements, such as food and protection and escape cover against harsh environmental conditions and predators. Key Words: Alberta, Black Spruce, Caribou, Late-winter habitat, Little Smoky population, Lodgepole pine, Muskegs, Tamarack, Wildlife management.

Introduction treed muskegs where tamarack stands had >20% canopy closure, black spruce stands had >40% canopy closure, and tree were ≤20-m Environment Canada (2012) classified boreal woodland caribou high. Caribou also used adjacent upland stands, particularly for (Rangifer tarandus caribou) populations according to 51 ranges, i.e., travelling between muskegs (Proulx 2013). geographic areas occupied by a group of individuals that are subject In an effort to better assess the effects of future timber harvest to similar environmental factors and threats (Thomas and Gray on boreal woodland caribou, I tested queries (list of parameters 2002). Environment Canada (2012) also classified populations of from a vegetation database that are used to identify specific areas) boreal woodland caribou according to their self-sustainability. A to predict the distribution of woodland boreal caribou late-winter self-sustaining boreal woodland caribou population is defined as a habitat in ANC’s FMA area in late winters of 2013 and 2014. local population that, on average, demonstrates stable or positive For the purpose of this study, mid-January to mid-February was population growth over the short term, and is large enough to defined as late winter. At this time of year, bad weather and long withstand stochastic events and persist over the long-term, without deep freeze periods can become a major constraint for the species the need of ongoing intensive management. In western Alberta, (Fancy and White 1987) and the animals can be energetically Environment Canada (2012) rated the Little Smoky caribou stressed (Farnell and McDonald 1988). population “likely as not self-sustaining”. In this paper, I explain a stepwise approach I employed to Limited or contradictory information exists about habitats used identify polygons (homogenous areas encompassing one or more by the boreal woodland caribou of the Little Smoky population, forest stands with similar composition and structure) used by which inhabits the Alberta Newsprint Company (ANC)’s Forest woodland boreal caribou in late winter. The objectives of this Management Agreement (FMA) area in the Boreal Plains study were to 1) test a draft query based largely on the Government Ecozone. Edmonds (1993) found that, in winter, primary cover of Alberta’s woodland caribou Habitat Suitability Index (HSI); 2) types used by caribou were mostly lodgepole pine (Pinus contorta) verify habitat use by woodland caribou; 3) in the event that the and mixed pine/spruce (Picea spp.) stands; caribou made little use draft query proves to be unsuccessful to predict late-winter caribou of muskegs. Hervieux et al. (1994) suggested that caribou used distribution, identify stand characteristics that should be used in primarily >80-yeard-old lodgepole pine and lodgepole pine/spruce the development of an effective query; and 4) test the new query. forests interspersed with relatively small muskegs that provide Based on Rettie and Messier (2000) and Proulx (2013), I considered terrestrial and arboreal lichens. Neufeld (2006) found that caribou that tamarack and black spruce would be key components of stands in the Little Smoky range showed strong selection for lodgepole used by caribou. I hypothesized that caribou tracks would be pine, black spruce (Picea mariana) and tamarack (Larix laricina) found mostly in mixed tamarack and black spruce muskegs with at 1-km scale. Caribou avoided stands dominated by white spruce ≤20-m-high trees and >30% canopy closure, and in lodgepole (Picea glauca) and aspen (Populus tremuloides). No information is pine-tamarack-black spruce mixed stands. available on the composition (e.g., percentage of pine, spruce and tamarack) and structure (tree size, canopy cover, basal area, etc.) Study area of stands used by the Little Smoky boreal caribou population. The ANC’s FMA area is located near the town of Fox Creek However, in another region of the Boreal Plains Ecozone, Proulx (N 54o 23’, W 116o 40’), approximately 300 km northwest of (2013) showed that boreal caribou preferred lichen-rich open and Edmonton in Alberta (Figure 1). The area lies in the Alberta’s Proulx Page 83

Upper and Lower Foothills Natural Subregion (Natural Regions 94% of productive forests, 5% of non-forested and 1% of Committee 2006) within the Boreal Plains Ecozone (Environment recreational areas. Tree species found within the FMA include Canada 2012). The region is characterized by short and warm lodgepole pine, black spruce, white spruce, balsam fir (Abies summers, and cold winters (http://www.ecozones.ca/english/ balsamea), aspen, balsam poplar (Populus balsamifera) and birch zone/index.html). Average snow accumulations tabulated by (Betula spp.). Muskegs consist of white and black spruce, balsam Environment Canada from 1981 to 2012 are 20-25 cm in January fir, jack pine Picea( balsamifera), and tamarack. The Little Smoky and February, with extremes of 65-75 cm (http://climate.weather. caribou population occupies 143,932 ha (1,439 km2 – 38%) of the gc.ca/climate_normals/results_1981_2010_e.html?stnID=2632& ANC’s FMA area; it occurs predominantly in the western portion prov=&lang=e&dCode=1&dispBack=1&StationName=Whiteco of the FMA area. urt&SearchType=Contains&province=ALL&provBut=&month1 Major land activities include logging, oil and gas exploration and =0&month2=12). development, non-motorized outdoor recreation, off-road vehicle The FMA area is 378,397 ha (3,784 km2) in size and comprises use (snowmobile, all-terrain vehicles), recreational hunting, and

Figure 1. Location of the Little Smoky caribou population range in the Alberta Newsprint Company’s Forest Management Agreement area, western Alberta, Canada. Page 84 Proulx commercial trapping (Brown and Hobson 1998). Access in the Methods area exists in the form of all-weather and dry-weather resource roads, and right-of-ways for pipelines, powerlines and seismic lines Queries and predictive distribution maps (Smith et al. 2000). Only gravelled roads exist in the Little Smoky 2013 Query range. The development of the 2013 query to predict the

Table 1. Query largely based on Stepnisky et al.’s (2011) Habitat Suitability Index for the development of 2013 predictive distribution maps of late-winter habitats used by the Little Smoky boreal woodland caribou population, Alberta, Canada.

Parameters Description Weight Justification/reference

Age Mature or old (≥80 years) 3 Mature and old stands provide caribou of the Little Smoky Transition young-mature 1 population with arboreal and terrestrial lichen (Stepinsky (≥60 and ≤79 years) et al. 2011). It is unlikely that caribou use <60 years-old Young (<60 years) 0 stands on the basis of Smith’s (2004) findings.

Treed muskegs Wet ground with late-seral stands and 3 Proulx (2013) at least 30% crown closure

Acceptable Mixed black spruce (<90%) coniferous 3 Stepnisky et al. (2011) stand types stands Pure larch (≥90%) 3 Mixed larch (<90%) coniferous stands 3 Pure pine (≥90%) 3 Lodgepole pine/larch mix 3 Larch-leading mixedwood 2 Lodgepole Pine/spruce mix 2

Unacceptable Pure black spruce (≥90%) stands 0 Proulx (2013); however, Stepnisky et al. (2011) identified stand types pure black spruce stands as favorable. Deciduous-dominated stands 0 Neufeld (2006) Spruce mixedwood 0 Lichen are seldom found in densely shaded forests (Ahti and Hepburn 1967)

Spatial ≤2 km from treed muskegs High value Proulx (2013) proximity of >2 km from trees muskegs Low value upland mixedwood

Polygon ranking Rank Sum of weights High 5-6 Medium 4 Low 0-3 Proulx Page 85 distribution of late-winter habitats for the Little Smoky distances along survey transects and to record changes in habitat caribou population was largely based on Stepnisky et al.’s types in order to verify vegetation inventory accuracy. Habitats (2011) HSI for caribou of the Lower Athabasca range (Table (immature, young, mature and old; Proulx and Kariz 2005) were 1). Two queries were originally developed where black spruce classified as coniferous (coniferous species ≥90%; crown closure stands were considered pure when black spruce represented ≥20%), deciduous (deciduous species ≥90%; crown closure ≥20%), either 90% or 100% of the stand composition. However, as the or mixed (neither type ≥90%; crown closure ≥20%) according to analysis of this study’s caribou track distribution demonstrated AVI datasets. that stands with ≥90% black spruce were avoided by caribou, Each caribou track crossing survey transects was recorded; GPS the analysis of stands with 100% black spruce became obsolete. locations and linear distance along transects were recorded for all The query was based entirely on forest stand characteristics, tracks. When trails were found, they were investigated on both and the impact of the density of seismic lines/km2 on the sides of transects in order to find clear hoof prints and determine rating of the polygons (e.g., Stepnisky et al. 2011) was not the minimum number of animals involved. The combination taken into account (Table 1). of hoof print (shape, size, presence of dew claws) and trail (gait, The Silvacom Group (Edmonton, Alberta) produced late- distance between prints) characteristics were used to identify winter habitat vector maps using Alberta Vegetation Inventory the tracks and differentiate them from deer Odocoileus( spp.), elk (AVI; AESRD 2014). Weight values of most forest stands were (Cervus canadensis), and moose (Alces alces). based on Stepnisky’s et al.’s (2011) HSI. After comparing draft Woodland Caribou track distribution and data analyses maps to known caribou locations (e.g., Smith 2004), weight The proportions of polygon rating classes (i.e., high-, medium-, values were adjusted to increase or reduce the effect of a variable and low-quality) and habitat classes (i.e., stands with similar on the overall predicted distribution of the Little Smoky caribou vegetation composition and structure found within the polygons) population. Polygons were rated according to the sum of traversed by survey transects were used to determine the expected weighed parameters and classified into high- (treed muskegs and frequency of tracks/polygon class or habitat class if tracks were other specific forests), medium-, or low-quality rating classes distributed randomly with respect to polygon or habitat classes (Figure 2). A color code was used to differentiate polygon (Proulx 2013). Polygon preference or avoidance (based on observed rating classes. Although linear features were not considered vs. expected frequency of track intercepts) was tested with Chi- in the rating of polygons, the proportions of transects crossing square statistics with Yates correction (Siegel 1956; Zar 1999). linear features were recorded for data analyses (see below). When >20% of cells had an expected frequency smaller than 5, or 2014 Query when any expected frequency was smaller than 1, data classes were Following the 2013 field assessment of the draft query, which combined (Cochrane 1954). Then, habitat classes with similar was judged inadequate to predict the late-winter distribution of characteristics were pooled. When Chi-square analyses suggested caribou, a new query was developed based on stands that were an overall significant difference between the distributions of preferred or used as per availability by caribou in late winter 2013 observed and expected frequencies, comparisons of observed (Table 2). The new query identified stand composition, tree height to expected frequencies for each polygon or habitat class were and canopy closure, which all impact on arboreal and terrestrial conducted using the G test for correlated proportions (Sokal and lichen growth and persistence (Ahti and Hepburn 1967; Carroll Rohlf 1981). Probability values ≤0.05 were considered statistically and Bliss 1982; Campbell and Coxson 2001). As in the 2013 query, significant. the rating of polygons was based on inherent habitat suitability and For the purpose of comparing with the 2013 findings, an analysis did not factor in linear features. of caribou track distribution was conducted using the individual Snowtracking stand’s vegetation composition and structure, i.e., canopy closure Survey methodology followed Proulx (2013). A random and tree height which both influence arboreal and terrestrial lichen stratified approach was used to locate transects averaging ≥1-km growth and persistence. The analysis of caribou track distribution long and ≥1 km-apart (within and between years) that were was based on Proulx (2013) where habitat classes were determined accessible by truck or snowmobile by the inventory team (Proulx based on vegetation composition: 1) pure black spruce (M) and 2013). Transects were selected to cross suitable and unsuitable tamarack (L) stands; 2) ≥60%-<90% black spruce stands (60M); polygons as predicted by the queries. Transects were plotted on 3) ≥60%-<90% tamarack (60L) stands; 4) pure lodgepole pine predictive maps; UTMs were recorded at the beginning and the stands (Pl); ≥60%-<90% lodgepole pine stands with (60PlM) and end of transects using hand-held GPS. When snowshoeing, the without (60Plw/o) black spruce; 5) mixed coniferous stands with number and length of transects varied according to accessibility, <60% black spruce, tamarack or lodgepole pine, but with (CML) safety, and environmental conditions. Snowshoe transects or without C(w/o) ≥20% black spruce or tamarack; 6) immature crossed various habitat types; a hip chain was used to record linear sites (IS) encompassing sloughs and drainages with scarce trees Page 86 Proulx

Treed Muskeg - High Value

Upland Mixedwood (CD) within 2km - High Value

Upland Mixedwood (CD) within 2km - Low Value

High Value Forest

Medium Value Forest

Low Value Forest

Figure 2. Predicted late-winter habitat distribution of the Little Smoky boreal woodland caribou population using a query largely based on Stepnisky et al.’s (2011) Habitat Suitability Index for caribou of the Lower Athabasca range (see Table 1 for description of polygons).

Table 2. Query used in 2014 for the development of a predictive distribution map of late-winter habitats for the Little Smoky caribou population, Alberta, Canada.

Vegetation type and structure Ranking of Stands

• ≥60%-<90% tamarack, 31-50% canopy closure, 0-19-m-high trees. Preferred (high-quality polygons) • ≥60%-<90% black spruce, 6-30% canopy closure, 10-19-m-high trees. • ≥60%-<90% lodgepole pine with black spruce, 51-70% and 71-100% canopy closure, 10-19-m-high trees. • mixed coniferous stands, <60% black spruce, tamarack or lodgepole pine, and ≥20% black spruce or tamarack.

• ≥60%-<90% tamarack, 6-30% and 51-70% canopy closure, Used as per availability (medium-quality polygons) 0-19-m-high trees. • ≥60%-<90% black spruce, 31-50%, 51-70%, and 71-100% canopy closure, 10-19-m-high trees. • ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 10-19-m-high trees. • ≥90% tamarack, ≥80 years old, all canopy closures, 0-19-m-high trees.

• All other stands including immature stands (<40 years) and cutblocks. Avoided (low-quality polygons) Proulx Page 87 and pine plantations; and 7) linear features (LF) such as roads, Results seismic lines, pipelines, and oil and gas installations. Four canopy Surveys were conducted during mild and harsh winter conditions closures were used: 6-30%, 31-50%, 51-70%, and 71-100%. The along 26 transects in 2013, and 13 transects in 2014 (Table 3). The great majority of stands were coniferous and canopy closure values 2014 winter conditions were generally milder than those of 2013, were representative of stand conditions throughout the year. Three and only 3 transects were surveyed at temperatures of ≤ -21oC. tree height classes were used: 0-9 m, 10-19 m, and ≥20 m. In In 2014, snow accumulations were deeper and there was a thick this study, forested stands were all >50 years old, and could be crust near surface, particularly in forest stands with closed canopy considered as being mature according to Stepnisky et al. (2011). cover. Snow conditions differed between 2013 and 2014 (see Results). 2013 Therefore, the analysis of track distribution was done separately Frequency of caribou tracks per polygon rating class for each year. Nevertheless, in order to understand the general A total of 190 individual caribou tracks were encountered in 9 distribution of caribou over 2 winters, and increase the track out of 26 transects (Figure 3). There was a significant difference sample size, datasets from both years were also pooled together between observed and expected frequencies of tracks per polygon and analysed. Distances of caribou tracks from linear features rating class ( =43.1, df: 5, P<0.001). Tracks were significantly less were measured on 1:7,000 forestry maps. Mean distances of frequent than expected in medium-value forest polygons (G=30.5, caribou tracks to seismic lines and pipelines in 2013 and 2014 were df: 1, P<0.001). Caribou showed no selection for any polygon compared with Student t-test (Zar 1999). rating class other than linear features (G=6.0, df: 1, P<0.02)

Observed

Expected

Figure 3. Distribution of tracks/polygon rating class of the Little Smoky woodland caribou in the Alberta Newsprint Company’s Forest Management Agreement area, late winter 2013, western Alberta, Canada. (*): significant difference between observed and expected frequencies, P<0.05.

Table 3. Transect characteristics and environmental conditions in the Little Smoky study area during late winters 2013 and 2014.

Parameters 2013 2014 Dates 15 January - 13 February 15 January - 6 February Number of transects surveyes 26 13 Total distance (m) 56,025 31,130 Number of caribou tracks 190 78 Range of To -34o to +6oC -28o to +5oC Snow depths (cm) - Openings and muskegs 35-50 60-100 - Forests 15-40 15-50 Snow crust No Yes Page 88 Proulx

Code Habitat Class M ≥90% Black spruce 60M 6-30% cc 10-19 m ≥60%-<90% black spruce, 5-30% canopy closure, 10-19-m-high trees 60M 31-50%/51-70% cc 10-19 m ≥60%-<90% black spruce, 31-50%, 51-70% and 71-100% canopy closure, 10-19-m-high trees L 80+ years 0-19 m ≥90% tamarack, ≥80 years old, all canopy closures, 0-10-m-high trees L 60-79 years 31-50% cc 0-10 m ≥90% tamarack, 60-79 years old, 31-50% canopy closures, 0-19-m high trees 60L 31-50% 0-19 m ≥60%-<90% tamarack, 31-50% canopy closure, 0-10-m-high trees 60L 6-30% 0-19 m ≥60%-<90% tamarack, 6-30% canopy closure, 0-19-m-high trees 60L 51-70% 0-10 m ≥60%-<90% tamarack, 51-70% canopy closure, 0-10-m-high trees PI 80+ years 6-30% cc ≥90% lodgepole pine, ≥80 years old, 6-30%, canopy closure PI 80+ years 31-50% cc 10-19 m ≥90% lodgepole pine, ≥80 years old, 31-50% canopy closure, 10-19-m-high trees PI 80+ years 51-70% cc >9 m ≥90% lodgepole pine, ≥80 years old, 51-70% canopy closure, >9-m-high trees PI 80+ years 71-100% cc 10-19 m ≥90% lodgepole pine, ≥80 years old, 71-100% canopy closure, 10-19-m-high trees PI 60-79 years 31-50%/71-100% cc ≥60%-<90% lodgepole pine, 31-50%/71-100% canopy closure 60PI M 6-30%cc 10-19 m ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 10-19-m-high trees 60PI M 31-50%cc 10-19 m ≥60%-<90% lodgepole pine with black spruce, 31-50% canopy closure, 10-19-m-high trees 60PI M 51-70/71-100% cc 10-19 m ≥60%-<90% lodgepole pine with black spruce, 51-70 and 71-100% canopy closure, 10-19-m-high trees PI (w/o) ≥60%-<90% lodgepole pine without black spruce C ML Mixed coniferous stands, <60% black spruce, tamarack or lodgepole pine, and ≥20 black spruce or tamarack C (w/o) Mixed coniferous stands and deciduous stands, <60% black spruce, tamarack or lodgepole pine, and <20 black spruce or tamarack IS Immature sites LF Linear features

40 Observed 35 Expected 30 25 20 15 10

umbercaribou of tracks 5 N

0 M 60M 6-30%60 M cc 31-50%/51-70% 10-19L 80+ myearsL 60-79 0-19 60Lcc yearsm 10-19 31-50% 31-50%60L m 6-30% 0-19 cc60L 0-10 m cc 51-70% 0-19 mPI 80+ m cc years PI0-10 80+ 6-30% yearsPI 80+cc 31-50% yearsPI 80+ 51-70% cc years 10-19PI 60-79 71-100% cc m >960PI(M) years m cc 31-50%/71-100%60PI(M) 10-19 6-30%cc m60PI(M) 31-50%cc 10-19PI 51-70/71-100%mcc (w/o) 10-19C(ML) m C(w/o) cc 10-19IS m LF

Habitat Classes

Figure 4. Distribution of Little Smoky woodland caribou tracks (n=190) according to habitat classes, late winter 2013, western Alberta, Canada. (*): significant difference between observed and expected frequencies,P <0.05. Proulx Page 89

Preferred Stands1

Stands Used as per Availability

Little Smoky Caribou Range

Figure 5. Predicted late-winter habitat distribution of the Little Smoky boreal woodland caribou population using a query based on 2013 surveys (see Table 2 for description of polygons).

Table 4. Distances of Little Smoky woodland caribou tracks from linear features, Alberta . Newsprint Company’s Forest Management Agreement area, late winters 2013 and 2014.

Anthropogenic infrastructures number of tracks (n) Mean distance – m Range of distances – m (standard deviation) 2013 ANC Road 89 267 (240) 0-598 Forestry road 4 82 (67) 25-149 Seismic line/Pipeline 72 120 (89) 0-355 Oil and Gas Installation & Non-forested 25 513 (312) 50-957 2014 Seismic line/Pipeline 74 78 (73) 1-357 Oil and Gas Installaion & Non-forested 4 51 (48) 27-123

(Figure 3). Eighteen of the tracks associated with linear features expected frequencies of tracks per habitat type ( =287.4, df: 19, were found along the ANC road right-of-way where caribou fed on P<0.001) (Figure 4). Tracks were significantly more frequent ground vegetation. Three caribou also used a seismic line to travel than expected (G>3.9, df: 1, P<0.05), or were present as expected between forest stands. (P>0.05) based on habitat availability, in ≥60%-<90% tamarack Frequency of caribou tracks per habitat class or black spruce muskegs, and in most ≥60%-<90% lodgepole pine There was a significant difference between observed and stands with black spruce (Figure 4). Tracks were, however, absent Page 90 Proulx from pure black spruce and pure lodgepole pine stands, and from all in 1 seismic line located in between 2 habitats that were used 60-79-years-old pure tamarack stands (Figure 4). Caribou tracks by caribou. Tracks were significantly more frequent than expected were more frequent than expected in immature sites (G= 4.3, df: (G=18.4, df: 1, P<0.001) in medium-quality polygons, and 1, P<0.05) and on linear features (G = 6.0, df: 1, P<0.02), due to significantly less frequent than expected (G=32.9, df: 1, P<0.001) the presence of 5-8 caribou travelling together or feeding on the in low-quality polygons (Figure 6). Caribou used “high-quality” ground vegetation near forest edge (Figure 4). polygons as they encountered them (P>0.05; Figure 6). The distribution of caribou tracks/polygon rating class suggested Frequency of caribou tracks per habitat type that the draft query that was developed using Stepnisky et al.’s There was a significant difference between observed and (2011) HSI was inadequate to predict the late-winter distribution expected frequencies of tracks per habitat type ( =125.1, df: 7, of the Little Smoky woodland caribou. The 2013 late-winter P<0.001). Tracks were significantly more frequent than expected distribution of caribou tracks/habitat class (Figure 4) was used to (P<0.05) in muskegs with ≥60%-<90% black spruce, and ≥60%- develop a new query (Table 2). The resulting predictive distribution <90% tamarack (G ≥8.6, df: 1, P<0.01; Figure 7). Caribou used map suggested that late-winter habitats used by woodland caribou ≥90% tamarack muskegs, and ≥60%-<90% lodgepole pine stands in 2013 occurred in highly fragmented landscapes (Figure 5). with black spruce, closed canopy and large trees, as per availability Distances of woodland caribou tracks from linear features (P>0.05). Otherwise, caribou avoided (P<0.05) all upland Woodland caribou tracks were recorded on or near seismic lines lodgepole pine stands (Figure 7). There were no caribou tracks and pipelines (0-355 m), and oil and gas installations (50-957 m) in ≥90% black spruce stands. Caribou used seismic lines as they (Table 4). encountered them, i.e., without attraction or avoidance (Figure 7). 2014 Distances of woodland caribou tracks from linear features Frequency of caribou tracks per polygon rating class Woodland caribou tracks were recorded on or near seismic lines Seventy-eight tracks were encountered in 5 out of 13 transects. and pipelines (1-357 m), and oil and gas installations (27-123 m) There was a significant difference between observed and expected (Table 4). frequencies of tracks per polygon rating class ( =104.9. df: 2, Frequency of caribou tracks per habitat type – 2013 & 2014 P<0.001). Only 3 tracks were recorded in low-quality polygons, A total of 268 tracks were recorded in 2 years. There was a

Figure 6. Distribution of tracks/polygon rating class of the Little Smoky woodland caribou in the Alberta Newsprint Company’s Forest Management Agreement area, late winter 2014, western Alberta, Canada. (*): significant difference between observed and expected frequencies, P<0.05. Proulx Page 91

Code Habitat Class M ≥90% Black spruce, all canopy closures and tree heights 60M ≥60%-<90% black spruce, all canopy closures and tree heights L ≥90% tamarack, all canopy closures, all tree heights 60L ≥60%-<90% tamarack, all canopy closures, all three heights 60PIM ≥60%-<90% lodgepole pine with black spruce, 51-70% canopy closure, >10-m-high trees PIS (w/o) ≥90% lodgepole pine, pine stands without black spruce, and mixed coniferous stands with <20% black spruce PIS ≥60%-<90% lodgepole pine with presence of black spruce, 6-50% canopy closure, and all tree heights, and mixed coniferous stands with ≥20% black spruce IS-LF Immature stands and linear features

40 Observed 35 Expected 30

15 umbercaribou of tracks

N 10

0 M 60M L 60L 60PI M PISw/o PIS IS-LF

Habitat Classes

Figure 7. Distribution of Little Smoky woodland caribou tracks (n=78) according to habitat types, late winter 2014, western Alberta, Can- ada. (*): significant difference between observed and expected frequencies,P <0.05. significant difference between observed and expected frequencies with 6-30% and 51-70% canopy closure, and ≥11-m-high trees. of tracks per habitat type ( =575.8, df: 18, P<0.001) (Table 5, Caribou avoided all other stands (Table 5, Figure 8). Figure 8). Tracks were significantly more frequent than expected Distances of woodland caribou tracks from linear features (P<0.05) in muskegs with ≥60%-<90% black spruce stands with In 2013, woodland caribou track average distances from seismic 31-50% canopy closure, and or ≥60%-<90% tamarack stands lines and pipelines ranged from 0 to 355 m (Table 4). In 2014, the with 30-50% canopy closure and 0-19-m-high trees, ≥60%- distance of tracks from these linear features ranged from 1 to 357 <90% lodgepole pine with black spruce stands (71-100% canopy m (Table 4). On average, caribou tracks were significantly closer closure, 11-19-m tree heights), and mixed coniferous stands with (t = 3.1, P<0.005) to seismic lines and pipelines in 2014 than in ≥20%-<60% black spruce or tamarack (6-30% canopy closure, 2013. 10-19-m-high trees) (Table 5, Figure 8). Finally, caribou tracks discussion were more frequent than expected in non-forested, immature stands, and linear features (Table 5, Figure 8). This study confirmed my hypothesis that caribou prefer mixed Caribou used as per availability all other ≥60%-<90% black tamarack and black spruce muskegs with ≤20-m-high trees and spruce and tamarack muskegs, ≥60%-<90% lodgepole pine stands >30% canopy closure, and lodgepole pine-tamarack-black spruce Page 92 Proulx

Table 5. Habitat type rating based on woodland caribou track distribution in the Alberta . Newsprint Company’s Forest Management Agreement area, late winters 2013 and 2014.

Habitat type Code Significant G Values Preferred stands ≥60%-<90% black spruce, 31-50% canopy closure, all tree heights 60M(31/50) 26.0, P<0.001 ≥60%-<90% tamarack, 31-50% canopy closure, 0-19-m-high trees 60L(31/50) 21.1, P<0.001 ≥60%-<90% lodgepole pine with black spruce, 71-100% canopy closure, 60PLM(71/100) 7.9, P<0.01 11-19-m-high trees Mixed coniferous stands, <60% black spruce, tamarack or CML(6/30) 18.3, P<0.001 lodgepole pine, and ≥20% black spruce or tamarack, 6-30% canopy closure, 11-19-m-high trees Non-forested sites, immature stands, and linear features IS-LF 10.2, P<0.01 Used as per availability ≥60%-<90% black spruce, 6-30%, all tree heights 60M(6/30-51/70) P>0.05 ≥60%-<90% black spruce, 51-70 and 71-100, 0-19-m-high trees 60M(51-100) P>0.05 ≥60%-<90% tamarack, 6-30%, 0-19-m-hight trees 60L(6/30) P>0.05 ≥60%-<90% tamarack, 51-70% canopy closure, 0-10-m-hight trees 60L(51-70) P>0.05 ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 60PLM(6/30-11/19) P>0.05 11-19-m-high trees ≥60%-<90% lodgepole pine with black spruce, 51-70% canopy closure, 60PLM(51/70) P>0.05 ≥11-m-high trees Avoided ≥90% black spruce M 0 tracks ≥90% tamarack L 6.8, P<0.01 ≥90% lodgepole pine, 6-30% and 31-50% canopy closure, ≥ 11-m-high trees PL(6/50) 19.8, P<0.001 ≥90% lodgepole pine, 51-70% canopy closure, ≥ 11-m-high trees PL(51/70) 18.3, P<0.001 ≥90% lodgepole pine, 71-100% canopy closure, ≥ 11-m-high trees PL(71/100) 16.1, P<0.001 ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 60PLM(6/30-20) 0 tracks ≥20-m-high trees ≥60%-<90% lodgepole pine with black spruce, 31-50% canopy closure, 11-19-m-high trees 60PLM(31/50) 5.9, P<0.05 Mixed coniferous stands, <60% black spruce, tamarack or lodgepole pine, CML(31) 11.0, P<0.01 and ≥20% black spruce or tamarack, ≥31% canopy closure, all tree heights Mixed and deciduous stands, <60% black spruce, tamarack or lodgepole pine, PLSw/o 0 tracks and <20% black spruce or tamarack, and ≥60%-<90% lodgepole pine without black spruce mixed stands. It also showed that the new query was adequate to on 1 seismic line located within a medium-quality polygon where predict late-winter habitat use by the Little Smoky boreal woodland signs of caribou feeding and bedding were recorded. The difference caribou inhabiting the ANC’s FMA area. In 2014, most caribou in the distribution of tracks among high- and medium-quality tracks were found in polygons that had been rated as either high- polygons between 2013 and 2014 appears related to different or medium-quality based on forest stand characteristics only. Only environmental conditions. In 2014, there was twice more snow on 3 tracks were found outside these polygons, and they were present the ground than in 2013. Also, in upland stands, the combination Proulx Page 93

Code Habitat Class Code Habitat Class M ≥90% black spruce PI (71-100) ≥90% lodgepole pine, 71-100% canopy closure, ≥11-m-high trees 60M (6/30) ≥60%-<90% black spruce, 6-30%/51-70%, all tree heights 60PIM (6/30-11/19) ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 11-19-m-high trees 60M (31) ≥60%-<90% black spruce, 31-50% canopy closure, all tree heights 60PIM (6/30-20) ≥60%-<90% lodgepole pine with black spruce, 6-30% canopy closure, 11-19-m-high trees 60M (51-100) ≥60%-<90% black spruce, 51-70 and 71-10, 0-19-m-high trees 60PIM (31/50) ≥60%-<90% lodgepole pine with black spruce, 31-50% canopy closure, 11-19-m-high trees L ≥90% tamarack 60PIM (51/70) ≥60%-<90% lodgepole pine with black spruce, 51-70% canopy closure, ≥11-m-high trees 60L (6/30) ≥60%-<90% tamarack, 6-30% canopy closure, 0-19-m-high trees 60PIM (71/100) ≥60%-<90% lodgepole pine with black spruce, 71-100% canopy closure, 11-19-m-high trees 60L (31/50) ≥60%-<90% tamarack, 31-50% canopy closure, 0-19-m-high trees CML (6/30) Mixed coniferous stands, <60% black spruce, tamarack or lodgepole pine, and ≥20% black spruce or tamarack, 6-30% canopy closure, 11-19-m-high trees 60L (51-70) ≥60%-<90% tamarack, 51-70% canopy closure, 0-10-m-high trees CML (31) Mixed coniferous stands, <60% black spruce, tamarack or lodgepole pine, and ≥20% black spruce or tamarack, ≥31% canopy closure, all tree heights PI (6/50) ≥90% lodgepole pine, 6-30% and 31-50% canopy closure, ≥11-m-high PIS w/o Mixed and deciduous stands, <60% black spruce, tamarack or lodgepole pine, and ≥20% black spruce trees or tamarack, and ≥60%-<90% lodgepole pine without black spruce PI (51-70) ≥90% lodgepole pine, 51-70% canopy closure, ≥11-m-high trees IS-LF Non-forested sites, immature stands, and linear features

umbercaribou of tracks 20 N

10 60PIM(31/50) CML(6/30) CML(31) PISw/o IS-LF 60M(6/30) L PI(6/50) 60M(31) 60M(51-100) 60L(6/30) 60L(31/50) PI(71/100) 60PIM(6/30-20) 60PIM(51/70) 60PIM(71/100) M PI(51/70) 60PIM(6/30-11/19) 0 60L(51-70)

Habitat Classes

Figure 8. Distribution of Little Smoky woodland caribou tracks (n=268) according to habitat classes, late winters 2013 and 2014, western Alberta, Canada. (*): significant difference between observed and expected frequencies, P<0.05.

of deep snow and heavy crust interfered with caribou feeding on used as per availability; the value of these stands for caribou may terrestrial lichens. As a result, some mixed coniferous stands and change, but still remain acceptable from year to year based on ≥60-<90% lodgepole pine stands with black spruce that had been winter severity. For this reason, basing conservation programs on rated as high-quality habitats in 2013, were less used in 2014; models that neglect the influence of potentially important resources however, they were still being used as per availability. Caribou that are used by caribou in proportion to their availability may be re-directed their activities to specific black spruce and tamarack misleading (Neufeld 2006). muskegs that provided them with food (e.g., arboreal lichens) The 2-year study in ANC’s FMA area led to the identification and cover, many of them having been rated as medium-quality of muskegs and upland stands that are favourable to caribou when in 2013. These findings point out the importance of managing snow accumulations are slightly above average as in 2013, or deep landscapes to conserve preferred habitat types, and those that are as in 2014. During both years, caribou preferred muskegs with Page 94 Proulx

≥60-<90% black spruce and ≥60-<90% tamarack muskegs, 31-50% caribou with food and cover, and a good line of sight to detect canopy closure and ≤19-m-high trees (Figures 9 and 10). However, predators. This is in agreement with Hervieux et al. (1994) and they also used other ≥60-<90% black spruce and ≥60-<90% Neufeld (2006) who indicated that mixed lodgepole pine stands tamarack muskegs with different canopy closure and tree heights with black spruce were valuable to caribou. However, when snow as they encountered them. All these muskegs offered cover to is deep or there is a thick crust on surface, caribou may not be caribou to move around and bed down (Figure 10), some terrestrial able to access terrestrial lichens in these habitats. Canopy closure lichens growing at the base of conifer clusters (Proulx, personal and tree height impact on habitat use by caribou, and in ANC’s observations), and accessible arboreal lichens. These findings are in FMA area, Kansas and Brown (1992) found that lichen growth agreement with Proulx (2013) who found that woodland caribou of was very irregularly distributed on the landscape with large areas northwestern Saskatchewan preferred stands with ≥60% tamarack supporting little or no lichen. Ideal terrestrial lichen conditions with >20% canopy closure and ≤20-m high trees. Neufeld (2006) are characterized by 60-100 year old lodgepole pine forest with an also found that caribou in the Little Smoky range showed strong open to semi-open canopy, located on rapidly drained sites with selection for black spruce and tamarack at the 1-km scale sandy soil textures (Kansas and Brown 1992). Caribou also preferred mixed coniferous stands with ≥20-<60% Contrary to expectations of Stepnisky et al. (2011) and black spruce that had structural characteristics similar to those Environment Canada (2012), caribou avoided pure (≥90% in the of the preferred muskegs. During both years, caribou preferred present study) black spruce or lodgepole pine stands in ANC’s lodgepole pine stands with black spruce that had high canopy FMA area. Travel is difficult and the line of sight is poor in pure closure and 11-19-m-high trees (Figure 11). Such habitats provide black spruce stands (Figure 12); food is sparse or absent in pure

Figure 9. Preferred muskeg with ≥60-<90% tamarack muskegs, 31-50% canopy closure and ≤19-m-high trees, Little Smoky caribou population, western Alberta, Canada. Proulx Page 95

Figure 10. Example of a ≥60-<90% black spruce muskeg where Little Smoky caribou fed and bedded down near a seismic line, late-winter 2014. lodgepole pine stands on typical moderately well-drained soils This study suggests that caribou use immature stands, non- (Figure 13). In 2013, muskegs with ≥90% tamarack had been used forested sites and linear features that are adjacent or within high- as per availability. In 2014, however, they were avoided. Field and medium-quality polygons mainly for travel. As caribou use observations suggest that, in these habitats, lichen forage (access muskegs, they extend their movements into nearby adjacent upland and quantity) often is limited. During both years, mixed stands and disturbed area (Proulx 2013). In this study, the selection of without black spruce, white spruce, and deciduous stands were linear features by caribou was due to a large number of individuals avoided by caribou. This is in agreement with Neufeld’s (2006) tracks recorded along the ANC Road where caribou used findings. Such stands usually have a well-developed understory intensively the south-facing grassy right-of-way in 2013. Neufeld that interferes with caribou movements, and they offer little or no (2006) reported that caribou preferred grass-covered locations to food. Caribou may also avoid these stands to distance themselves forested locations. In 2014, caribou did not show any avoidance of from other ungulates and large predators who prefer such forests linear features. In fact, some animals fed and bedded down beside (Proulx and Kariz 2005). one of them. Wasser et al. (2011) found that negative effects on the physiological health of caribou were related to human activity. This Page 96 Proulx

Figure 11. Preferred lodgepole pine stands with black spruce with high canopy closure and 11-19-m-high trees, Little Smoky caribou population, western Alberta, Canada. suggests that not all linear features are the same. Two corridors rich in browse, and avoided lodgepole pine stands, as found by of similar shape and size may have different impacts on caribou Proulx and Kariz (2005). Overall, in late winter, muskegs may because of different environmental conditions. Some linear protect caribou from competition and predation (McLoughlin features get more human traffic, some are newly constructed, and et al. 2005). Tamarack-dominated muskegs may be particularly others have been built many decades ago and become part of the important in the conservation of woodland caribou since they are ambient environmental conditions. In 2014, caribou tracks were unsuitable for logging and most of them remain undisturbed by significantly closer to seismic lines than in 2013. Greater snow wildfire (Proulx 2013). accumulation in 2014 resulted in lesser use of seismic lines by other A new predictive late-winter habitat suitability distribution map ungulates and predators. In 2014, moose travelled with difficulty (Figure 14) was developed using the pooled 2013 and 2014 track on unopened seismic lines, sinking as much as 90 cm (Proulx, survey data (Table 5). This map suggests that habitats that are unpublished data). Deer also avoided these seismic lines and the preferred or used as per availability by caribou are not as extensive muskegs used by caribou. In 2013, I found that moose travelled on as the high-quality areas that had been determined using the seismic lines that crossed muskegs, but they rarely ventured deep initial query based on Stepnisky et al.’s (2011) HSI. With the into tamarack stands. Moose preferred mixed coniferous stands new query, polygons that are preferred or used as per availability Proulx Page 97

Figure 12. Caribou of the Little Smoky population avoided pure black spruce stands in late winters 2013 and 2014, western Alberta, Canada. Page 98 Proulx

Figure 13. Caribou of the Little Smoky population avoided pure lodgepole pine stands in late winters 2013 and 2014, western Alberta, Canada. by caribou amount to 28,183 ha (20%) of the 143,932 ha of the recovery strategies that have been endorsed by various organizations Little Smoky population range included in the ANC’s FMA area. need to be reviewed to properly assess the existing suitability and High-suitability caribou range is naturally highly fragmented potential of landscapes inhabited by the Little Smoky caribou and adequate habitats are scattered across the landscape, which population. Environment Canada (2012) considers that, in order to suggests that the natural carrying capacity of the FMA area for achieve self-sustainability of a caribou population, landscapes and the boreal woodland caribou is relatively low. The size of the anthropogenic disturbances must be restored to provide landscapes caribou population inhabiting this area may be naturally limited with at least 65% of undisturbed land. In the calculation of this by the amount of available late-winter habitat. Finally, the new 65%, cumulatively, the total disturbed area that is avoided by predictive distribution map suggests that preferred and suitable boreal caribou includes the anthropogenic footprint and a 500-m caribou habitat may be found outside the caribou range recognized buffer (therefore a seismic line compromises a 1-km-wide corridor by Alberta Environment (Figure 14) on the basis of Smith (2004). of land), and areas where a fire has occurred in the past 40 years (no Indeed, I found caribou tracks east of the boundaries of the caribou buffer applied). The remaining habitat within a range is considered range. More investigations outside the recognized range should be undisturbed. At the landscape level, intact areas (i.e., without conducted to determine the extent of the distribution of the boreal any anthropogenic or natural disturbance) of >200 ha (Alberta caribou population. Sustainable Resource Development, undated), and even >1,000 ha (West Central Alberta Caribou Landscape Planning Team 2008), Management implications with >80 years-old stands have been suggested. On the basis of The results of this study suggest that current boreal caribou 2008 satellite imagery, timber harvesting footprint in west-central Proulx Page 99

Alberta caribou ranges would currently cover 38% of the Little (2005), and Neufeld (2006). In the present study, caribou did not Smoky range (ASRD/ACA 2010). Also, linear feature density show any sign of avoidance of linear features that were located (mostly seismic lines) was estimated at 3.7 km/km2 in the Little within habitats that they preferred or used as per availability. Many Smoky range. Using Environment Canada’ s (2012) minimum anthropogenic disturbances in the ANC’s FMA area occurred >15 amount of undisturbed habitat, and the concept of large intact years ago, and today`s caribou were born in ecosystems that were areas with >80 years-old, the landscapes that are actually part of already disturbed by human activities. This does not mean that the Little Smoky caribou range would have no potential for the caribou may not be affected by new habitat alterations and linear species. Yet, this study shows that caribou inhabit the muskegs and features. However, it is possible that old seismic lines do not affect adjacent upland habitats found in highly fragmented landscapes. the Little Smoky woodland caribou in the selection of their late- Caribou would, of course, benefit from less fragmented landscapes winter habitats. and larger habitat patches. However, this will not happen in the At a landscape level, caribou require habitat patches that are short-term. The value of the current habitat conditions must interconnected (O’Brien et al. 2006). My field observations suggest therefore be assessed according to the actual caribou distribution, that the Little Smoky caribou travel between muskegs using habitat use, and behaviour during the critical late-winter period. adjacent upland stands (e.g., Figure 11). Proulx (2013) also found The minimum size of an area that is used by caribou on a daily that boreal caribou used upland stands to travel between muskegs, basis is not well understood. Proulx (2013) reported the presence particularly when the upland forests encompassed a coulee or of caribou in muskegs of all sizes. Similarly, in the Little Smoky lowlands interconnecting bogs and fens. I suggest that caribou range, caribou tracks were found in small and large muskegs do not travel through extensive upland areas to reach “preferred” bisected by seismic lines. Habitat use by caribou has been lodgepole pine or mixed coniferous stands with black spruce. confirmed within 250 m of seismic lines by Dyer (1999), Saher Caribou likely use these stands when they are adjacent to high-

Figure 14. Predicted late-winter habitat distribution of the Little Smoky boreal woodland caribou population based on 2013 and 2014 track surveys (see Table 5 for description of polygons). Page 100 Proulx and medium-quality black spruce and tamarack muskegs (Proulx 2 types of corridors: 2013), particularly if they support lichen (Kansas and Brown 1992). - Type 1: Corridors that interconnect late-winter habitat patches, Preferred upland stands that are located ≤2 km from valuable i.e., muskegs and adjacent upland stands (≤2 km) that are preferred muskegs should therefore be protected. Such stands could provide or used as per availability by caribou. (Figure 15); and caribou with protection while traveling between muskegs, and with - Type 2: Riparian corridors that should be protected for terrestrial lichens in early winter, and later if snow accumulations biodiversity conservation. These include corridors along are not considerable. watercourses, gullies, marshes, etc. (Raedeke 1989; Naiman et al. The primary factors causing extinctions are habitat loss and 1993; Proulx 2013) (Figure 15). However, because moose feed on habitat fragmentation (e.g., Andrén 1994; Brooks et al. 2002; riparian vegetation, and wolves use riparian corridors to travel and Butchart et al. 2010; Haddad et al. 2015), and the recovery of the hunt (Latham et al. 2011), riparian corridors may be disconnected Little Smoky caribou population must start with the conservation from nearby high-quality habitats to decrease predation risks for of valuable habitats. Boreal caribou do not migrate, and caribou, and their vegetation should be modified to reduce their movements between calving and winter ranges are relatively short suitability for moose and deer. so that seasonal home ranges overlap (Brown et al. 2000; Rettie Habitat management zones and Messier 2001; Dyke 2008). Conserving late-winter habitats The distribution of late-winter habitats and movement corridors identified in this study may actually provide caribou with spring may be used to identify different management zones (Figure 15). and summer habitats, and most of the critical elements required for • High-priority zone: This zone corresponds to late-winter the persistence of their population. In other regions of the Boreal habitats that are selected or used as per availability by caribou. Plains Ecozone, caribou calving habitat has been associated with This zone encompasses tamarack and black spruce muskegs that muskegs. In the Smoothstone-Wapaweka region of Saskatchewan, are used by caribou, and mixed coniferous stands (with ≥20% caribou selected peatlands and black spruce-dominated stands black spruce or tamarack) that are located ≤2 km from these for calving (Rettie 1998). In the Wabowden area of Manitoba, muskegs. I recommend that logging be prohibited within the caribou selected black spruce- or tamarack-treed muskegs (Hirai high-priority zone. Oil & gas activities should be limited to 1998). Late-winter caribou habitats identified in this study provide already existing seismic lines, and should not be conducted when the animals with food, protection, and escape cover, i.e., necessary snow accumulations in seismic lines may be compacted by vehicles, conditions for ideal calving habitat (Morash and Racey 1990). which would allow predators to travel deep into caribou muskegs. No new seismic lines, pipelines or roads should be established in Habitat conservation plan the high-priority zone. Movement corridor networks • Medium-priority Zone: This zone is a buffer zone, which Natural and anthropogenic disturbances break landscapes into protects the high-priority zone from man activities and a series of disconnected areas comprised of old and immature permanent infrastructures, or interconnects late-winter stands, and upland and lowland habitats. To counteract the effects habitats. Not more than 30% of medium-priority areas that fragmentation may have on caribou movements, connectivity should be impacted by harvest activity within 120 years. corridors need to be identified or established to facilitate Timber harvesting should aim to increase the functionality movements within and between landscapes (Rosenberg et al. 1995; of movement corridors. Harvesting and silvicultural methods Racey and Arsenault 2007). Here, I define corridors as linear should promote the retention of terrestrial lichens without landscape portions that facilitate movement (Noss 1987; Merriam jeopardizing the hydrological characteristics of the region. 1991) between valuable late-winter habitats. Deciduous regeneration in plantations should be controlled The critical features of a movement corridor may vary from to minimize the production of browse or grass for competing one region to another, but independently of their width, length ungulates. Upland stands that may be avoided by caribou, but and structure, all corridors must fulfill their role as facilitators used by other ungulates, should be modified through vegetation of movement. Therefore, corridors must be easily accessible and management to decrease habitat use by predators, deer, elk and recognizable by caribou, and they must allow them to travel moose near high-priority zones. efficiently across landscapes (Rosenberget al. 1995). Knowing that • Low-priority Zone: This zone includes habitats that are not late-winter habitats that are preferred or used as per availability used by caribou either because of their composition and structure, by caribou represent only 20% of ANC’s FMA area, one negative or they are located >2 km from late-winter habitats and their effect of the highly fragmented habitats may be an increase in connectivity corridors. There are no harvest limits other than those distances between valuable muskegs (see Andrén 1994). Therefore, determined by AESRD Lands & Forests. However, cut block size identifying connectivity corridors that will ensure caribou and distribution should not compromise the buffering effect of movements within and between landscapes is essential. I recognize the medium-priority zone. The low-priority zone should never be Proulx Page 101 Figure 15. Example of a landscape with caribou movement corridors and habitat management zones.management habitat landscapecorridorsand a withcaribou Figure Exampleof movement 15. Page 102 Proulx adjacent to the high-priority zone. the dispersal, germination and establishment of tamarack, Selected harvest and silviculture practices and planting techniques according to hydrology and water The major effect of timber harvesting on caribou, other than chemistry (e.g., Duncan 1954; Brown et al. 1988; Mietti et al. removing suitable habitat that provides protection against bad 2005). weather and predation, is the loss of terrestrial and arboreal lichens. - Restoration of seismic lines at muskeg level – Seismic This is particularly true in clearcuts where arboreal lichens are lines crossing valuable muskegs may be restored by planting lost over large expanses. However, extensive research in British tamarack and black spruce seedlings (depending on the Columbia and USA demonstrated that partial cutting (group composition of the stand) to eliminate man-made gaps within or individual tree selections) can be used to harvest mountain stands. pine beetle (Dendroctonus ponderosae)-killed stands and create - Management of coniferous stands adjacent to valuable environmental conditions that are favourable to both arboreal and muskegs by opening up black spruce stands (Figure 15). terrestrial lichens (Rominger et al. 1994; Delong et al. 1999; Miège - Restoration of seismic lines at their entry points in et al. 2001a, b). Partially harvested stands might have adequate muskegs with thick plantations of mixed conifer trees, so that arboreal and terrestrial lichen growths approximately 15 years predator and human access to muskegs via seismic lines may after harvest, a time period similar to what has also been observed be hindered (e.g., Charlebois et al. 2015). in some clearcuts (e.g., Maykawa and Kershaw 1976; Snyder and Mixed coniferous stand management – Mixed coniferous stands Woodard 1992; Coxson et al. 1999; Sulyma 2000; Seip and Jones with ≥20% black spruce are valuable habitats for caribou in late 2009), but markedly shorter than the ≥70-year period reported winter. However, some of these stands may not be usable because in burns for arboreal (e.g., Goward and Campbell 2005) and of recent blow-downs. Fallen trees should be removed near the terrestrial (Johnson 1981; Klein 1982 ; Morneau and Payette 1988; edge of muskegs to allow free movements by caribou. Rowe 1984) lichens. Medium-priority zone Although assessment data are lacking, various vegetation Stands of the medium-priority zone will act as a buffer to management strategies (e.g., high density planting of coniferous protect the high-priority zone from extensive industrial activities. trees, restoration of disturbed sites such as seismic lines, and Management actions proposed for mixed coniferous stands in vegetation control – West Central Alberta Caribou Landscape the high-priority zone may be applied here. Stands at the border Planning Team 2008; ANC 2011) have been recommended in of muskegs can be managed to promote growth of arboreal and the past to decrease habitat quality for competing ungulates and terrestrial lichens. Finally, stands of the medium-priority zone to minimize access by predators. However, to my knowledge, no will serve as connectivity corridors between functional habitats. recommendations have been proposed to further improve existing Whenever necessary, these stands should be managed to facilitate caribou habitats and incite animals to extend their activities in caribou movements and minimize resistance to movements adjacent areas. There is therefore a need to integrate strategies that resulting from blow-downs, the growth of alder thickets, or the may improve the quantity and quality of caribou habitats. presence of bare grounds resulting from past anthropogenic Vegetation management strategies for the Little Smoky caribou activities. The vegetation of recent (<20 years old) cut blocks that population range are present within the medium-priority zone should be managed Harvest systems and silviculture practices will vary according to (e.g., vegetation control) to minimize utilization by moose, elk and management zones. deer (e.g., Racey et al. 1991; Rea 2003). Finally, timber harvesting High-priority zone (30% per 120-year periods) should employ the use of partial or Muskeg management – Muskegs change with time and while group selection cuts to promote the maintenance of terrestrial and some of them develop into high-quality habitats for caribou, others arboreal lichens (Figure 15; Coxson et al. 2015). become less desirable due to a loss of accessible arboreal lichens, Low-priority zone a greater predominance of a single species (e.g., black spruce), or Industrial activities in the low-priority zones are not subject a change in hydrological characteristics that jeopardize tamarack to restrictions. However, clearcuts in this zone should be stands and ground vegetation. The following management actions amalgamated to minimize the creation of forest mosaics, which may be considered to maintain or improve muskeg areas that are provide moose, elk and deer with protection cover and food currently used by caribou: (Proulx and Kariz 2005; Proulx 2008). When in proximity to - Increase the size of valuable muskegs by establishing the medium-priority zone, vegetation should be controlled in cut tamarack seedlings in muskegs that are avoided by caribou blocks to minimize food quantity and quality for ungulates (Racey because they have ≥90% black spruce, or lichen-rich branches et al. 1991; Rea 2003). are out of reach, etc. (Figure 15). Some of this work may be Adaptive Management empirical and will require adopting techniques associated with Adaptive management refers to the structured process of Proulx Page 103 adjusting management in response to implementation of a Canada. monitoring program to test stated hypotheses, and revision of Alberta Environment and Sustainable Resource Development management based on the monitoring results (Dzus et al. 2010). (AESRD). Undated. Appendix 5: Caribou habitat When modifying the composition and structure of muskegs and assessment. Available at: http://srd.alberta.ca/LandsForests/ forest stands, or when restoring seismic lines, it is essential to ForestManagement/ForestManagementPlans/documents/ formulate hypotheses and assess the success of these management Forest ManagementUnitE8/Appen5_Caribou.pdf. Accessed 2 actions. Through the monitoring of the distribution of animals July 2013. across landscapes, and the use of new and old habitats by animals, Alberta Newsprint Company (ANC). 2011. Mitigation of management actions may be modified and adapted to local forest management impacts on caribou habitat. Mimeograph, environmental characteristics to meet the habitat requirements of Whitecourt, Alberta, Canada. the species. Andrén, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable Conclusion habitat: a review. Oikos 71: 355-366. The alteration and destruction of late-winter habitats of the Brooks, T. M., R. A. Mittermeier, C. G. Mittermeier, G. A. B. Little Smoky caribou population have occurred over decades, and da Fonseca, A. B. Rylands, W. R. Konstant, P. Flick, J. Pilgrim, landscapes are now significantly fragmented. The restoration of S. Oldfield, G. Magin, and C. Hilton-Taylor. 2002. Habitat these habitats will take decades to accomplish. However, this loss and extinction in the hotspots of biodiversity. Conservation study identified muskeg habitats that are preferred by caribou in Biology 16: 909-923. late winter, and possibly during the calving period. Therefore, Brown, K. G., C. Elliott, and F. Messier. 2000. Seasonal management actions should be taken now to conserve and expand distribution and population parameters of woodland caribou in these habitats. While this is being done, there is a need to adapt central Manitoba: implications for forestry practices. Rangifer industrial activities to a habitat management action plan where Special Issue No. 12: 85-94. priority must be given to the conservation of the late-winter Brown, K. R., D. B. Zobel, and J. C. Zasada. 1988. Seed habitats of the Little Smoky caribou population. dispersal, seedling emergence, and early survival of Larix laricina (DuRoi). Canadian Journal of Forest Research 18: 306-314. Acknowledgments Brown, W. K., and D. P. Hobson. 1998. Caribou in west-central This study was funded by ANC. I thank Rick Morse for Alberta – Information review and synthesis. Terrestrial & field assistance, Ryan Spooner and Steve Reed from The Aquatic Environmental Managers Ltd. Report submitted to Silvacom Group for the development of caribou habitat West-central Alberta Caribou Standing Committee, Calgary, predictive maps, and the ANC staff, particularly Greg Branton Alberta, Canada. and Ian Daisley, for their logistical support. I am grateful to Butchart, S. H. M., M. Walpole, B. Collen, A. Van Strien, J. Associate Editor Pauline Feldstein, and referees John Kansas P. W. Scharlemann, R. E. A. Almond, J. E. M. Baillie, B. and Paul Paquet for their comments on an earlier version of Bomhard, C. Brown, and J. Bruno. 2010. Global biodiversity: this manuscript. indicators of recent declines. Science 328: 1164. Campbell, J., and D. S. Coxson. 2001. Canopy microclimate and Literature cited arboreal lichen loading in subalpine spruce-fir forest. Canadian Ahti, T., and R. L. Hepburn. 1967. Preliminary studies on Journal of Botany 79: 537-555. woodland caribou range, especially on lichen stands, in Ontario. Carroll, S. B., and C. Bliss. 1982. Jack pine-lichen woodland Ontario Department of Lands and Forests, Research Report on sandy soils in northern Saskatchewan and northern Alberta. (Wildlife) No. 74, Toronto, Ontario, Canada. Canadian Journal of Botany 60: 2270-2282. Alberta Environment and Sustainable Resource Development Charlebois, M. L., H. G. Skatter, J. L. Kansas, and D. P. Crouse. (AESRD). 2014. Alberta Vegetation Inventory Standards and 2015. Using LiDar, coloured infrared imagery, and ground truth Data Model Documents. Available at: http://esrd.alberta.ca/ data for mapping and characterizing vegetation succession on lands-forests/vegetation-inventory-standards.aspx. Accessed 19 disturbance types: implications for woodland caribou (Rangifer January 2015. tarandus caribou) management. Canadian Wildlife Biology & Alberta Sustainable Resource Development and Alberta Management 4: 119-136. Conservation Association (ASRD/ACA). 2010. Status of the Cochran, W. J. 1954. Some methods for strengthening the woodland caribou (Rangifer tarandus caribou) in Alberta: update common χ2 tests. Biometrics 10: 417-451. 2010. Alberta Sustainable Resource Development, Wildlife Coxson, D. S. 2015. Using partial-cut harvesting to conserve Status Report No. 30 (Update 2010), Edmonton, Alberta, terrestrial lichens in managed landscapes. Canadian Wildlife Page 104 Proulx

Biology & Management 4: 150-162. D. J. Levey, C. R. Margules, B. A. Melbourne, A. O. Nicholls, Coxson, D. S., J. Marsh, and R. Sulyma. 1999. Factors affecting J. L. Orrock, D.-X. Song, and J. R. Townshend. 2015. Habitat terrestrial lichen abundance in woodland caribou habitat. fragmentation and its lasting impact on earth’s ecosystems. University of Northern British Columbia, Prince George, British Science Advances: e1500052. Columbia, Canada. Hervieux, D., J. Edmonds, R. Bonar, and J. McCammon. DeLong, D., J. Miller, B. McLellan, and E. Terry. 1999. Partial 1994. Successful and unsuccessful attempts to resolve caribou cutting prescriptions for mountain caribou in the ESSF of management and timber harvesting issues in west central southeastern BC – a retrospective investigation. FRBC report Alberta. Rangifer 9: 185-191. KB97, Nelson, British Columbia, Canada. Hirai, T. 1988. An evaluation of woodland caribou (Rangifer Duncan, D. P. 1954. A study of some of the factors affecting the tarandus caribou) calving habitat in the Wabowden area, natural regeneration of tamarack (Larix laricina) in Minnesota. Manitoba. MSc thesis, University of Manitoba, Winnipeg, Ecology 35: 498-521. Manitoba. Dyer, S. J. 1999. Movement and distribution of woodland caribou Kansas, J. L., and W. K. Brown. 1992. Preliminary lichen (Rangifer tarandus caribou) in response to industrial development habitat suitability assessment for the Little Smoky primary in northeastern Alberta. MSc thesis, University of Alberta, caribou winter range. Report prepared by Sentra Consultants Edmonton. Alberta, Canada. Ltd. and Terrestrial and Aquatic Environmental Managers Ltd. Dyke, C. 2008. Spatial and temporal characterization of for Alberta Newsprint Company Ltd., Whitecourt, Alberta, woodland caribou (Rangifer tarandus caribou) calving habitat in Canada. the boreal plains and boreal shield ecozones of Manitoba and Klein, D. R. 1982. Fire, lichens, and caribou. Journal of Range Saskatchewan. MSc thesis, University of Manitoba, Winnipeg, Management 35: 390-395. Canada. Latham, A. D. M., M. C. Latham, M. S. Boyce, and S. Boutin. Dzus, E., J. Ray, I. Thompson, and C. Wedeles. 2010. Caribou 2011. Movement responses by wolves to industrial linear features and the National Boreal Standard: report of the FSC Canada and their effect on woodland caribou in northeastern Alberta. Science Panel. Report prepared for FSC Canada, Toronto, Ecological Applications 21: 2854-2865. Ontario, Canada. Maikawa, E., and K. A. Kershaw. 1976. Studies on lichen- Edmonds, E. J. 1993. Winter distribution of woodland caribou in dominated systems. XIX. The postfire recovery sequence of black the Little Smoky area, November 1992 to March 1993. Alberta spruce-lichen woodland in the Abotau Lake Region, N. W. T. Fish and Wildlife Division, Edson, Alberta. Canadian Journal of Botany 54: 2679-2287. Environment Canada. 2012. Recovery strategy for the woodland McLoughlin, P. D., J. S. Dunford, and S. Boutin. 2005. Relating caribou (Rangifer tarandus caribou), Boreal population, in Canada. predation to broad-scale habitat selection. Journal of Animal Species at Risk Act Recovery Strategy Series. Environment Ecology 74: 701-707. Canada, Ottawa, Ontario, Canada. Merriam, G. 1991. Corridors and connectivity: animal Fancy, S. G., and R. G. White. 1985. Energy expenditures by populations in heterogeneous environments. Pages 133-142 in caribou while catering in snow. Journal of Wildlife Management D. A. Saunders and R. J. Hobbs, editors. Nature conservation 2: 49: 987-993. the role of corridors. Surrey Beatty & Sons, Norton, Australia. Farnell, R., and J. McDonald. 1988. The influence of wolf Miège, D. J., H. M. Armleder, M. J. Waterhouse, and T. Goward. predation on caribou mortality in Yukon’s Finlayson caribou 2001a. A pilot study of silvicultural systems for northern caribou herd. Proceedings North American Caribou Workshop 3: winter range. BC Ministry Forest Research Program, Working 52-70. Paper 56/2001, Victoria, British Columbia, Canada. Goward, T., and J. Campbell. 2005. Arboreal hair lichens in Miège, D. J., T. Goward, M. J. Waterhouse, and H. M. Armleder. a young mid-elevation conifer stand with implications for the 2001b. Impact of partial cutting on lichen diversity in lodgepole management of mountain caribou. The Bryologist 108: 427- pine forests on the Chilcotin Plateau of British Columbia. 434. Ministry of Forest Research Program, Working Paper 55/2001, Johnson, E. A. 1981. Vegetation organization and dynamics of Victoria, British Columbia, Canada. lichen woodland communities in the Northwest Territories, Mietti, T. E., C. E. Cameron, C. R. Picard, K. M. Mulac, A. Canada. Ecology 62: 200-215. Landaw, and L. H. Fraser 2005. Hydrology, water chemistry, Haddad, N. M., L. A. Brudvig, J. Clobert, K. F. Davies, A. and vegetation characteristics of a tamarack bog in Bath Gonzalez, R. D. Holt, T. E. Lovejoy, J. O. Sexton, M. P. Township, Ohio: towards restoration and enhancement. The Austin, C. D. Collins, W. M. Cook, E. I. Damschen, R. M. Ohio Journal of Science 105: 21-30. Ewers, B. L. Foster, C. N. Jenkins, A. J. King, W. F. Laurance, Morash, P. R., and G. D. Racey. 1990. The Northwestern Ontario Proulx Page 105

Forest Ecosystem Classification as a descriptor of woodland Ecography 23: 466-478. caribou (Rangifer tarandus caribou) range. Northwestern Ontario Rettie, W. J., and F. Messier. 2001. Range use and movement Technology Development Unit Technical Report No. 55. rates of woodland caribou in Saskatchewan. Canadian Journal Morneau, C., and S. Payette. 1988. Postfire lichen-spruce of Zoology 79: 1933-1940. woodland recovery at the limit of the boreal forest in northern Rominger, E. M., L. Allen-Johnson, and J. L. Oldemeyer. 1994. Quebec. Canadian Journal of Botany 67: 2770-2782. Arboreal lichen in uncut and partially cut subalpine fir stands Naiman, R. J., H. Decamps, and M. Pollock. 1993. The role in woodland caribou habitat, northern Idaho and southeastern of riparian corridors in maintaining regional biodiversity. British Columbia. Forest Ecology Management 70: 195-202. Ecological Applications 3: 209-212. Rosenberg, D. K., B. R. Noon, and E. C. Meslow. 1995. Toward Natural Regions Committee. 2006. Natural regions and a definition of biological corridor. Pages 436-446 in J. Bissonette subregions of Alberta. Compiled by D. J. Downing and W. W. and P. R. Krausman, editors. Integrating people and wildlife for Pettapiece. Government of Alberta. Publication no. T/852. a sustainable future. The Wildlife Society, Bethesda, Maryland, Neufeld, L. M. 2006. Spatial dynamics of wolves and woodland USA. caribou in an industrial forest landscape in west-central Alberta. Rowe, J. S. 1984. Lichen woodlands in northern Canada. Pages MSc Thesis, University of Alberta, Edmonton, Alberta, 225-237 in R. Olson, R. Hastings, and F. Geddes, editors. Canada. Northern ecology and resource management. University of Noss, R. F. 1987. Corridors in real landscapes: a reply to Alberta Press, Edmonton, Alberta, Canada. Simberloff and Cox. Conservation Biology 1: 159-164. Saher, D. J. 2005. Woodland caribou habitat selection during O’Brien, D., M. Manseau, A. Fall, and M.-J. Fortin. 2006. winter and along migratory routes in west-central Alberta. MSc. Testing the importance of spatial configuration of winter habitat thesis, University of Alberta, Edmonton, Alberta, Canada. for woodland caribou: an application of graph theory. Biological Seip, D., and E. Jones. 2009. Response of woodland caribou to Conservation 130: 70-83. partial retention logging of winter ranges attacked by mountain Proulx, G. 2008. Late-winter habitat use by mule deer, Odocoileus pine beetle. FIA/FSP Project # Y091010, 2008-09, Prince hemionus, in central interior British Columbia. Canadian Field- George, British Columbia, Canada. Naturalist 122: 205-211. Siegel, S. 1956. Nonparametric statistics for the behavioural Proulx, G. 2013. Late-winter habitat use by boreal woodland sciences. McGraw-Hill Book Co., New York, New York, USA. caribou (Rangifer tarandus caribou), in northwestern Smith, K. G. 2004. Woodland caribou demography and Saskatchewan. Canadian Wildlife Biology & Management 2: persistence relative to landscape change in west-central Alberta. 11-22. MSc thesis, University of Alberta, Edmonton, Alberta, Proulx, G., and R. Kariz. 2005. Winter habitat use by moose, Canada. Alces alces, in central interior British Columbia. Canadian Field- Smith, K. G., E. J. Ficht, D. Hobson, T. C. Sorensen, and D. Naturalist 119: 186-191. Hervieux. 2000. Winter distribution of woodland caribou in Racey, G. D., K. Abraham, W. R. Darby, H. R. Timmermann, relation to clear-cut logging in west-central Alberta. Canadian and Q. Day. 1991. Can woodland caribou and the forest Journal of Zoology 78: 1433-1440. industry coexist: the Ontario scene. Rangifer, Special Issue No. Snyder, J., and P. M. Woodard. 1992. Lichen regeneration 7: 108-115. rates in Alberta following various types of logging and wildfire Racey, G. D., and A. A. Arsenault. 2007. In search of a critical disturbance. University of Alberta report, Edmonton, Alberta habitat concept for woodland caribou, boreal population. Canada. Rangifer 17: 29-37. Sokal, R. R. and F. J. Rohlf. 1981. Biometry. 2nd edition. W.H. Raedeke, K. Editor. 1989. Streamside management: riparian Freeman and Company. San Francisco, California, USA. wildlife and forestry interactions. Institute Forest Resources, Stepnisky, D., K. Smith, C. Stambaugh, T. Powell, D. Hervieux, University of Washington, Seattle, Contribution No. 59. B. Wilson, and C. Hampel. 2011. Caribou Habitat Suitability Rea, R.V. 2003. Modifying roadside vegetation management Index (HSI). A GIS modeling procedure for the Lower practices to reduce vehicular collisions with moose Alces alces. Athabasca LUF Region. Unpublished mimeograph, Alberta Wildlife Biology 9: 81-91. Fish & Wildlife, Whitecourt, Alberta, Canada. Rettie, W. J. 1988. The ecology of woodland caribou in central Sulyma, R. 2000. Interim report: a retrospective study of Saskatchewan. PhD thesis, University of Saskatchewan, terrestrial lichen development in harvested areas located in the Saskatoon, Saskatchewan, Canada. Omineca Region of north central British Columbia. Forest Floor Rettie, W. J., and F. Messier. 2000. Hierarchical habitat selection Contracting Ltd. Report submitted to BC Ministry of Forests by woodland caribou: its relationship to limiting factors. and Slocan Forest Products Ltd., Vanderhoof, British Columbia, Page 106 Proulx

Canada. Thomas,D . C., and D. R. Gray. 2002. Update COSEWIC status report on the woodland caribou Rangifer tarandus caribou in Canada. Committee on the Status of Endangered Wildlife in Canada, Ottawa, Ontario, Canada. Wasser, S. K., J. L. Keim, M. L. Taper, and S. R. Lele. 2011. The influences of wolf predation, habitat loss, and human activity on caribou and moose in the Alberta oil sands. Frontiers in Ecology and Environment 9: 546-551. West Central Alberta Caribou Landscape Planning Team. 2008. West central Alberta caribou landscape plan. Mimeograph prepared for the Alberta Caribou Committee, Alberta, Canada. Zar, J. H. 1999. Biostatistical analysis. 4th edition. Prentice Hall, New Jersey, USA. About the author Gilbert Proulx, PhD, is Director of Science at Alpha Wildlife Research & Management, and Editor-in Chief of the Canadian Wildlife Biology & Management journal. Gilbert has 39 years of field experience as a wildlife biologist. He has published more than 130 scientific articles, 5 textbooks, and 8 field guides (species at risk). His main research interest focuses on mammals, particularly in forest and agriculture ecosystems, and on technology development, mainly on mammal trapping and detection methods.

Received 18 February 2015 – Accepted 28 May 2015