RANGIFER Research, Management and Husbandry of Reindeer and Other Northern Ungulates

15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

RANGIFER Research, Management and Husbandry of Reindeer and other Northern Ungulates

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License VolumeRangifer, 35, 35, Spec.2015 Iss. - No. Special 23, 2015 Issue EditorNo. in Chief: Birgitta23 Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 1 Rangifer

Publisher: Swedish University of Agricultural Sciences Department of Animal Nutrition and Management Reindeer Husbandry Unit P.O. Box 7024, 750 07 Uppsala Sweden

Organisation number: SE 202100-2817

Editor in Chief: Birgitta Åhman

Editor: Eva Wiklund

Address: Swedish University of Agricultural Sciences Department of Animal Nutrition and Management Reindeer Husbandry Unit P.O. Box 7024, 750 07 Uppsala Sweden

E-mail: [email protected]

Telephone: +46 18 67 23 08

Mobile phone: +46 70 251 53 07 (Editor in Chief); +46 70 683 22 77 (Editor)

As from Volume 28 (2008), Rangifer is published as on open access online journal (ISSN 1890-6729 at www.rangiferjournal.com). A complete archive of all published issues of Rangifer including Rangifer Special Issues and Rangifer Report is also available at www. rangiferjournal.com.

The publication of Rangifer is supported by funds from the Swedish Ministry of Enterprise and Innovation.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 2 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Special Issue No. 23

RANGIFER

Proceedings of the 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

RANGIFER

Proceedings of the 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

Caribou Conservation and Management: What’s Working? DOI 10.7557/2.35.2.3704

Editor in Chief of Rangifer: Birgitta Åhman

Technical Editor of Rangifer: Eva Wiklund

Issue editors: Troy Hegel Yukon Department of Environment, Canada Fiona Schmiegelow University of Alberta and Yukon College, Canada

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 4 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 RANGIFER

Special Issue 2015 No. 23

Contents Page

NACW at Thirty: A Work in Progress…………………………..……….....…………...... 6

Past North American Caribou Workshops…………….………...... ……....………………... 8

Organisation and sponsors .....……………………………….……………....….………….... 8

Articles

Susan K. Stevenson & Darwyn S. Coxson - Can partial‐cut harvesting be used to manage terrestrial lichen habitat? A review of recent evidence……..……....………....………..……....11

Emily K. Gonzales, Patrick Nantel, Arthur R. Rodgers, Martha L. Allen & Christine C. Drake - Decision-support model to explore the feasibility of using translocation to restore a woodland caribou population in Pukaskwa National Park, Canada………...... 27

Justina C. Ray, Deborah B. Cichowski, Martin-Hugues St-Laurent, Chris J. Johnson, Stephen D. Petersen, & Ian D. Thompson- Conservation status of caribou in the western mountains of Canada: Protections under the Species At Risk Act, 2002-2014...... 49

Kim G. Poole, Anne Gunn, Jack Wierzchowski & Morgan Anderson - Peary caribou distribution within the Bathurst Island Complex relative to the boundary proposed for Qausuittuq National Park, Nunavut…………...... …………...... 81

Christine B. Robichaud & Kyle H. Knopff- Biodiversity offsets and caribou conservation in Alberta: opportunities and challenes...... 99

Paula Bentham & Brian Coupal - Habitat restoration as a key conservation lever for wood‑ land caribou: A review of restoration programs and key learnings from Alberta………...... 123

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 6 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 The organising committee received a strong response to our call for contributions, resulting in a final selection of over 70 presentations and 50 posters representing a broad range of topics including co-management, caribou habitat, predator-prey dynamics, population dynamics, status assessment, environmental impact assessment, land use planning, and monitoring. Related exchanges were lively, and often spilled into the breaks, which thanks to phenomenal spring weather, were held on the banks of the Yukon River. While no one was complaining about the warmth, it was a poignant reminder of the impacts of climate change on conditions facing caribou in the coming decades. In addition to the main program, three pre-conference technical workshops were delivered on structured decision making, engaging local indigenous communities, and wildlife photography. As well, an evening event titled “A Celebration of Caribou” was hosted for the public and received very positive feedback. It included videos and stories and was an opportunity for people of all ages to learn more about caribou. These events highlight the increasing importance of social dimensions of caribou management on conservation outcomes. It is traditional for contributors to NACW to have an opportunity to publish their work in proceedings from the conference. At one time, this captured much of the caribou research taking place. However, caribou have moved into the mainstream, with related theoretical and applied research appearing in a broad array of journals, testimony to the breadth and quality of work being conducted, as well as its sheer volume. The contributions profiled here highlight the diversity of approaches being applied to contemporary caribou conservation, including active habitat management (Stevenson & Coxson, Bentham & Coupal), legislative tools for protection (Ray et al., Poole et al.), and application of new, and in some cases controversial, conceptual frameworks (Gonzales et al., Robichaud & Knopff). These speak to the opportunities and challenges that managers face in finding workable solutions to sustaining caribou now and into the future. On behalf of the 15th NACW organising committee we thank the sponsors, volunteers, and delegates for their shared commitments to the real Rangifer. We look forward to a continued tradition of inclusive and constructive dialogue and demonstration of innovative approaches to caribou conservation and management at future NACW meetings.

Troy Hegel, Yukon Department of Environment Fiona Schmiegelow, University of Alberta and Yukon College

Organisation

Organising Committee Troy Hegel (Conference Chair) – Yukon Department of Environment Natasha Ayoub – Carcross Tagish First Nation Brian Bell – Ta’an Kwäch’än Council Matt Clarke – Yukon Department of Environment Anne MacLeod – EDI Environmental Dynamics Inc. Don Reid – Wildlife Conservation Society Canada Kelsey Russell – EDI Environmental Dynamics Inc. Kyle Russell – Yukon Department of Environment Fiona Schmiegelow – University of Alberta and Yukon College Mike Setterington – EDI Environmental Dynamics Inc. Shannon Stotyn – Canadian Wildlife Service

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 8 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Event Planner Leaf Solutions

Presentation Management Alberto Suárez-Esteban – University of Alberta

Editorial Assistance Katherine Thiesenhausen

Conference Session Chairs Andrew deVries – Sustainable Forestry Initiative Mike Gill – Canadian Wildlife Service Troy Hegel – Yukon Department of Environment Chris Johnson – University of Northern British Columbia Thomas Jung – Yukon Department of Environment Bjørn Kaltenborn – Norwegian Institute for Nature Research Chris Ritchie – British Columbia Ministry of Forests, Lands, and Natural Resource Operations Fiona Schmiegelow – University of Alberta and Yukon College Mike Setterington – EDI Environmental Dynamics Inc. Ron Sumanik – Yukon Department of Energy, Mines, and Resources

Plenary Speakers Michael J. Conroy – University of Georgia Anne Gunn Stephen Virc – Environment Canada

GOLD SPONSORS ($10,000+)

SILVER SPONSORS ($5,000+)

BRONZE SPONSORS ($2,000+)

CONTRIBUTORS (< $2,000)

Silvacom Ltd. Telonics Inc Yukon Mines and Resources Yukon Brewing Yukon Department of Energy Advanced Telemetry Systems

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 10 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

Can partial‐cut harvesting be used to manage terrestrial lichen habitat? A review of recent evidence

Susan K. Stevenson1§ & Darwyn S. Coxson1 1 University of Northern British Columbia, Ecosystem Science and Management Program, 1 3333 University Way, Prince George, B.C. V2N 4Z9, Canada, (Corresponding author: [email protected]). § Deceased, August 29, 2014.

Abstract: Recent research suggests that partial-cut harvesting techniques can be used to alter successional trajectories in pine- and spruce-lichen woodlands, allowing forest managers to extend the period of reindeer lichen growth in mid- to late seral boreal forest stands. In Quebec, a fully replicated partial-cutting trial found that terrestrial lichen abundance remained at least as high in the partial cut as in the clearcuts or unlogged stands, and that the partial cut appeared to be on a trajectory to have even more terrestrial lichen due to sustained higher growth rates. In Alberta, a retrospective study found higher terrestrial lichen abundance in an early horse-logged partial cut than in undisturbed adjacent old forests or in clearcuts. Follow-up studies of partial-cut harvesting trials in British Columbia found that group selection plots 10 years after harvesting had lichen cover equivalent to that of undisturbed forest. In contrast, studies on lichen woodlands that have been defoliated by mountain pine beetle showed a major decline in reindeer lichen cover and a corresponding increase in vascular plant cover, similar to the results of previous studies on clear-cut logging impacts. Taken together these studies provide qualified support for the hypothesis that partial-cut harvesting can be used to enhance, or at least maintain, terrestrial lichen mats used as forage by caribou.

Key words: forest management; lichen woodlands; partial-cut harvesting; terrestrial lichens; woodland caribou.

Rangifer, 35, Special Issue No. 23, 2015: 11-26 DOI 10.7557/2.35.2.3461

Introduction are most relevant to our objectives, they are the The changes in stand structure that occur dur- most limited, and there are also valuable lessons ing succession in boreal and sub-boreal lichen to be learned from other successional series. woodlands are accompanied by major shifts in the composition of forest floor lichen commu- Post-fire succession in the boreal forest nities. We review the impacts of three major Ahti (1959) described 5 stages in the develop- types of disturbances: fire, partial-cut logging, ment of boreal forest terrestrial lichen com- and canopy mortality due to mountain pine munities after fire. 1) The first stage was a bare beetle attack. Although data from partial cuts soil or organic substrate stage in the immediate

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 12 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 & Marsh, 2001). As the understory becomes cooler and moister feather-moss mat fronds can infiltrate and eventually bury existing reindeer lichen mats (Fig.1). Coxson & Marsh (2001) documented the shift from dominance by Cladina morphotype lichens in mature stands to dominance by pleurocarpous mosses in older stands (Fig. 2) within sub-boreal forests in central-interior BC. The biomass of Cladina morphotype lichens, including C. mitis, C. rangiferina, C. stellaris, and C. uncialis, in 50 - 100 year old stands exceeded 1700 kg/ha, falling to less than 300 kg/ha after replacement by feather-moss mats had occurred in the stands > 100 years in age (Coxson & Marsh, 2001). Several regional exceptions to this pattern of canopy closure and domination by feather- moss mats in old-growth lichen woodlands occur. In the clay belt of northwestern Quebec Figure 2. Mean cover (%) by functional group within paludification in old stands can lead to reduced 50-year age-class intervals (n= 4, 4, 5, and 2 stands re- vigour and lower tree densities, with the forest spectively) for pine-lichen woodlands in central-interior floor surface gradually becoming dominated by British Columbia. TheCladina morphotype functional Sphagnum et al., group consisted of Cladonia arbuscula, C. rangiferina, mats (Boudreault 2002; Harp- and C. stellaris (adapted from Coxson and Marsh, 2001). er et al., 2003). In cooler oceanic climates, such as in boreal forests in the Grande rivière de la tive to stand nutrient availability. Lichens can Baleine area of northeastern Quebec, no transi- access stand level nutrient reserves indirectly, tion to feather-moss mats occurred even in very from exposure to through-flow precipitation in old stands, with C. stellaris mats dominating drip-zones under shrubs and trees (Haughian the forest floor in stands at least 250 years old & Burton, 2015), and for some species directly, (Morneau & Payette, 1989). using rhizines to mobilize soil or bark nutrient Successional changes in forest floor lichen reserves (Cornelissen et al., 2007). communities reflect both stochastic factors, such as the availability of propagule sources Abundance and viability of terrestrial lichens af- over time (Hilmo & Såstad, 2001), and the re- ter partial cutting sponse of individual lichen species to gradients Partial-cutting provides a means of bringing of temperature, moisture, and light availabil- about changes in stand structure through re- ity (Tegler & Kershaw, 1980; Kershaw, 1985). moval of selected trees. If the objective of for- Generally, early successional lichen species are est managers is to prolong the period of rein- thought to be more tolerant of heat extremes deer lichen growth in forest stands, the impact associated with post-fire surface microclimate, of changes in stand structure from partial-cut while late-successional lichen species are more harvesting should be greatest in late succes- sensitive to extremes of desiccation and heat sional stands, especially if canopy closure has exposure (Kershaw, 1977; Kershaw, 1985). already started to occur. Selective removal of Terrestrial lichen communities are also sensi- trees in late-seral stands should alter forest floor

Table 1. Mean microplot values (± 1S.E.) for leaf area index in moss- versus lichen-dominated plots in pine lichen woodlands from central-interior British Columbia. Significance values are shown for t-test comparisons between ‘Moss’ and ‘Lichen’ plots at each location (from Sulyma & Coxson 2001).

Plot variable Site Moss plots Lichen plots Significance values Leaf Area Index (m2·m2) West Ger- 1.617 6 1.551 P < 0.0001 mansen (0.016) (0.014) Germansen 1.700 6 1.637 P = 0.0004 Lake (0.024) (0.0126) 1.961 6 1.787 P = 0.0003 Manson (0.030) (0.020)

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 14 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 tial removal of canopy structure comes from morphotype species (C. stellaris, C. mitis, and several studies. Snyder and Woodard (1992) C. rangiferina) in three treatment types: partial did a retrospective study of lichen abundance cuts, clearcuts, and controls. Both unlogged in 18 different-aged clearcuts, one 20-year-old stands and partial cuts had higher percent cover horse-logged partial cut, and two unlogged of Cladina spp. than clearcuts. Growth rates of stands within the Subalpine Ecoregion of Al- the Cladina morphotype species were higher berta. Terrestrial lichen cover, in general, was in both the partial-cut and clearcut sites than greater in the partial cut than in the unlogged in the control sites (Fig. 3). In this forest type, stands. Lichen cover in the unlogged stands it appeared that partial cuts offered the best was similar to that in the 20- and 30-year-old combination of retention of preharvest forage clearcuts, but greater than that in the 10-year- lichens, and environmental conditions pro- old clearcuts. Percent cover of Cladonia spe- moting lichen growth after harvest. Interest- cies palatable to caribou, however, was greater ingly, measurements for C. stellaris and C. mitis in the partial cut than in the unlogged stands, showed negative growth rates (loss of biomass) and greater in the unlogged stands than in the in the control stands, perhaps indicative of an clearcuts (Snyder & Woodard, 1992), suggest- already declining status for Cladina morpho- ing that in this case, partial cutting not only type species in these late seral stands. As Moser retained preferred forage lichens but actually et al. (1978) noted, however, negative growth enhanced them compared to clearcutting. rates may be a periodic feature of even healthy In a replicated study in Québec black spruce lichen mats, during episodes of unfavourable forests, Boudreault et al. (2013) compared the climatic conditions. abundance and growth rates of three Cladina One of the most comprehensive studies to

Figure 3. Mean growth rates (+ 1 S.E.) of Cladonia stellaris, C. rangiferina, and C. mitis in black spruce boreal forests of western Québec from spring 2005 to fall 2006. Differences in growth rate that were significantP ( < 0.05) according to least squares means Tukey’s honestly significant difference tests are indicated by different letters (excerpted from Boudreault et al., 2013).

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 15 date on the impact of partial-cutting on ter- Although not part of the experimental design, restrial lichen mats was conducted by British three adjacent clearcuts were also monitored, Columbia’s Ministry of Forests and Range, who beginning in 2001. began a replicated silvicultural systems study in Cover of forage lichens was significantly 1995, described by Miège et al. (2001), with lower in the three harvested treatments than in follow-up studies described by Waterhouse et the unharvested control at the first reassessment al. (2011). The study was carried out in the after harvesting, and subsequently increased. western portion of the range of the Itcha-Ilga- Declines in the harvested units were greater in chuz caribou herd in British Columbia’s central the openings than in the residual forest stand. interior plateau. Compared to other forested In the IGS-SO treatment, healthy forage li- BC ecosystems, the climate of the Dry, Cold chens had declined to 49% of preharvest levels sub-Boreal Pine Spruce biogeoclimatic zone in 1998, and increased to 68% of preharvest (SBPSxc) and Very Dry Very Cold Montane levels by 2004. In the IGS-WT treatment, Spruce biogeoclimatic zone are very dry and healthy forage lichens had declined to 57% cold (Meidinger & Pojar, 1991), with light lev- of preharvest levels in 1998, and increased to els within stands sufficient to allow pine regen- 71% of preharvest levels by 2004. In the GS- eration in the understorey (Waterhouse et al., SO treatment, healthy forage lichens had de- 2010). clined to 53% of preharvest levels in 1998, but Waterhouse et al. (2011) described results had nearly reached preharvest levels again by from their measurements taken in 2004 (8.5 2004 (Fig. 4). Total moss cover also showed a years after logging), reporting on the following treatments: • Irregular group shelterwood with stem-only harvesting (IGS-SO) • IGS with whole tree harvesting (IGS-WT) • Group selection with stem-only harvesting (GS-SO) • No harvest The irregular shelterwood prescription called for 50% removal in openings ranging from 20 - 30 m in diameter, and was designed to provide partial shade for terrestrial lichens in the openings. The group selection prescription called for 33% removal in openings about 15 m in diameter, and was designed to maintain arboreal lichens. In stem-only treatments, debris from topping and delimbing was aggregated and left in the harvested openings. In the whole tree treatment, debris from topping and delimb- Figure 4. Percent cover of mosses (all moss species) and ing was piled and burned at the roadside. Post- lichens (preferred forage species) in control versus irreg- harvest surveys showed that the actual area cut ular group shelterwood (stem-only: GS-SO) harvesting was 39% in the irregular group shelterwood plots under pre-harvest (1995) and post-harvest (1998, and 28% in the group selection, and that the 2000, 2004) conditions on caribou winter range in west-central British Columbia. Adapted from Water- opening sizes were within the targeted range. house et al., 2011.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 16 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 similar pattern of decline and then recovery in moderate to high in all four areas, ranging from the GS-SO treatment, though values remained 46 - 75% in the Itcha-Ilgachuz to 78 - 96% well below that of the control stands in 2004. in the Tweedsmuir-Entiako (Cichowski, 2011). This recovery to levels below that of control Cover of terrestrial forage lichens in un- stands is consistent with a negative response of logged stands decreased after the mountain moss mats to reduced canopy cover, reflecting, pine beetle attack in all four study areas. In the among other factors, the loss of stored carbon Tweedsmuir-Entiako, mean lichen cover was reserves in feather moss-mats when rewetting 16% in 2001, 13% in 2003, 11% in 2005, and follows intense desiccation events (Wilson & 10% in 2007. The rate of lichen decline had Coxson, 1999). diminished or stopped in most site series by Dwarf shrub cover, initially at 8 - 10%, de- 2007 – about 7 or 8 years after the initial attack clined slightly in the partial cuts immediately (Cichowski et al., 2008). In the northeastern after harvest, then increased to slightly (2 - 6%) portion of the Itcha-Ilgachuz range, mean li- above preharvest levels. Herb cover was only 2 - chen cover declined from 20.5% on plots es- 3% in the preharvest stand, and increased to 3 - tablished in 2005 and 2006 to 16.4% in 2008 4% in the partial cuts by 2004. Cover of dwarf (Cichowski et al., 2009). In the western portion shrubs and herbs had a significant negative re- of the Itcha-Ilgachuz, terrestrial lichen cover in lationship with cover of healthy forage lichens the no-harvest controls for the silvicultural sys- at all measurement times. Seven to eight years tems study of Waterhouse et al. (2011) declined post-harvest in the adjacent clearcuts, forage li- from the 2004 pre-mountain pine beetle cover chen cover was low (2.8%), dwarf shrub cover of 11.3% to 9.7% in 2008, but this change was was similar (16.9%), and herb cover was high not statistically significant (Waterhouse, 2011). (17.4%), compared to the partial cuts. At Kennedy Siding, terrestrial lichen cover was 24% in 2006, 17% in 2007, and 12% in 2008 Lessons from mountain pine beetle attacks and 2009 (Cichowski, 2011). The mountain pine beetle epidemic has killed Concurrently, cover of dwarf shrubs in- at least 726 million m3 of timber in BC’s increased in all four study areas. On most sites in terior forests, affecting more than 17.5 mil- the Tweedsmuir-Entiako range, kinnikinnick lion ha and killing most of the mature lodge- (Arctostaphylos uva-ursi) cover increased from pole pine in the Central Plateau region of the about 30% to about 60% between 2001 and province (British Columbia Ministry of For- 2007 (Cichowski et al., 2008). In some, but ests, Lands and Natural Resource Operations, not all, subzones, the rate of increase levelled 2012). Cichowski (2011) reviewed literature off after 2005, and dieback of kinnikinnick was on the potential effects of the epidemic on observed in some plots. Informal observations caribou. In addition to the western portion of in 2010 suggested that the kinnikinnick die- the Itcha-Ilgachuz winter range (Waterhouse, back was continuing and was more widespread 2011), data on the effects of mountain pine (Cichowski, 2011). In the northeastern por- beetle mortality on terrestrial lichens are avail- tion of the Itcha-Ilgachuz, kinnikinnick cover able from three other study areas: the Tweeds- increased significantly from 3.6% in 2005-06 muir-Entiako winter range (Cichowski et al., to 6.2% in 2008, and twinflower Linnaea( bo- 2008), the Kennedy Siding winter range (Seip realis) cover increased significantly from 3.4% & Jones, 2009), and the northeastern portion to 8.5% (Cichowski et al., 2009). In the west- of the Itcha-Ilgachuz winter range (Cichowski ern portion of the Itcha-Ilgachuz, mean percent et al., 2009). The level of pine mortality was cover of dwarf shrubs increased from 10.1% in

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 17 2004 to 17.3% in 2008 (Waterhouse, 2011); woodlands consistently show a positive rela- species that increased included kinnikinnick, tionship between terrestrial forage lichen abun- twinflower, crowberry Empetrum( nigrum) and dance and canopy openness, mountain pine grouseberry (Vaccinium scoparium) (Cichowski, beetle mortality, which increases light trans- 2011). At Kennedy Siding, velvetleaf huck- mission, has been associated with declines in leberry (Vaccinium myrtilloides) exhibited a terrestrial lichen cover. In the northeastern por- 17.5% increase in cover from 2006 to 2008 tion of the Itcha-Ilgachuz range, Cichowski et (Seip & Jones, 2009). Moss cover generally de- al. (2009) found that the greater the increase in clined (northeastern Itcha-Ilgachuz, Kennedy light transmission following the mountain pine Siding) or remained the same (Tweedsmuir- beetle epidemic, the greater the decrease in ter- Entiako, western Itcha-Ilgachuz) in mountain restrial lichen abundance. Between 2001 and pine beetle-attacked stands (Cichowski, 2011). 2007 in the Tweedsmuir-Entiakio Range, more The death of some, but not all, of the trees mountain pine beetle mortality was associated in the forest canopy, either through mountain with a greater decrease in lichen cover, but pine beetle attack or through partial cutting, once the effect of mountain pine beetle mortal- may affect terrestrial lichen abundance in sev- ity was taken into account, higher light levels eral ways. The amount of light reaching the for- had a positive effect. A separate examination of est floor is increased. There is increased deposi- trends between 2005 and 2007 indicated that tion of needles and woody debris on the forest changes in light transmission or interception floor, although the timing of deposition differs had a significant effect on changes in lichen between insect mortality and logging, and with abundance, suggesting that by seven to eight logging method. Water relationships may also years after the initial attack, increased light was change. Surface drying increases with increased beginning to create favourable conditions for irradiance, but soil moisture may increase as lichen recovery (Cichowski et al., 2009). there are fewer canopy trees to transpire mois- Armleder and Waterhouse (2008) and Wa- ture into the atmosphere. Significant changes terhouse et al. (2011) have asserted that lichen in surface temperature and humidity profiles morphotypes that have been growing in sub- can occur with loss of canopy cover. Greater in- dued light are adversely affected by a sudden solation exposure in sites with dry continental increase in exposure to light. Lichens of the climates can lead to cooler night-time tempera- same species growing under different light con- tures and more intense early morning dewfall, ditions exhibit both morphological and physi- triggering wetting events in those lichen spe- ological differences. Lichens have the capabil- cies capable of rehydration from dewfall and/ ity of responding to changes in the light regime or high humidity events alone, similar to those by increasing the production of phenolic com- observed by Bidussi et al. (2013) for epiphytic pounds believed to have a protective function; lichens in interior BC. However, these same in Cladina morphotypes newly exposed to en- sites with reduced canopy cover will typically hanced levels of ultraviolet light, accumulation experience lower mid-day humidity and thus of phenolics increased and the penetration of more intense mid-day drying of forest floor ultraviolet radiation into the lichen thallus was moss and lichen mats. All these changes may more strongly attenuated than in lichens not affect lichens directly, or may affect them in- exposed to enhanced levels of radiation (Buf- directly by altering relationships between the foni Hall et al., 2002). Gauslaa et al. (2012) lichens and their competitors. noted that a rapid transition to high light ex- Although studies in undisturbed lichen posure in desiccated lichens (this in work on

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 18 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Lobaria pulmonaria) resulted in chlorophyll (skirts) in the high needle litter deposition zone degradation, an interaction that was exacer- beneath live lodgepole pine trees. Cichowski et bated under low relative humidity conditions. al. (2008) observed that the rapid expansion of This same high light exposure while the lichens kinnikinnick in the Tweedsmuir-Entiako study were desiccated gradually, however, triggered area following mountain pine beetle attack cor- changes in the photochemistry of the lichens responded to the massive needle deposition by that improved their tolerance of excess irradi- canopy trees and to the increased availability of ance in the desiccated state (Stepigova et al., nutrients, light and water. They suggested that 2008). the needle-pulse combined with the increase in It is therefore possible that the increase in available nutrients allowed kinnikinnick to ex- light levels brought about by beetle mortal- pand beyond its ‘skirt’ at the base of trees, into ity of canopy trees reported by Cichowski et areas normally too poor in nutrients to support al. (2009) induced mortality in some lichens, kinnikinnick. Needle deposition throughout and that those that survived had acclimatized their study area has slowed down, and the ex- to increased light levels seven to eight years af- tra nutrient pulse that may have allowed kinni- ter the initial beetle attack. Differences in in- kinnick to establish in marginal habitats is no creased light exposure could also explain why longer available, resulting in a slowing of kinni- forage lichens in 15-m wide openings recov- kinnick growth or even die-back. ered faster than those in 30-m wide openings Over time, woody debris levels will in- between 2000 and 2004 (Waterhouse et al., crease after mountain pine beetle outbreaks as 2011), but were then adversely affected by a branches gradually break off the dead trees and further increase in light exposure from moun- the trees begin to fall, but the mountain pine tain pine beetle attack between 2004 and 2008, beetle-affected stands reviewed by Cichowski while those in the 30-m wide openings were (2011) had not yet reached this stage. After not (Waterhouse, 2011). Gauslaa et al. (2006) stem-only partial cutting, there is a pulse of lit- noted that lichen response to changing canopy ter that is not confined to pre-existing needle- conditions is a delicate balance between growth fall zones, and is composed of both logging potential (more with increasing light) and des- slash and needlefall. It is therefore more likely iccation damage, which is greater under high to cover lichen mats than the needles that fall light conditions. from beetle-killed trees, and was identified by Miège et al. (2001) as a factor adversely affect- Interactions with understory shrubs and woody ing post-harvest lichen abundance. By 2008, debris however, the difference between the group Mortality of canopy trees from mountain pine selection and the two irregular group shelter- beetle attacks is followed by a period of two or wood treatments, which differed in level of three years during which the needles fall from removal and opening size, was far greater than the dead trees, covering ground vegetation and the difference between stem-only harvesting producing a pulse of nutrients. Although li- and whole-tree harvesting (Waterhouse et al., chens do not generally dominate areas imme- 2011), and the authors observed that woody diately under the crown-radius projection of debris left over from stem-only harvesting may canopy trees (Haughian, 2010), they may be have ameliorated microclimate conditions for affected by vegetation changes within those the lichens. areas. Kinnikinnick, a major competitor with forage lichens, often forms circular colonies

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 19 Lichen management using partial cuts moderated, and appeared to provide suitable Current literature supports the concept that microenvironments for lichens, whereas lichens partial cutting can be used to maintain, and present in road-side delimbed sites without in some cases perhaps enhance, the preharvest cover often appeared to be suffering damage lichen community. This maintenance strategy from exposure. of partial-cut harvesting has gained acceptance Waterhouse et al. (2011) observed that slash as a viable management approach in BC, espe- on the ground and suspended low above the cially in the relatively dry and cold Montane ground adversely affected lichens, but high sus- Spruce Zone (Meidinger & Pojar, 1991). Par- pended slash and areas adjacent to slash piles tial cutting with the objective of maintaining li- may provide refugia for lichens. An important chen forage is a key part of the management ap- point raised by Waterhouse et al. (2011) was proach on more than 181,000 ha in the range that whole-tree (WT) and stem-only (SO) of the Itcha-Ilgachuz caribou herd (Armleder harvesting systems have the potential of affect- & Waterhouse, 2008), and was recommended ing lichens differently. Whole tree skidding is by McNay (2011) for high-elevation plateaus likely to cause more damage to the lichen mat in the Montane Spruce Zone in B.C. Specif- than skidding of delimbed stems, and results in ically, McNay (2011) recommended an irreg- roadside processing areas that are severely dis- ular group shelterwood system with openings turbed. The slash generated from on-site pro- not exceeding two tree lengths wide by three or cessing covers lichens. Percent cover of logging four tree lengths long for the purpose of main- slash had a significant negative relationship taining a sustainable supply of terrestrial forage with abundance of healthy forage lichens at all lichens. He added the following caveat: “Prac- measurement times, but forage lichen abun- titioners are advised that while this specific sil- dance in the two irregular group shelterwood vicultural regime has been shown to maintain (IGS) treatments did not differ significantly terrestrial forage lichens, further monitoring from one another at any time. The authors not- may be necessary to prove the regime does not ed that slash deposited on the ground crushed subject caribou to greater spatial overlap with lichen, while low suspended slash prevented an early-seral predator-prey system” (McNay, light and precipitation from reaching lichen 2011: p. 68). mats. The influence of slash on lichen regrowth Particular care should be taken with respect may be broadly similar in other disturbance to the adverse impacts of logging slash (needles features, such as seismic lines, which at a mi- and woody debris) on terrestrial forage lichens, crosite level can lead to increased canopy open- as described by Gough (2010). The exclusion ness. However, as discussed below, linear dis- of woody debris that occurs under a branch turbance features have other attributes which and stem harvesting system probably has mixed can significantly reduce their value as habitat impacts on terrestrial forage lichens. In a com- for caribou. parison of several different clearcut harvesting Harvesting with a high level of retention and methods in Alberta, Kranrod (1996) found small patch size, as recommended by Gough that stump-side delimbing in combination (2010) is a key component of any successful with winter harvest and without scarification partial-cutting lichen retention strategy. Terres- left more terrestrial lichen cover the summer trial forage lichens in the 30% removal treat- after timber harvest than any other treatment ment of Waterhouse et al. (2011) recovered to combination. He observed that environmental preharvest levels, whereas lichens in the 50% conditions under piles and at pile edges were removal treatment with 30-m openings did

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 20 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 not (Waterhouse et al., 2011). Sulyma (2002) of ground cover and limiting future use. The provided a summary of the predicted interac- design of access roads for hauling timber from tions between harvesting methods, harvest- partial-cutting harvest blocks raises further ising season, site preparation, and regeneration sues of predator access and changes in land- method (Table 2). Sulyma recommended that scape patterns. Previous studies suggest that harvesting occur during the winter season to the impacts of linear features such as roads and minimize disturbance of lichen mats during seismic lines may be greatest in the late-winter harvesting with whole tree removal preferred period, when woodland caribou were found by to minimize the amount of residual debris left Dyer et al. (2002) to cross active roads 6 times after harvesting. less frequently than simulated road networks. Dyer et al. (2001) also found significant avoid- The broader context of caribou habitat manage- ance effects in woodland caribou, up to 500 m, ment from roads and seismic lines, although these Caribou populations are affected by many fac- effects should decline with time as vegetation tors other than the abundance of terrestrial recolonizes deactivated roads and seismic lines. forage lichens, and forest harvesting can affect A strategy for road deactivation, road access re- caribou in ways other than through its impact strictions, and restoration of road corridors, if on lichens. Here we discuss mainly the impacts required, is therefore a vital part of any partial- of forest harvesting on terrestrial forage lichens, cutting harvest design in caribou winter habi- but a landscape-level management plan should tat, in common with other linear disturbance take into consideration the entire array of hu- features in caribou habitat. man-mediated influences on caribou populations. Conclusion A cornerstone of present-day caribou man- The use of partial-cut logging to reduce can- agement in North America has been to con- opy closure in mid- to late-seral lichen wood- sider interactions between habitat modifica- lands provides a short- to mid-term strategy for tions and predator response. Current findings maintaining or even enhancing forage lichen suggest that factors which increase predator availability for caribou. This strategy would populations, for instance, creation of early seral best be considered in landscapes where there habitats which favor growth of moose or deer are few mid-seral stands that can replenish car- populations, or those which allow more ready ibou habitat in coming decades. Under these predator dispersal, such as the creation of road circumstances, partial-cut harvesting may be networks, can have major detrimental impacts used as a tool to maintain lichen forage avail- on caribou (Apps et al., 2013; Whittington et ability in locations where it might be lost due to al., 2011). Care must also be taken that partial- successional change. Eventually, however, these cut harvesting intensity does not exceed the stands will require resetting by stand level dis- threshold at which a flush of early seral vegeta- turbance factors such as wildfire to develop fu- tion might occur (Frey et al., 2003). ture lichen communities. Partial-cutting, there- The design of partial-cut logging blocks fore, cannot be used as a long-term substitute must also include considerations of changes in for natural disturbance dynamics that maintain road access which would be required and fu- a broad age-class distribution of stands after ture use of these roads. At a small scale, within disturbances such as fire. blocks, planned skid trails would optimally be winter access only, minimizing disturbance

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 21 Acknowledgments Beetle facts: Infestation information. Available We would like to thank the Government of the from: http://www.for.gov.bc.ca/hfp/moun- Yukon, Forest Management Branch, for fund- tain_pine_beetle/Updated-Beetle-Facts_ ing support for an earlier draft of this review. May2012.pdf (Accessed 10 April 2014). Buffoni Hall, R.S., Bornman, J.F. & Björn, L.O. 2002. UV-induced changes in pigment References content and light penetration in the fruticose Ahti, T. 1959. Studies on the caribou lichen lichen Cladonia arbuscula ssp. mitis. — Jour- stands of Newfoundland. — Ann. Bot. Soc. nal of Photochemistry and Photobiology B: Bi- 'Vanamo' 30: 1-43. ology 66: 13-20. http://dx.doi.org/10.1016/ Apps, C.D., McLellan, B.N., Kinley, T.A., S1011-1344(01)00270-6 Serrouya, R., Seip, D. & Wittmer, H. Cichowski, D. 2011. Literature Review — 2013. Spatial factors related to mortality and Research on Mountain Pine Beetles and population decline of endangered mountain Caribou. Report prepared for B.C. Min- caribou. — Journal of Wildlife Management istry of Environment, Williams Lake by 77: 1409-1419. http://dx.doi.org/10.1002/ Caribou Ecological Consulting, Smith- jwmg.601 ers, B.C. Available from: http://a100.gov. Armleder, H. & Waterhouse, M. 2008. Moun- bc.ca/appsdata/acat/documents/r22847/ tain pine beetle and northern caribou: The Cichowski_1303238545657_e822d6b- Itcha- Ilgachuz experience. — BC Journal of 59c6854aeb61e99363115a4aa9016b33c- Ecosystems and Management 9: 102-105. 98cf2984cbb2bfb8c0cea71c.pdf (Accessed Bidussi, M., Goward, T. & Gauslaaa, Y. 10 April 2014). 2013. Growth and secondary compound Cichowski, D., Macadam, A. & Haeussler, investments in the epiphytic lichens Lobar- S. 2009. Mountain pine beetle/lichen pro- ia pulmonaria and Hypogymnia occidentalis ject, Quesnel TSA: Year 4—2008/09. Report transplanted along an altitudinal gradient in prepared for B.C. Ministry of Environ- British Columbia. — Botany 91: 621-630. ment, Williams Lake, B.C. FIA-FSP Pro- http://dx.doi.org/10.1139/cjb-2013-0088 ject Y091176. Available from: http://www. Boudreault, C., Bergeron, Y., Gauthier, Y. for.gov.bc.ca/hfd/library/FIA/2009/FSP_ & Drapeau, P. 2002. Bryophyte and lichen Y091176c.pdf (Accessed 10 April 2014). communities in mature to old-growth stands Cichowski, D, Williston, P. & Haeussler, S. in eastern boreal forests of Canada. — Ca- 2008. The response of caribou terrestrial for- nadian Journal of Forest Research 32: 1080- age lichens to mountain pine beetles and forest 1093. http://dx.doi.org/10.1139/x02-027 harvesting in the East Ootsa and Entiako area: Boudreault, C., Zouaoui, S., Drapeau, P., Annual Report – 2007/08 – Year 7. A report Bergeron, Y. & Stevenson, S. 2013. Can- to Morice-Lakes Innovative Forest Prac- opy openings created by partial cutting in- tices Agreement, Prince George, B.C., the crease growth rates and maintain the cover of Bulkley Valley Centre for Natural Resourc- three Cladonia species in the Canadian bore- es Research and Management, Smithers, al forest. — Forest Ecology and Management B.C., and Ministry of Environment, Prince 304: 473-481. http://dx.doi.org/10.1016/j. George, B.C. 46p. Available from: http:// foreco.2013.05.043 www.sernbc.ca/pdf/Caribou/Year%207%20 British Columbia Ministry of Forests, Lands Report%20on%20the%20Response%20 and Natural Resource Operations. 2012. of%20Caribou%20Terrestrial%20For-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 22 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 age%20Lichens%20to%20MPB.pdf (Ac- pulmonaria in forests with contrasting light cessed 10 April 2014). climates. — Oecologia 147: 406-416. http:// Cornelissen, J.H.C., Lang, S.I., Soudzilovs- dx.doi.org/10.1007/s00442-005-0283-1 kaia, N.A. & During, H.J. 2007. Com- Gough, A. 2010. Effects of historical timber parative Cryptogam Ecology: A Review harvesting practices on caribou forage li- of Bryophyte and Lichen Traits that Drive chen abundance near Marsh Lake, Yukon: Biogeochemistry. – Annals of Botany 99: Phase 1 final report. – Unpublished report 987-1001. http://dx.doi.org/10.1093/aob/ prepared for Forest Management Branch, mcm030 Government of Yukon. 56 pp. Available Coxson, D. S. & Marsh, J. 2001. Lichen from: http://www.emr.gov.yk.ca/forestry/ chronosequences (post-fire and post-harvest) pdf/Lewes_Lake_Caribou_Lichen_Report. in lodgepole pine (Pinus contorta) forests of pdf (Accessed 10 April 2014). northern-interior British Columbia. — Ca- Harper, K., Boudreault, C., DeGrandpré, L., nadian Journal of Botany 79: 1449-1464. Drapeau, P., Gauthier, S. & Bergeron, Y. http://dx.doi.org/10.1139/b01-127 2003. Structure, composition, and diversity Dyer, S.J., O’Neill, J.P., Wasel, S.M. & Bou- of old-growth black spruce boreal forest of tin, S. 2001. Avoidance of industrial devel- the Clay Belt region in Quebec and Ontar- opment by woodland caribou. — Journal of io. — Environmental Reviews 11: S79-S98. Wildlife Management 65: 531–542. http:// http://dx.doi.org/10.1139/a03-013 dx.doi.org/10.2307/3803106 Haughian, S.R. 2010. Understory vegetation- Dyer, S.J., O’Neill, J.P., Wasel, S.M. & Bou- environment relationships of lichen rich forests tin, S. 2002. Quantifying barrier effects of in north central British Columbia. — M.Sc. roads and seismic lines on movements of thesis. University of Northern British Co- female woodland caribou in northeastern lumbia, Prince George, B.C. Alberta. — Canadian Journal of Zoology 80: Haughian, S.R. & Burton, P. 2015. Micro- 839-845. http://dx.doi.org/10.1139/z02- habitat associations of lichens, feathermoss- 060 es, and vascular plants in a caribou winter Frey, B.R., Lieffers, V.J., Munson, A.D. & range, and their implications for understo- Blenis, P.V. 2003. The influence of partial ry development. — Botany. http://dx.doi. harvesting and forest floor disturbance on org/10.1139/cjb-2014-0238 nutrient availability and understory veg- Hilmo, O. & Såstad, S.M. 2001. Coloniza- etation in boreal mixedwoods. — Canadian tion of old-forest lichens in a young and an Journal of Forest Research 33: 1180-1188. old boreal Picea abies forest: an experimental http://dx.doi.org/10.1139/x03-042 approach. — Biological Conservation 102: Gauslaa, Y., Coxson, D.S. & Solhaug, K.A. 251-259. http://dx.doi.org/10.1016/S0006- 2012. The paradox of higher light tolerance 3207(01)00100-8 during desiccation in rare old forest cyanoli- Jonsson Čabrajić, A.V., Moen, J. & Palmqvist, chens than in more widespread co-occur- K. 2010. Predicting growth of mat-forming ring chloro- and cephalolichens. — New lichens on a landscape scale – compar- Phytologist 195: 812–822. http://dx.doi. ing models with different complexities. org/10.1111/j.1469-8137.2012.04221.x — Ecography 33: 949-960. http://dx.doi. Gauslaa, Y, Lie, M., Solhaug, K.A. & Ohl- org/10.1111/j.1600-0587.2009.06079.x son, M. 2006. Growth and ecophysiologi- Kershaw, K.A. 1977. Studies on lichen-dom- cal acclimation of the foliose lichen Lobaria inated systems. XX. An examination of

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 23 some aspects of the northern boreal lichen chen–spruce woodland recovery at the limit woodlands in Canada. — Canadian Jour- of the boreal forest in northern Quebec. — nal of Botany 55: 393-410. http://dx.doi. Canadian Journal of Botany 67: 2770-2782. org/10.1139/b77-050 http://dx.doi.org/10.1139/b89-357 Kershaw, K. A. 1985. Physiological Ecology of Moser, T.J. & Nash III, T.H. 1978. Photosyn- Lichens. Cambridge Univ. Press. Cambridge, thetic patterns of Cetraria cucullata (Bell.) U.K. 293pp. Ach. at Anaktuvuk Pass, Alaska. — Oeco- Kershaw, M., Wiltshire, R. & Hollsted, C. logia 34: 37-43. http://dx.doi.org/10.1007/ 1994. Critical silvics of feathermoss as related BF00346239 to vegetation management. — Technical Note Rouse, W.R. 1976. Microclimatic Chang- TN-29, Ontario Ministry of Natural Re- es Accompanying Burning in Subarc- sources. 16pp. Available from: https://www. tic Lichen Woodland. — Arctic and Al- mnr.gov.on.ca/stdprodconsume/groups/ pine Research 8: 357-376. http://dx.doi. lr/@mnr/@nwsi/documents/document/ org/10.2307/1550439 mnr_e005383.pdf (downloaded 10 April Seip, D. & Jones, E. 2009. Response of wood- 2014). land caribou to partial retention logging of Kranrod, K.A. 1996. Effects of timber harvest- winter ranges attacked by mountain pine bee- ing methods on terrestrial lichens and under- tle. — Report on FIA/FSP Project Y091010 story plants in west-central Alberta. – M.Sc. (2008-09). Available from: http://www. thesis. University of Alberta, Edmonton. for.gov.bc.ca/hfd/library/FIA/2009/FSP_ 138pp. Y091010a.pdf (Accessed 10 April 2014). Maikawa, E. & Kershaw, K. A. 1976. Stud- Skatter, H.G., Kansas, J.L., Charlebois, M.L. ies on lichen-dominated systems. XIX. The & Balicki, B. 2014. Recovery of terres- postfire recovery sequence of black spruce – trial lichens following wildfire in the boreal lichen woodland in the Abitau Lake Region, shield of Saskatchewan: Early seral foraging N.W.T. — Canadian Journal of Botany 54: availability for Woodland Caribou (Rangifer 2679-2687. http://dx.doi.org/10.1139/b76- tarandus caribou) — Canadian Wildlife Biol- 288 ogy and Management 3: 1-14. McNay, R.S. 2011. Silviculture options for use Snyder, J. & Woodard, P.M. 1992. Lichen rein ranges designated for the conservation of generation rates in Alberta following various northern caribou in British Columbia. — types of logging and wildfire disturbances: final BC Journal of Ecosystems and Management report. — Unpublished report, University of 12: 55-73. Alberta, Edmonton. 118pp. Meidinger, D. & Pojar, J. 1991. Ecosystems of Stepigova, J., Gauslaa, Y., Cempirkova- British Columbia. BC Ministry of Forests, Vrablikova, H. & Solgaug, K.A. 2008. Victoria, British Columbia. Irradiance prior to and during desiccation Miège, D., Goward, T., Waterhouse, M. & improves the tolerance to excess irradiance Armleder, H. 2001. Impact of partial cut- in the desiccated state of the old forest li- ting on lichen diversity in lodgepole pine chen Lobaria pulmonaria. — Photosynthetica forests on the Chilcotin Plateau in British 46: 286-290. http://dx.doi.org/10.1007/ Columbia. — BC Ministry of Forests, Re- s11099-008-0050-8 search Branch, Victoria, B.C. Working Pa- Sulyma, R. 2002. Adaptive management of per 55. forestry practices in pine-lichen woodlands in Morneau, C. & Payette, S. 1989. Postfire li- north-central British Columbia, Site 1 (98-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 24 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Mile, Mackenzie) and Site 2 (Malaput – Van- time-to-event approach. — Journal of Ap- derhoof) Pretreatment Assessments. — Un- plied Ecology 48: 1535–1542. http://dx.doi. published Report. Forest Floor Contracting, org/10.1111/j.1365-2664.2011.02043.x Fort St. James, B.C. 32pp. Available from: Wilson, J.A. & Coxson, D.S. 1999. Carbon https://www.for.gov.bc.ca/hfd/library/ flux in a subalpine spruce-fir forest: pulse re- fia/2006/LBIP_2418011d.pdf (Accessed 10 lease from Hylocomium splendens feather- April 2014). moss mats. — Canadian Journal of Botany Sulyma, R. & Coxson, D.S. 2001. Micro- 77: 564-569. http://dx.doi.org/10.1139/ site Displacement of Terrestrial Lichens b99-002 by Feather Moss Mats in Late Seral Pine- Lichen Woodlands of North-central Brit- ish Columbia. — The Bryologist 104: 505-516. http://dx.doi.org/10.1639/0007- 2745(2001)104[0505:MDOTLB]2.0.CO;2 Tegler, B. & Kershaw, K.A. 1980. Studies on lichen-dominated systems. XXIII. The control of seasonal rates of net photosynthesis by moisture, light, and temperature in Cladonia rangiferina. — Canadian Journal of Botany 58: 1851-1858. http://dx.doi.org/10.1139/ b80-214 Waterhouse, M. 2011. Lichen and vegetation response to mountain pine beetle damage in partial cuts and mature forests in the West Chilcotin in 2008. EP1208 File Report. — Unpublished draft on file at Ministry of Forests, Lands and Natural Resource Opera- tions, Williams Lake, B.C. 23pp. Waterhouse, M.J., Armleder, H.M. & Nemec, A.F.L. 2011. Terrestrial lichen response to partial cutting in lodgepole pine forests on caribou winter range in west-central British Columbia. — Rangifer Special Issue 19: 119- 134. http://dx.doi.org/10.7557/2.31.2.1996 Waterhouse, M. J., Wallich, E. C., Daintith, N. M. & Armleder, H. M. 2010. Planted stock performance 10 years after partial cutting in west-central British Columbia. — Forestry Chronicle 96: 118-129. http:// dx.doi.org/10.5558/tfc86118-1 Whittington, J., Hebblewhite, M., DeCesare, N.J., Neufeld, L., Bradley, M., Wilmshurst, J. & Musiani, M. 2011. Caribou encounters with wolves increase near roads and trails: a

Decision-support model to explore the feasibility of using translocation to restore a woodland caribou population in Pukaskwa National Park, Canada

Emily K. Gonzales1, Patrick Nantel2, Arthur R. Rodgers3, Martha L. Allen4 & Christine C. Drake5 1 Ecological Restoration Division, Parks Canada, 300-300 West Georgia Street, Vancouver, BC, V6B 6B4, Canada 1 (Corresponding author: [email protected]). 2 Office of the Chief Ecosystem Scientist, Parks Canada, 25 Eddy, 4th Floor 25-4-S, Gatineau, QC, K1A 0M5, 2 Canada. 3 Centre for Northern Forest Ecosystem Research, Ontario Ministry of Natural Resources and Forestry, 955 Oliver 3 Road, Thunder Bay, ON, P7B 5E1, Canada. 4 Bruce Peninsula National Park, Parks Canada, P.O. Box 189, 120 Chi sin tib dek Road, Tobermory, ON, N0H 2R0, 4 Canada. 5 Pukaskwa National Park, Parks Canada, P.O. Box 212, Heron Bay, ON, P0T 1R0, Canada.

Abstract: The distribution and abundance of woodland caribou (Rangifer tarandus caribou) have declined dramatically in the past century. Without intervention the most southern population of caribou in eastern North America is expected to disappear within 20 years. Although translocations have reintroduced and reinforced some populations, approximately half of caribou translocation efforts fail. Translocations are resource intensive and risky, and multiple interrelated factors must be considered to assess their potential for success. Structured decision-making tools, such as Bayesian belief networks, provide objective methods to assess different wildlife management scenarios by identifying the key components and relationships in an ecosystem. They can also catalyze dialogue with stakeholders and provide a record of the complex thought processes used in reaching a decision. We developed a Bayesian belief network for a proposed translocation of woodland caribou into a national park on the northeastern coast of Lake Superior, Ontario, Canada. We tested scenarios with favourable (e.g., good physical condition of adult caribou) and unfavourable (e.g., high predator densities) conditions with low, medium, and high numbers of translocated caribou. Under the current conditions at Pukaskwa National Park, augmenting the caribou population is unlikely to recover the species unless wolf densities remain low (<5.5/1000 km2) or if more than 300 animals could be translocated.

Key words: Bayesian belief network; decision-support; endangered species; expert opinion; process model; protected areas; reintroduction; species at risk; structured decision-making; threatened species; woodland caribou.

Rangifer, 35, Special Issue No. 23, 2015: 27-48 DOI: 10.7557/2.35.2.3626

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 27 Introduction the population; however, translocations have Boreal populations of woodland caribou mixed success as a management tool to recover (Rangifer tarandus caribou) (hereafter “wood- caribou. Wildlife translocation is one of the land caribou”) historically occupied the boreal more complex management actions used to re- forest across North America but are now extir- store or reinforce populations of species at risk pated from the southern limits of that range (Decesare et al., 2011). The long-term success (Bergerud, 1974). Due to the declines in the of translocations requires managing the behav- distribution and abundance of this species, the iour, habitat, metapopulation, and ecosystem Committee on the Status of Endangered Wild- level issues that initially led to the decline of life in Canada (COSEWIC) and the Commit- the population (Armstrong & Seddon, 2008). tee on the Status of Species at Risk in Ontario Since 1982, the Ontario Ministry of Natural (COSSARO) assessed woodland caribou as Resources and Forestry has restored or intro- Threatened (2000, 2005 and 2014, respective- duced woodland caribou from the Slate Islands ly). Between 1900 and 1950, boreal caribou to a number of islands and the shoreline of east- retracted northward from Lake Superior (Crin- ern Lake Superior with little success (G. Eason, gan, 1957). They disappeared from the western personal communication; Gogan & Cochrane, shore of Lake Superior between 1905 and 1912 1994). (Riis, 1938a, b, c, d) and were declining and Failures of caribou translocation projects scarce on the Sibley Peninsula by 1914 (Crin- have been attributed to disease, predation, an- gan, 1957). Three unconnected populations thropogenic disturbance and/or insufficient around northeastern Lake Superior persist as and fragmented habitats (Bergerud & Mercer, the species’ most southern representatives in the 1989; Gogan & Cochrane, 1994; Compton et eastern half of North America. These popula- al., 1995). In a review of 33 caribou introductions became disjunct from the northern herds tions in eastern North America from 1924 to in the 1950s or 1960s (Bergerud, 1988). Today, 1985, introductions inevitably failed when ani- two populations are located on islands (Slates mals, released in proximity to white-tailed deer and Michipicoten); they were the products of (Odocoileus virginianus), contracted meningeal translocations and are considered to be persis- brain worm (Parelaphostrongylus tenuis) and tent with >200 individuals that fluctuate with died (Bergerud & Mercer, 1989). For example, the availability of vegetation (Environment a herd of 51 caribou, released in Cape Breton Canada, 2012). The natural population on the Highlands National Park, Nova Scotia in 1968 mainland is now restricted to a narrow band and 1969, was extinct by 1973 due to menin- along Lake Superior coast that includes Pu- geal brain worm (Dauphiné, 1975). Similar kaskwa National Park (48°N, 85°W). Biennial results occurred on Anticosti Island, Que- surveys in the Park since the late 1970s have bec (145 reindeer introduced in 1924), Great revealed a steady decline from 30 individuals to Cloche Island, Ontario (12 caribou released in only 4 in 2009 (Bergerud et al., 2007; Patterson 1970), and southern Wisconsin (14 caribou in et al., 2014). With little to no recruitment for an enclosure with white-tailed deer) (Bergerud over a decade, Bergerud et al. (2007) suggested & Mercer, 1989). that extirpation is the likely outcome for this Predation was also a key factor in failed population by 2018. Parks Canada must decide translocations. Wolf (Canis lupus), cougar (Felis between a costly intervention or risk extirpa- concolor), and occasionally bear (Ursus ameri- tion of a species from a national park. canus) predation were credited with the loss Translocation has been proposed to augment of translocated caribou in Ontario, Quebec,

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 28 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 and British Columbia in Canada and Maine projects that identify knowledge gaps and ad- in the United States (Bergerud & Mercer, dress uncertainty coupled with multidiscipli- 1989; Gogan & Cochrane, 1994; Compton et nary teams of resource managers and scientists al., 1995). Cougar predation was the primary (Seddon et al., 2007). cause of death for 60 woodland caribou trans- The planning, documenting, and decision- located from British Columbia to northern support for translocation is well served by struc- Idaho between 1987 and 1992 (Compton et tured decision analysis (Pérez et al., 2012; Con- al., 1995). Wolf predation caused the failure of verse et al., 2013). With such a tool, planners translocations in the Lake Superior region, On- and advisors can explore the factors expected to tario, including the Gargantua Peninsula (39 influence the success of a caribou translocation caribou released in 1989) (Gogan & Cochrane, and examine various combinations of environ- 1994) and Bowman Island (6 caribou released mental settings and introduction scenarios. in 1985) (Bergerud & Mercer, 1989). Preda- Federal programs to recover species at risk also tion is also the primary limiting factor for al- benefit from clear communication with stake- most all natural woodland caribou populations holders and the public. The framing of protec- (McLoughlin et al., 2003; Wittmer et al., 2005; tion and recovery of species at risk is critical Festa-Bianchet et al., 2011). Wolves and white- because it alters the way we think, talk, and tailed deer are absent from Newfoundland, approach the issue (Nie, 2001). Decision sup- which has the highest rate of successful trans- port tools are transparent, repeatable, and help locations (Bergerud & Mercer, 1989). From conceptualize the key factors and their relation- 1961 to 1982, 384 caribou were released at 22 ships – all of which facilitates framing and un- sites and 17 of these releases were successful. derstanding the issue. It was under this premise The failures in Newfoundland were attributed that we developed a Bayesian belief network to to illegal hunting and anthropogenic distur- explore the feasibility of a successful transloca- bance (Bergerud & Mercer, 1989). tion of woodland caribou into Pukaskwa Na- The failure of caribou translocations is con- tional Park. sistent with reintroductions in general. An early Bayesian belief networks (BBNs) are graphi- review of reintroduction projects suggested that cal models that represent a set of variables linked the majority failed to establish viable popula- by conditional probability relationships (Mc- tions due to poor planning and insufficient Cann et al., 2006; McNay et al., 2006; Rumpff consideration of the biological and ecologi- et al., 2011; Conroy & Peterson, 2012). They cal factors needed for success (Griffith et al., facilitate communication at the interface of sci- 1989; Wolf et al., 1998). A more recent review ence, politics and community to enhance the (1990-2005) of 454 projects found most rein- decision making process (Reckhow, 1999). A troduction programs to be ad hoc rather than BBN starts with an influence diagram, which an organized attempt to assess risk, advance un- is an intuitive graphical representation of the derstanding in the field of reintroduction biol- probabilistic dependence among variables (or ogy, or to improve reintroduction success (Sed- nodes). In a BBN, a node leading to another don et al., 2007). The authors described most one is a parent node, and the dependent node research in the field of reintroduction biology is a child node; the most external nodes (with to be retrospective, that is, opportunistic pro- no parent nodes) are used as the input to the ject evaluations and post hoc interpretation of model. Those diagrams are an effective meth- monitoring (Seddon et al., 2007). They recom- od of modeling potential causal relationships/ mended an increased role for formally planned conditional dependencies (Reckhow, 1999).

Figure 1. Influence diagram underlying a Bayesian Belief Network for a proposed woodland caribou translocation into Pukaskwa National Park. Grey shaded nodes are those presented in Table 1; the resultant (outcome) node has a thicker border.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 30 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 reached a consensus (Fig. 1). terminant of the population recovery potential. Whenever possible we were parsimonious The rate of predation was primarily determined with the model because the conditional prob- by the densities of wolves and bears (Ballard, ability table (CPT) of a child node becomes 1994), and the accessibility of caribou to difficult to parameterize with increasing num- wolves, which are considered their most signifi- bers of parent nodes. Also, the more links there cant predator (Bergerud et al., 2007). Predation are among nodes, the less tractable the model rate is also likely to be affected by the experi- becomes (Marcot et al., 2006). Parsimony was ence the introduced animals have with preda- also appropriate given the degree of precision tors (Frair et al., 2007). If caribou translocated available for each node. into Pukaskwa National Park were sourced from nearby predator-free islands, these indi- Developing the influence diagram (Fig. 1) viduals would be naïve and more susceptible to The general structure of the BBN is consistent predators. Given that the experience of translo- with other efforts to identify key variables for cated animals with predators could affect their caribou in Ontario (Rodgers et al., 2008). Cari- persistence, predator-experienced vs. predator- bou declines are ultimately caused by habitat naive caribou was a factor included as a parent alterations and proximately caused by preda- (input) node in this model. tion. More specific divisions can be traced back Wolf density is in turn affected by the den- to these two broad effects (Festa-Bianchetet al., sity of their main prey species, which could be 2011). moose or caribou depending on their relative The most external parent nodes of the BBN, availability (Bergerud & Elliott, 1986). For the also called “input nodes” herein, are the key period 1974–1988, the dynamics at Pukaskwa ecosystem variables and processes that affect National Park suggested that wolf predation caribou persistence and that either are deter- depended on caribou density (Bergerud, 1996). mined by the local conditions or can be modi- Caribou recruitment declined and adult mor- fied through management. These include -de tality increased when wolf numbers increased scriptors of the caribou’s environment, such as beyond 20 individuals (Bergerud, 1996). Pre- amount of escape habitat, the extent of linear dation dynamics can partly offset the effect that features, and landfast ice, which all influence a larger initial population of caribou would the access of wolves to caribou (Bergerud et have on recovery potential. This is why the al., 2007). Other parent nodes include “log- model includes the intermediate child node ging and prescribed fire”, “wildfire” and “qual- “number of translocated caribou surviving” ity of non-winter forage”, which all influence between the nodes “number of caribou intro- moose and bear density and the amount and duced” and “recovery potential” (Fig. 1). The quality of habitat for caribou (Rodgers et al., number of surviving animals (over ~ 5 yrs) is 2008; Environment Canada, 2012; Pinard et modulated by the survival rate of adults, and al., 2012) (Fig. 1). For details on each node’s therefore links the short-term dynamics to the states, thresholds used to separate states, and longer-term projection. conditional probability values, see tables in the Vors and Boyce (2009) reviewed a variety Appendix. of potential responses by caribou to climate The child nodes are key variables and pro- change, such as indirect, density-independent cesses that influence population dynamics effects of extreme weather events that cause more or less directly, such as rate of predation unpredictable access to forage, or freezing rain and adult survival, which in turn are a main de- events that eliminate access to grazing due to

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 31 an impenetrable layer of ice. Therefore, we in- them (Bergerud et al., 2007). These factors cluded physical condition (or body mass) as were included in the model, incorporating the a qualitative biological integrator of key en- circumstances more to less favourable for cari- vironmental variables: amount and quality of bou. habitat, parasites and diseases, and human dis- Presently, disease is not considered the pri- turbances through recreation and resource ex- mary limiting factor in the Lake Superior range traction. Those key environmental variables are mainly because white-tailed deer, the vectors determined by the local conditions and/or can of brain worm, which is lethal to caribou (An- be altered through management, so they are set derson & Strelive, 1968), were not histori- as external parent nodes in the model. cally abundant (Whitlaw & Lankester, 1994). The physical condition of caribou has impli- Nonetheless, we included disease as an input cations for determining adult survival and re- node in the model because white-tailed deer cruitment, as relationships between body mass are expanding their distribution (Thompsonet and survival and fertility have shown (Taillon al., 1998) and have been increasingly detected et al., 2012). Caribou may skip reproduction in Pukaskwa National Park (C. C. D., unpub- if they are in poor physical condition due to lished data). insufficient food resources (Bergerudet al., The terminal child node of the model is the 2007; NCASI, 2007; Taillon et al., 2012). recovery potential. It is defined as the long-term Caribou are also susceptible to anthropogenic probability of persistence of the population disturbances; they avoid resorts and recreation (i.e., whether a population will be self-sustain- activities (Nellemann et al., 2000; Carr et al., ing). As such, the node has as parent nodes the 2011), active logging (Schaefer & Mahoney, population trends, the environmental variation 2007), and are subject to increased bear preda- (which drives the random variation in population near campsites (Pitt & Jordan, 1996). In tion trends), and the number of translocated Pukaskwa National Park, human recreational caribou surviving. A high recovery potential activities could include tourists on foot and in could be defined as a time to extinction longer boats around islands and coastlines. than 50 yrs, or a 95% chance of persistence Timber volumes harvested in Ontario over over the next 50 yrs. Although the time scale the last decade have declined by more than of the processes included in the model is short- 40%, including from lands adjacent to the term (~5 yrs), the end result is a projection into park (Ontario Ministry of Natural Resources, the future. When the result of a BBN scenario 2012). Although wildfire and prescribed fires is a high probability for “high recovery poten- are permitted in some circumstances in the tial”, it suggests that this scenario will produce park, the fire cycle has departed significantly a successful translocation. from what it would have been without human Other factors that might be relevant for oth- influence, and as a result, an older-than-usual er caribou populations, such as predation by fe- forested landscape prevails. Fires are infrequent lids (Compton et al., 1995), vehicle collisions, (Perera & Baldwin, 2000) and typically smaller or avalanches (Hebblewhite et al., 2007), were in size along the coast (C. C. Drake, unpub- not relevant at Pukaskwa National Park. Ge- lished data), which is largely believed to be netic diversity was not included in the model beneficial for caribou (Environment Canada, because, although it is lower in isolated popula- 2012). Fire improves habitat for moose, which tions, there is no immediate concern for con- attracts predators. The predators consume servation (Courtois et al., 2003; McLoughlin moose but also caribou, when they encounter et al., 2004) nor did participants at the 2010

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 32 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 caribou workshop feel this was a significant all the most external parent nodes according factor in the success of a translocation (Parks to these 10 scenarios: least favourable vs. most Canada, 2011). favourable environmental conditions with two levels of translocation effort (4 scenarios), cur- Parameterizing the model rent conditions at Pukaskwa National Park with The links among the model’s nodes reflect the three levels of translocation effort (3 scenarios), knowledge we have about the probable influ- and current conditions at Pukaskwa National ence that a given parent node has on one or Park with low wolf densities with three levels more child nodes. These links are assumed to of translocation effort (3 scenarios) (Table 1). be causal. All the links in this BBN are through The decline in logging, less frequent wild- CPTs, which we conceived as contingency ta- fire, combined with limited prescribed fire in bles. For example, the probabilities of a popula- the park over the last decade (Kuchta, 2012), tion decline were determined by the number of has created older growth forests adjacent to observed cases in which a decline was observed and within the park that are favourable to cari- under each combination of two states of adult bou. Therefore, the probability of limited log- survival, recruitment, and three states of migra- ging and prescribed fire and wildfire was set tion (positive, negligible, negative). at 100%. Terrestrial lichen, a year-round food For the node “Population trends”, we used source for caribou (NCASI, 2007), is abundant data from population surveys and modeling, at Pukaskwa National Park but entirely absent categorized each case, and compiled a con- on Michipicoten Island (Bergerud et al., 2007) tingency table (Appendix). For all other child where caribou numbers are high. Therefore the nodes, data were less available so we first asked probability of good “quality of non-winter for- experts to determine what threshold values age” was set at 100%. could be used to tell each state apart. Wher- Several nodes have high levels of uncertainty ever possible, these thresholds were drawn first or show important variation among years, so from the literature. We then asked the experts virtual evidence was used as input for those to consider how nodes would interact so that nodes. Parasites or disease being transmitted we could parameterize the CPTs. For example, by deer is unlikely to seriously threaten the we asked, “among all the possible cases where physical condition of caribou in the near fu- number of caribou introduced were high, in ture because of the current low density of deer how many cases would the wolf density have in the park and surrounding landscape, but the remained low?”. Experts were asked to consider situation could change rapidly. Therefore, the the breadth of the caribou literature, not spe- probability of low “parasites and diseases” was cifically caribou in Pukaskwa National Park. set at 90%. Experts were also invited to review each other’s In Alberta, human activities alter caribou be- assessments. Most often there was consensus havior and mediate the effects of wolf predation or suggestions for additions, fine-tuning of the on caribou (Hebblewhite et al., 2005; Wasser et model, or increased precision in a threshold al., 2011). However, Pukaskwa National Park based on a new literature reference. has low human use at sensitive times (calving/ rutting), so the probability of low “human rec- Exploring scenarios reation/activity” was set to 80%. To explore the properties of the model and to In the Lake Superior range, caribou remain apply it specifically to caribou translocation at vulnerable because escape habitat is limited Pukaskwa National Park, we set evidence in and, importantly, habitat in their range has

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 33 been altered by human disturbance (Vors et al., “available” for a translocation into Pukaskwa 2007). Near-shore islands may serve as a prima- National Park would originate from islands ry escape habitat from predators (Ferguson et where caribou are abundant, such as the near- al., 1988; Carr et al., 2011) and limited linear by Slate and Michipicoten Islands and many of features likely keeps predator access low in the those naïve individuals would be lost annually. area (Bergerud, 1985). Trends toward warmer winters resulting in less landfast ice may have Results further limited the access of wolves to caribou The least favourable scenario produced only a in the coastal region (Thompsonet al., 1998). 1% probability of population recovery (Table To take into account the variation and uncer- 1). The most favourable scenario resulted in tainty in those factors, the probability of plenty 58% probability of population recovery when vs. little for the node “amount of escape habi- fewer than 50 animals were translocated and tat” was set to 20:80; the probability of limited 90% when more than 50 animals were translo- “linear features” was set to 90% and to 50% for cated (Table 1). limited “landfast ice”. This set of values gives a Under current conditions in Pukaskwa Na- probability of low “access of wolf to caribou” of tional Park, the chance of high recovery poten- about 50% (Appendix). tial increased with the number of translocated Based on population size time series, envi- animals to a high of 46% (Table 1). When we ronmental variation (i.e., the long-term yearly set the probability of high “wolf density” to random fluctuation in population growth rate 100%, regardless of its parent nodes, the prob- due to variation in survival, recruitment, and ability of high predation rate reached 72%. migration), remains low; therefore, the prob- This combination of inputs suggests that even ability of low “environmental variation” was set introducing 300 caribou would not increase at 80%. the probability of population recovery beyond Once the values for the input nodes were set, 50% (Table 1). With the same set of evidence, we examined how the probability of recovery but with probability of low wolf density set potential would increase following the intro- at 100%, introducing more than 50 caribou duction of an increasing number of caribou: raised the probability of population recovery less than 50, 50–300, and >300. These values above 50% (Table 1). were drawn from a non-spatial population viability analysis which concluded that a popu- Discussion lation of 300 animals with moderate calf and Interestingly, the probability of high recov- adult female survival had a 10% probability ery potential under the current conditions, of quasi-extinction, and that large populations and with even a large translocation effort, are (≥ 300) had a high probability of persistence roughly consistent with the 50% failure rate under favourable demographic conditions (En- of caribou translocations in eastern North vironment Canada, 2012). It could be argued America (Bergerud & Mercer, 1989; Gogan & that introducing such large numbers of animals Cochrane, 1994) as well as estimates of translo- is unrealistic, but one has to consider the (con- cation success in western North America (De- ceptual) 5 year time frame of the model, which cesare et al., 2011). However, the mechanisms would allow for a lower number of animals to leading to that result vary from one application be introduced annually over 5–10 yrs rather to another, so we cannot claim that our model than all at once during a one-time translocation emulates or explains the more general result of event. We also assumed that caribou that were many historic translocations.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 34 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Table 1. Probability of recovery potential (%) under different model scenarios and number of caribou introduced. Percent probability of five child nodes are also presented. Model scenario # of Predation Physical Adult Recruit- Population Recovery caribou rate condition of survival ment trends Potential introduced caribou Low/High Good/Bad Low/High Low/High Decline/Stab- High/Low le/Increase Least favourable1 <50 0/100 0/100 90/10 100/0 96/2/2 1/99 Most favourable2 <50 99/1 100/0 1/99 1/99 0/25/75 58/42 Most favourable2 50-300 99/1 100/0 1/99 1/99 0/25/75 90/10 Current3 <50 30/70 91/9 38/62 72/28 47/32/21 21/79 Current3 50-300 30/70 91/9 38/62 72/28 47/32/21 38/62 Current3 >300 30/70 91/9 38/62 72/28 47/32/21 46/54 Low wolf density4 <50 87/13 91/9 7/93 17/83 29/28/43 35/65 Low wolf density4 50-300 87/13 91/9 7/93 17/83 29/28/43 58/42 Low wolf density4 >300 87/13 91/9 7/93 17/83 29/28/43 67/33 1 Input nodes adjusted to the least favourable environmental conditions or worst case scenario 2 Input nodes adjusted to the most favourable environmental conditions or best case scenario 3 Input nodes adjusted to reflect the current conditions at Pukawska National Park, based on best available information. For those scenarios, wolf density node is input as 100:0 high, regardless of the value of its parent nodes. 4 A hypothetical scenario with same input as current, but in which wolf density node is set at 0:100 low, regardless of the value of its parent nodes.

The differences among the probabilities of of parasites and disease, human disturbance, high recovery potential for the most favour- and the ability of predator-naïve caribou suc- able scenario (90%) and the current conditions cessfully eluding predators. The complexity of (38%) at Pukaskwa National Park suggest that the relationships among the nodes of this BBN the translocation of caribou into Pukaskwa Na- coupled with knowledge gaps highlights the tional Park would be highly risky unless some importance of uncertainty. Complexity and un- of the unfavourable conditions were altered. Al- certainty are “familiars” in ecology; the advan- though reducing predation would increase the tage of a BBN over ad hoc decision-making is probability of recovery potential by 12–21%, that it identifies and prioritizes research needs. this increase may be insufficient to warrant a The parts of our BBN that are based mainly on potentially unpopular and ecologically harmful expert experience can be used to generate testa- management option such as predator control, ble hypotheses and can be advanced with itera- particularly in a national park. Alternatively, tive testing and updating of the model (Marcot Parks Canada could try managing the preda- et al., 2006; Martin et al., 2012). tion rate on caribou indirectly. For example, the Our BBN is a representation of a collec- park could manage habitat to reduce alternate tively agreed upon reality as opposed to a test prey (moose and deer) that attract predators, of causal relationships. We could not formally improve escape habitat, limit linear features estimate the predictive accuracy of the model that facilitate access of wolves to caribou, and since observation data are unavailable to com- provide safe sites for caribou to calve. pare predictions with observations. This may be Typical of species at risk, elements of uncer- an unsatisfying outcome for those who value tainty remain that affect recovery potential. The the precision of quantitative models; as data be- probability of recovery and persistence of trans- come available, this model can certainly be im- located caribou in Pukaskwa National Park proved. However, a network of variables with hinges on key uncertainties such as the risk numerical probabilities is not an intuitive way

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 35 to interpret results for all stakeholders (Renooij this caribou BBN could include the addition of & Witteman, 1999). Eliciting expert input for socio-economic factors. BBNs requires experts to express their beliefs in probabilistic terms that describe dependen- Conclusion cies among different factors. It has been argued Species at risk of extirpation or extinction pre- that inferential reasoning is the mechanism by sent unique challenges to land managers given which people integrate and interpret subjec- their paucity coupled with political scrutiny tive and incomplete data from various sources and economic realities (Armstrong & McCa- (Pearl, 1988). Some of our experts did not feel rthy, 2007). It is often necessary to make de- familiar enough with the concept of probability cisions for species at risk under considerable or they felt it was too difficult to quantify their uncertainty (i.e., limited demographic data and beliefs. As a result, the probabilities of the out- lack of information on dispersal (Beissinger & comes in this BBN are generally described in a Westphal, 1998) and failing to acknowledge or relative sense. The model’s precision could be address uncertainty can lead to poor decisions improved in the future; presently, it is consist- and outcomes (Regan et al., 2005). Despite ent with available data and the level of uncer- the ad hoc nature of these projects, programs tainty of the experts. to recover endangered species are expected to The translocation of caribou is logistically maximize species survival and minimize finan- difficult and expensive to implement. Recov- cial cost, while under the scrutiny of stakehold- ery of caribou requires public funds and so ers and jurisdictions with divergent opinions it is important to have local support for cari- (Maguire, 1986). We presented a BBN for a bou translocation programs (Schneider et al., potential caribou translocation in Pukaskwa 2010). In this area, the majority of regional National Park to provide structured decision residents support conservation actions for support for resource managers. caribou in Pukaskwa National Park, however, This BBN suggests that any size of transloca- only 51% would support translocation (Parks tion is unlikely to help recover the population Canada, 2011). The lack of strong support may of caribou in Pukaskwa National Park under be driven, in part, by local hunters. Over the the current conditions. Although the long- past century, caribou have declined and moose term recovery and persistence of an augmented have increased and local hunters in this region population of caribou in Pukaskwa National have shifted their harvest to moose. Hunters are Park is unknown, most of the short-term sce- aware that managing for caribou habitat does narios explored in the BBN resulted in low to not favour moose habitat, which could result moderate success, which suggests that long- in lower moose densities and fewer moose tags term recovery and persistence may be unlikely (C. C. Drake, personal observation). The social either with or without translocation. Impor- challenges of translocations can be even more tantly, long-term recovery and survival of cari- daunting than the biological ones (Reading & bou may be hampered by the lack of contiguity Clark, 1996), and successful programs benefit with more northern populations and habitat from approaches that integrate the social and conditions beyond the boundaries of Pukaskwa biological sciences (Miller et al., 1999). BBNs National Park. are well-suited to incorporating social and Although this BBN was developed for Pu- economic analyses by including model nodes kaskwa National Park’s proposed translocation, for costs and utilities (Levontin et al., 2011; we also made it flexible enough to be applied Haines-Young, 2011). A future application for to other caribou populations. It represents and

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 36 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 combines empirical data with experts’ under- Conroy, C. Hiltner, and anonymous reviewers standing of caribou ecology. It graphically ex- for comments on earlier versions of this paper. presses complex relationships and challenges for caribou recovery and management. It addresses, in a structured way, uncertainties that plague attempts to solve these problems. It evaluates alternative decisions within a context References of risk assessment to help identify options with Anderson, R. & Strelive, U.R. 1968. The ex- caribou translocation. It also fosters communi- perimental transmission of Parelaphostron- cation among ecologists, decision-makers, and gylus tenuis to caribou (Rangifer tarandus stakeholders who may lack common training, terraenovae). — Canadian Journal of Zool- terminology, or experience (Cain, 2001). ogy 46: 503-510. http://dx.doi.org/10.1139/ Regardless of whether the caribou popula- z68-070 tion in Pukaskwa National Park is augmented Armstrong, D.P. & McCarthy, M.A. 2007. through translocation, it is apparent that the Big decisions and sparse data: adapting sci- factors driving the decline of caribou and the entific publishing to the needs of practical fate of their recovery in this region will not be conservation. — Avian Conservation and easily resolved. On-going development of this Ecology 2:14. http://www.ace-eco.org/vol2/ BBN based on empirical data, as it becomes iss2/art14 available, could be an important tool in facili- Armstrong, D.P. & Seddon, P.J. 2008. Direc- tating the decision-making process for caribou tions in reintroduction biology. — Trends in management in Pukaskwa National Park and Ecology and Evolution 23: 20-25. http:// more broadly, as many caribou populations in dx.doi.org/10.1016/j.tree.2007.10.003 Canada are declining (Environment Canada, Ballard, W.B. 1994. Effects of black bear pre- 2012). dation on caribou - a review. — Alces 30: This model was developed using the freely 25-35. downloaded software GeNie 2.0 (http://ge- Bergerud, A.T. 1974. Decline of caribou in nie.sis.pitt.edu/). We invite readers to explore North America following settlement. — their own scenarios. Our inputs are available in Journal of Wildlife Management 38: 757- Appendix. Contact the authors to request the 770. http://dx.doi.org/10.2307/3800042 model. Bergerud, A.T. 1985. Antipredator strategies of caribou: dispersion along shorelines. — Acknowledgements Canadian Journal of Zoology 63: 1324-1329. Experts who contributed to one or more ver- http://dx.doi.org/10.1139/z85-199 sions of the model included G. Eason (retired Bergerud, A.T. 1988. Caribou, wolves and Area Biologist, Ontario Ministry of Natural man. — Trends in Ecology and Evolution 3: Resources), N. Freeman (Inventory Specialist, 68-72. http://dx.doi.org/10.1016/0169- BC Ministry of Sustainable Resource Manage- 5347(88)90019-5 ment), I. Schmelzer (Terrestrial Ecologist, Gov- Bergerud, A.T. 1996. Evolving perspectives on ernment of Newfoundland and Labrador), S. caribou population dynamics, have we got Hayes and L. Parent (PNP), and one anony- it right yet? — Rangifer 16: 95-116. http:// mous expert. We thank G. Seutin, Chief Eco- dx.doi.org/10.7557/2.16.4.1225 system Scientist at Parks Canada, S. Ban, M. Bergerud, A.T., Dalton, W.J., Butler, H., Camps, L. & Ferguson, R. 2007. Wood-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 37 land Caribou persistence and extirpation in & Breton, L. 2003. Significance of cari- relic populations on Lake Superior. — bou (Rangifer tarandus) ecotypes from a Rangifer Special Issue No. 17: 57-78. http:// molecular genetics viewpoint. — Conser- dx.doi.org/10.7557/2.27.4.321 vation Genetics 4: 393-404. http://dx.doi. Bergerud, A.T. & Elliott, J.P. 1986. Dynamics org/10.1023/A:1024033500799 of caribou and wolves in northern British Cringan, A.T. 1957. History, food habits and Columbia. — Canadian Journal of Zoology range requirements of the woodland cari- 64: 1515-1529. http://dx.doi.org/10.1139/ bou of continental North America. Transac- z86-226 tions of the North American Wildlife Confer- Bergerud, A.T. & Mercer, W.E. 1989. Caribou ence, March 4-6, 1957. J. B. Trefethan (ed). introductions in eastern North America. — Washington, DC. USA. 22, 485-501pp. Wildlife Society Bulletin 17: 111-120. Dauphiné Jr, T.C. 1975. The disappearance of Beissinger, S.R. & Westphal, M.I. 1998. On caribou reintroduced to Cape Breton High- the use of demographic models of population lands National Park. — Canadian Field-Nat- viability in endangered species management. uralist 89: 299-310. — Journal of Wildlife Management 62: 821- Decesare, N.J., Whittington, J., Hebble- 841. http://dx.doi.org/10.2307/3802534 white, M., Robinson, H., Bradley, M., Cain, J. 2001. Planning improvements in Neufeld, L. & Musiani, M. 2011. The role natural resources management. Centre for of translocation in recovery of woodland Ecology and Hydrology, Wallingford, UK. caribou populations. — Conservation Biol- http://64.40.97.24/downloads/BBN%20 ogy 25: 365-373. Guidelines%20-%20Cain.pdf (accessed Environment Canada. 2012. Recovery Strat- March 2013) . egy for the Woodland Caribou (Rangifer Carr, N.L., Rodgers, A.R., Kingston, S.R. & tarandus caribou), Boreal population, in Lowman, D.J. 2011. Use of island and main- Canada. Species at Risk Act Recovery Strat- land shorelines by woodland caribou during egy Series. Environment Canada, Ottawa. the nursery period in two northern Ontario http://www.acee-ceaa.gc.ca/050/docu- parks. — Rangifer, Special Issue No. 19: 49- ments/p21799/83574E.pdf (accessed March 61. http://dx.doi.org/10.7557/2.31.2.1989 2013) Compton, B.B., Zager, P. & Servheen, G. Ferguson, S.H., Bergerud, A.T. & Ferguson, 1995. Survival and mortality of translocated R. 1988. Predation risk and habitat selection woodland caribou. — Wildlife Society Bul- in the persistence of a remnant caribou pop- letin 23: 490-496. ulation. — Oecologia 76: 236-245. http:// Conroy, M.J. & Peterson, J.T. 2012. Decision dx.doi.org/10.1007/BF00379957 making in natural resource management: Festa-Bianchet, M., Ray, J.C., Boutin, S., A structured, adaptive approach. John Wiley Côté, S.D. & Gunn, A. 2011. Conservation & Sons. of caribou (Rangifer tarandus) in Canada: Converse, S.J., Moore, C.T. & Armstrong, an uncertain future. — Canadian Jour- D.P. 2013. Demographics of reintroduced nal of Zoology 89: 419-434. http://dx.doi. populations: estimation, modeling, and de- org/10.1139/z11-025 cision analysis. — The Journal of Wildlife Frair, J.L., Merrill, E.H., Allen, J.R. & Boyce, Management 77: 1081-1093. http://dx.doi. M.S. 2007. Know thy enemy: experience org/10.1002/jwmg.590 affects elk translocation success in risky land- Courtois, R., Bernatchez, L., Ouellet, J. P., scapes. — Journal of Wildlife Management 71:

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 38 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 541-554. http://dx.doi.org/10.2193/2006- economic, and sociological knowledge by 141 Bayesian belief networks: the interdiscipli- Gogan, P.J. & Cochrane, J.F. 1994. Resto- nary evaluation of potential management ration of woodland caribou to the Lake plans for Baltic salmon. — ICES Journal of Superior region. — In: Bowles, M.L. & Marine Science 68: 632-638. http://dx.doi. Whelan, C.J. (eds). Restoration of endan- org/10.1093/icesjms/fsr004 gered species: conceptual issues, planning and Maguire, L.A. 1986. Using decision analysis implementation. pp 219-242. Cambridge to manage endangered species populations. University Press, Cambridge. http://dx.doi. — Journal of Environmental Management 22: org/10.1017/CBO9780511623325.012 345-360. Griffith, B., Scott, J.M., Carpenter, J.W. & Marcot, B.G., Steventon, J.D., Sutherland, Reed, C. 1989. Translocation as a species G.D. & McCann, R.K. 2006. Guidelines conservation tool: status and strategy. for developing and updating Bayesian belief — Science 245: 477-480. http://dx.doi. networks applied to ecological modeling and org/10.1126/science.245.4917.477 conservation. — Canadian Journal of For- Haines-Young, R. 2011. Exploring ecosys- est Research 36: 3063-3074. http://dx.doi. tem service issues across diverse knowl- org/10.1139/x06-135 edge domains using Bayesian Belief Martin, T.G., Burgman, M.A., Fidler, F., Networks. — Progress in Physical Ge- Kuhnert, P.M., Low‐Choy, S., McBride, ography 35: 681-699. http://dx.doi. M. & Mengersen, K. 2012. Eliciting expert org/10.1177/0309133311422977 knowledge in conservation science. — Con- Hebblewhite, M., White, C.A., Nietvelt, servation Biology 26: 29-38. http://dx.doi. C.G., McKenzie, J.A., Hurd, T.E., Fryxell, org/10.1111/j.1523-1739.2011.01806.x J.M., Bayley, S.E. & Paquet, P.C. 2005. McCann, R.K., Marcot, B.G. & Ellis, R. Human activity mediates a trophic cascade 2006. Bayesian belief networks: applications caused by wolves. — Ecology 86: 2135-2144. in ecology and natural resource manage- http://dx.doi.org/10.1890/04-1269 ment. — Canadian Journal of Forest Research Hebblewhite, M., Whittington, J., Bradley, 36: 3053-3062. http://dx.doi.org/10.1139/ M., Skinner, G., Dibb, A. & White, C.A. x06-238 2007. Conditions for caribou persistence in McLoughlin, P.D., Dzus, E., Wynes, B. & the wolf-elk-caribou systems of the Cana- Boutin, S. 2003. Declines in populations of dian Rockies. — Rangifer 27: 79-91. http:// woodland caribou. — Journal of Wildlife dx.doi.org/10.7557/2.27.4.322 Management 67: 755-761. http://dx.doi. Kuchta, B. 2012. Bottom-up and top-down forc- org/10.2307/3802682 es shaping caribou forage availability on McLoughlin, P.D., Paetkau, D., Duda, M. the Lake Superior Coast. M.Sc.F. Thesis. & Boutin, S. 2004. Genetic diversity and Lakehead University, Thunder Bay, Ontario. relatedness of boreal caribou populations in Kujala, H., Burgman, M.A. & Moilanen, A. western Canada. — Biological Conservation 2013. Treatment of uncertainty in conserva- 118: 593-598. http://dx.doi.org/10.1016/j. tion under climate change. — Conservation biocon.2003.10.008 Letters 6: 73-85. http://dx.doi.org/10.1111/ McNay, R.S., Marcot, B.G., Brumovsky, V. j.1755-263X.2012.00299.x & Ellis, R. 2006. A Bayesian approach to Levontin, P., Kulmala, S., Haapasaari, P. & evaluating habitat for woodland caribou in Kuikka, S. 2011. Integration of biological, north-central British Columbia. — Canadi-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 39 an Journal of Forest Research 36: 3117-3133. Perera, A.H. & Baldwin, D.J.B. 2000. Spa- http://dx.doi.org/10.1139/x06-258 tial patterns in the managed forest landscape Miller, B., Ralls, K., Reading, R.P., of Ontario. In: Perera, A.H., Euler, D.L. & Scott, J.M. & Estes, J. 1999. Biologi- Thompson, I.D. (eds). Ecology of a Managed cal and technical considerations of carni- Terrestrial Landscape: Patterns and Processes of vore translocation: a review. — Animal Forest Landscapes in Ontario. pp 74-99. UBC Conservation, 2: 59-68. http://dx.doi. Press, Vancouver, BC. org/10.1111/j.1469-1795.1999.tb00049.x Pérez, I., Anadón, J.D., Díaz, M., Nicola, NCASI (National Council for Air and Stream G.G., Tella, J.L. & Giménez, A. 2012. Improvement). 2007. A review of ungulate What is wrong with current translocations? nutrition and the role of top-down and bot- A review and a decision-making proposal. — tom-up forces in woodland caribou popula- Frontiers in Ecology and the Environment 10: tion dynamics. Technical Bulletin No. 934. 494-501. http://dx.doi.org/10.1890/110175 Research Triangle Park, NC. Pinard, V., Dussault, C., Ouellet, J.P., Fortin, Nellemann, C., Jordhøy, P., Støen, O.G. & D. & Courtois, R. 2012. Calving rate, calf Strand, O. 2000. Cumulative impacts of survival rate and habitat selection of forest- tourist resorts on wild reindeer (Rangifer dwelling caribou in a highly managed land- tarandus tarandus) during winter. — Arctic scape. — Journal of Wildlife Management 53: 9-17. http://dx.doi.org/10.14430/arc- 76: 189-199. http://dx.doi.org/10.1002/ tic829 jwmg.217 Nie, M.A. 2001. The sociopolitical dimensions Pitt, W.C. & Jordan, P.A. 1996. Influence of of wolf management and restoration in the campsites on black bear habitat use and po- United States. — Human Ecology Review 8: tential impact on caribou restoration. — Res- 1-12. toration Ecology 4: 423-426. http://dx.doi. Ontario Ministry of Natural Resources. org/10.1111/j.1526-100X.1996.tb00195.x 2012. State of Ontario’s Forests. Toronto: Reading, R.P. & Clark, T.W. 1996. Carnivore Queen’s Printer for Ontario. 73pp. reintroductions. — In: Gittleman, J.L. (ed.). Parks Canada. 2010. Caribou translocation Carnivore behavior, ecology, and evolution, plan for Pukaskwa National Park. Unpub- Vol. 2: 296–336. Ithaca, NY: lished report, ON. 8 pp. Comstock Publishing Associates. Parks Canada. 2011. Pukaskwa National Park Reckhow, K.H. 1999. Water quality predic- Regional Resident Survey. Parks Canada tion and probability network models. — Ca- Agency, Cornwall, Ontario. nadian Journal of Fish and Aquatic Sciences Patterson, L.D., Drake, C.C., Allen, M.L. & 56: 1150-1158. http://dx.doi.org/10.1139/ Parent, L. 2014. Detecting a population de- f99-040 cline of woodland caribou (Rangifer tarandus Regan, H.M., Ben-Haim, Y., Langford, B., caribou) from non‐standardized monitoring Wilson, W.G., Lundberg, P., Andelman, data in Pukaskwa National Park, Ontario. S.J. & Burgman, M.A. 2005. Robust de- — Wildlife Society Bulletin 38: 348-357. cision-making under severe uncertainty for http://dx.doi.org/10.1002/wsb.402 conservation management. — Ecological Pearl, J. 1988. Probabilistic reasoning in intel- Applications 15: 1471-1477. http://dx.doi. ligent systems: networks of plausible infer- org/10.1890/03-5419 ence. Morgan Kaufmann Publishers, San Renooij, S. & Witteman, C.L.M. 1999. Francisco. 75pp. Talking probabilities: communicating

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 40 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 probabilistic information with words and grounds must account for spatial changes numbers. — International Journal of Approx- over time. — Biological Conservation 147: imate Reasoning 22: 169-194. http://dx.doi. 163-173. http://dx.doi.org/10.1016/j.bio- org/10.1016/S0888-613X(99)00027-4 con.2010.04.002 Riis, P.B. 1938a. Woodland caribou and time. Thompson, I.D., Flannigan, M.D., Wot- — Parks and Recreation 21: 529-535. ton, B.M. & Suffling, R. 1998. The effects Riis, P.B. 1938b. Woodland caribou and time. of climate change on landscape diversity: — Parks and Recreation 21: 594-600. an example in Ontario forests. — Environ- Riis, P.B. 1938c. Woodland caribou and time. mental Monitoring and Assessment 49: 213- — Parks and Recreation 21: 639-645. 233. http://dx.doi.org/10.1016/j.bio- Riis, P.B. 1938d. Woodland caribou and time. con.2010.04.002 — Parks and Recreation 22: 23-30. Vors, L.S. & Boyce, M.S. 2009. Global declines Rodgers, A.R., Berglund, N.E., Wade, K.D., of caribou and reindeer. — Global Change Allison, B.A. & Iwachewski, E.P. 2008. Biology 15: 2626-2633. http://dx.doi. Forest-Dwelling Woodland Caribou in On- org/10.1111/j.1365-2486.2009.01974.x tario: Experimental Design Workshop Re- Vors, L.S., Schaefer, J.A., Pond, B.A., Rodg- port. Information Paper CNFER IP-002. ers, A.R. & Patterson, B.R. 2007. Wood- Ontario Ministry of Natural Resources, land caribou extirpation and anthropogenic Thunder Bay, Ontario. landscape disturbance in Ontario. — Jour- Rumpff, L., Duncan, D.H., Vesk, P.A., nal of Wildlife Management 71: 1249-1256. Keith, D.A. & Wintle, B.A. 2011. State- http://dx.doi.org/10.2193/2006-263 and-transition modeling for adaptive man- Wasser, S.K., Keim, J.L., Taper, M.L. & Lele, agement of native woodlands. — Biological S.R. 2011. The influences of wolf predation, Conservation 144: 1224-1236. http://dx.doi. habitat loss, and human activity on caribou org/10.1016/j.biocon.2010.10.026 and moose in the Alberta oil sands. — Fron- Schaefer, J.A. & Mahoney, S.P. 2007. Effects tiers in Ecology and the Environment 9: 546- of progressive clearcut logging on New- 551. http://dx.doi.org/10.1890/100071 foundland caribou. — Journal of Wildlife Whitlaw, H.A. & Lankester, M.W. 1994. The Management 71: 1753-1757. http://dx.doi. co-occurrence of moose, white-tailed deer, org/10.2193/2005-479 and Parelaphostrongylus tenuis in Ontario. — Schneider, R.R., Hauer, G., Adamowicz, Canadian Journal of Zoology 72: 819-825. W.L. & Boutin, S. 2010. Triage for conserv- http://dx.doi.org/10.1139/z94-111 ing populations of threatened species: the Wittmer, H. U., Sinclair, A. R. & McLel- case of woodland caribou in Alberta. — Bio- lan, B.N. 2005. The role of predation in the logical Conservation 143: 1603-1611. http:// decline and extirpation of woodland cari- dx.doi.org/10.1016/j.biocon.2010.04.002 bou. Oecologia 144: 257-267. http://dx.doi. Seddon, P.J., Armstrong, D.P. & Maloney, org/10.1007/s00442-005-0055-y R.F. 2007. Developing the science of rein- Wolf, C.M., Garland, T. & Griffith, B. 1998. troduction biology. — Conservation Biology Predictors of avian and mammalian translo- 21: 303-312. http://dx.doi.org/10.1016/j. cation success: reanalysis with phylogeneti- biocon.2010.04.002 cally independent contrasts. — Biological Taillon, J., Festa-Bianchet, M. & Côté, Conservation 86: 243-255. http://dx.doi. S.D. 2012. Shifting targets in the tundra: org/10.1016/S0006-3207(97)00179-1 Protection of migratory caribou calving

Table A1. Conditional probability table for node recovery potential. Parent nodes and their state Recovery potential1

Number of Environmental Population High Low translocated caribou variation trends surviving Large Low Decline 0.25 0.75 Large Low Stable 0.75 0.25 Large Low Increase 1.00 0.00 Large High Decline 0.00 1.00 Large High Stable 0.50 0.50 Large High Increase 1.00 0.00 Medium Low Decline 0.00 1.00 Medium Low Stable 0.75 0.25 Medium Low Increase 1.00 0.00 Medium High Decline 0.00 1.00 Medium High Stable 0.50 0.50 Medium High Decline 0.75 0.25 Small Low Decline 0.00 1.00 Small Low Stable 0.25 0.75 Small Low Increase 0.70 0.30 Small High Decline 0.00 1.00 Small High Stable 0.10 0.90 Small High Increase 0.50 0.50 1 Thresholds for recovery potential: Low Probability of extinction >5% over 50 yrs OR: time to extinction <= 20 yrs High Probability of extinction <5% over 50 yrs OR: time to extinction > 20 yrs

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 42 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Table A2. Conditional probability table for node population trends. Parent nodes and their state Population Trend

Migration1 Survival1,2 Recruitment3 Decline Stable Increase Negligible Low Low 0.875 0.125 0.000 Negligible Low High 0.571 0.286 0.143 Negligible High Low 0.200 0.600 0.200 Negligible High High 0.250 0.250 0.500 Positive Low Low 0.000 0.875 0.125 Positive Low High 0.200 0.500 0.300 Positive High Low 0.100 0.600 0.300 Positive High High 0.000 0.250 0.750 Negative Low Low 1.00 0.000 0.000 Negative Low High 0.800 0.200 0.000 Negative High Low 0.600 0.200 0.200 Negative High High 0.500 0.250 0.250 1 States for migration: Positive: immigration accounts for >10% of the population size over 5 years; Negative: emigra- tion accounts for >10% of the population size over 5 years; Negligible: migration is less than or equal to 10% over 5 years. 2 Survival (annual rate): Low: S < 0.88; High: S >= 0.88. 3 Recruitment (calf:adult ratio): Low: R < 0.105; High: R >= 0.105.

Table A3. Conditional probability table for nodes adult survival and recruitment. Parent nodes and their state Adult Survival Recruitment

Predation Rate Physical Condition Low High Low High Low Good 0.0 1.0 0.0 1.0 Low Bad 0.1 0.9 0.5 0.5 High Good 0.9 0.1 1.0 0.0 High Bad 0.9 0.1 1.0 0.0 Note: Probability values assume that predation affects recruitment much more than it affects survival of adults.

Adult Survival N caribou introduced1 Large Medium Small Low Large 0.0 1.0 0.0 Low Medium 0.0 0.0 1.0 Low Small 0.0 0.0 1.0 High Large 0.9 0.1 0.0 High Medium 0.0 0.9 0.1 High Small 0.0 0.0 1.0 1 N caribou introduced: Small <= 50 animals, Medium 50-300 animals, Large >300 animals.

Table A5. Conditional probability table for node predation rate. Parent nodes and their state Predation Rate Caribou Wolf Density Access of wolf Bear Density Low High predator to caribou experience Yes Low Low Low 1.0 0.0 Yes Low Low High 0.9 0.1 Yes Low High Low 1.0 0.0 Yes Low High High 0.9 0.1 Yes High Low Low 0.9 0.1 Yes High Low High 0.8 0.2 Yes High High Low 0.4 0.6 Yes High High High 0.3 0.7 No Low Low Low 0.9 0.1 No Low Low High 0.75 0.25 No Low High Low 0.9 0.1 No Low High High 0.2 0.8 No High Low Low 0.5 0.5 No High Low High 0.25 0.75 No High High Low 0.0 1.0 No High High High 0.0 1.0

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 44 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Table A6. Conditional probability table for node physical condition of caribou. Parent nodes and their state Physical Condition

Parasites & Amount & quality of Human Good Bad diseases1 habitat recreation/activity Low High Low 1.0 0.0 Low High High 0.9 0.1 Low Low Low 0.2 0.8 Low Low High 0.1 0.9 High High Low 0.3 0.7 High High High 0.1 0.9 High Low Low 0.0 1.0 High Low High 0.0 1.0 1 Thresholds for parasite and diseases: based on deer density: Low: < 6 deer/km2; High: > 6 deer/km2 (Bergerud & Mercer, 1989).

Table A7. Conditional probability table for node wolf density. Parent nodes and their state Wolf density1

Moose density N of caribou intro- Low High duced Low Large 0.75 0.25 Low Medium 0.9 0.1 Low Small 1.0 0.0 High Large 0.0 1.0 High Medium 0.25 0.75 High Small 0.5 0.5 1Low: <5.5/1000 km2; High >=5.5/1000 km2 Bergerud and Mercer (1989) have suggested that even in the absence of deer (the source for P. tenuis) when wolf densities exceed 10/1,000 km2, caribou re-introductions will fail. Bergerud and Elliot (1986) indicated that generally, in the absence of escape habitat, caribou populations cannot maintain their numbers when wolf densities are >=6.5/1,000 km2.

Logging & prescribed Wildfire Low High Low High fire Limited Limited 0.90 0.10 0.90 0.10 Limited Extensive 0.75 0.25 0.75 0.25 Extensive Limited 0.40 0.60 0.50 0.50 Extensive Extensive 0.20 0.80 0.20. 0.80 1 Thresholds for moose density: Low : <0.3 moose/km2; High = >0.3 moose/km2 2 Thresholds for bear density: Low: <10/100 km2; High = >10/100 km2

Table A9. Conditional probability table for node access of wolves (to caribou). Parent nodes and their state Access of wolves

Amount of escape habitat Linear features Landfast ice Good Bad Plenty Limited Limited 1.0 0.0 Plenty Limited Extensive 0.8 0.2 Plenty Extensive Limited 0.7 0.3 Plenty Extensive Extensive 0.5 0.5 Little Limited Limited 0.5 0.5 Little Limited Extensive 0.4 0.6 Little Extensive Limited 0.3 0.7 Little Extensive Extensive 0.0 1.0

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 46 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Table A10. Conditional probability table for node amount and quality of habitat. Parent nodes and their state Amount & quality of habitat

Quality of non-winter Logging & prescribed fire1 Wildfire1 High Low forage Good Limited Limited 1.0 0.0 Good Limited Extensive 0.2 0.8 Good Extensive Limited 0.2 0.8 Good Extensive Extensive 0.1 0.9 Poor Limited Limited 0.5 0.5 Poor Limited Extensive 0.0 1.0 Poor Extensive Limited 0.0 1.0 Poor Extensive Extensive 0.0 1.0 1 Logging and prescribed fire node, and for Wildfire node, the threshold for limitedvs. extensive is 40% of the range. When total disturbance exceeds 40% of the range, the probability that a Woodland Caribou population would be stable or increasing drops below 0.5 (Environment Canada, 2012).

Conservation status of caribou in the western mountains of Canada: Protections under the Species At Risk Act, 2002-2014

Justina C. Ray1, Deborah B. Cichowski2, Martin-Hugues St-Laurent3, Chris J. Johnson4, Stephen D. Petersen5 & Ian D. Thompson6

1 Wildlife Conservation Society Canada, 344 Bloor Street West #204, Toronto, Ontario, Canada, M5S 3A7 (Corresponding author: [email protected]). 2 Caribou Ecological Consulting, Box 3652, Smithers, B.C., Canada, V0J 2N0. 3 Université du Québec à Rimouski, Département de biologie, chimie & géographie, Centre for Northern Studies, 300 allée des Ursulines, Rimouski, QC, Canada, G5L 3A1. 4 Natural Resources and Environmental Studies Institute, University of Northern British Columbia, Prince George, BC, Canada, V2N 4Z9. 5 Assiniboine Park Zoo, 2595 Roblin Blvd, Winnipeg, MB, Canada, R3R 0B8. 6 Canadian Forest Service, 1219 Queen St. East, Sault Ste. Marie, Ontario, Canada, P6A 2E5.

Abstract: In April 2014, the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) reviewed the status of caribou in the western mountains of Canada, in keeping with the ten-year reassessment mandate under the Species at Risk Act. Assessed as two ‘nationally significant’ populations in 2002, COSEWIC revised the conservation units for all caribou in Canada, recognising eleven extant Designatable Units (DUs), three of which -- Northern Moun- tain, Central Mountain, and Southern Mountain -- are found only in western Canada. The 2014 assessment concluded that the condition of many subpopulations in all three DUs had deteriorated. As a result of small and declining population sizes, the Central Mountain and Southern Mountain DUs are now recognised as endangered. Recent declines in a number of Northern Mountain DU subpopulations did not meet thresholds for endangered or threatened, and were assessed as of special concern. Since the passage of the federal Species at Risk Act in 2002, considerable areas of habitat have been managed or conserved for caribou, although disturbance from cumulative human development activities has increased during the same period. Government agencies and local First Nations are attempting to arrest the steep decline of some subpopulations by using predator control, maternal penning, population augmentation, and captive breeding. Based on declines, future developments and current recovery effects, we offer the following recommendations: 1) where recovery actions are necessary, commit to simultaneously reducing human intrusion into caribou ranges, restoring habitat over the long term, and conducting short-term predator control, 2) carefully consider COSEWIC’s new DU structure for management and recovery actions, especially regarding translocations, 3) carry out regular surveys to monitor the condition of Northern Mountain caribou subpopulations and immediately implement preventative measures where necessary, and 4) undertake a proactive, planned approach coordinated across jurisdictions to conserve landscape processes important to caribou conservation.

Key words: Central Mountain; COSEWIC; Designatable Units; Northern Mountain; Rangifer tarandus; Southern Mountain; Species At Risk Act.

Rangifer, 35, Special Issue No. 23, 2015: 49-80 DOI 10.7557/2.35.2.3647

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 49 Introduction Caribou Recovery Team, 2005; ASRD & ACA, When Canada’s Species At Risk Act (SARA) 2010; Parks Canada, 2011a; Mountain Cari- came into force in 2003, the legal list (SARA, bou Recovery Implementation Plan Progress 2002) comprised 233 wildlife species (as de- Board, 2012). In the past decade, management fined under the Act) in Schedule 1. Among plans or recommendations have also been de- these were Woodland Caribou (Rangifer taran- veloped for individual subpopulations or sub- dus caribou) residing in the western mountains population groups (e.g., Chisana Caribou Herd of Canada, which the Committee on the Status Working Group, 2012; BC Ministry of Envi- of Endangered Wildlife in Canada (COSE- ronment, 2013). Under SARA, a Management WIC) considered as two “nationally significant” Plan for caribou in the Northern Mountain Southern Mountain and Northern Mountain population (Environment Canada, 2012), and populations (COSEWIC, 2002). Ranging a Recovery Strategy for the Southern Mountain from southern British Columbia and Alberta to population (Environment Canada, 2014) were Yukon and the Northwest Territories, caribou both released. Targeted measures, including historically had a relatively widespread distri- habitat and population management and pro- bution and occurred in large (>1,000 individu- tection, have also been implemented under the als) subpopulations (Spalding, 2000). By 2000, authority of various provincial legislation and about 30% of their early 1900s range was no policies (COSEWIC, 2014a). longer occupied (Figure 1; Spalding, 2000; First created in 1977, COSEWIC was for- Dzus, 2001). In 2002, COSEWIC assessed the mally established under SARA (SARA, 2002, Southern Mountain population as threatened s. 14), with the functions of conducting assess- and the Northern Mountain population as of ments, reassessments, and classifications of spe- special concern (COSEWIC, 2002) and they cies at risk “on the basis of the best available in- were listed on the SARA registry the next year formation on the biological status of a species, (Government of Canada, 2014). Subpopu- including scientific knowledge, community lations comprising the Southern Mountain knowledge and aboriginal traditional knowl- population were generally small in size, increas- edge” (SARA, 2002, s. 15). For each species, ingly isolated from one another, and subject to relevant information is assembled in a status threats, with the majority in decline (COSE- report, which is subjected to an extensive ex- WIC, 2002). Although numbers of Northern pert review process (COSEWIC, 2011b). Each Mountain caribou appeared to be stable, forest- species is assessed according to criteria based on ry, roads, gas, and other developments were be- the IUCN Red List system to measure the like- ginning to affect some subpopulations through lihood of species going extinct under prevail- habitat modification and increasing human ac- ing circumstances (Mace et al., 2008). Under cess (COSEWIC, 2002). SARA, the government of Canada considers Various recovery planning and actions di- COSEWIC’s designations within designated rected at these populations since listing under timeframes when establishing the legal list of SARA have been undertaken by provinces and wildlife species at risk (COSEWIC, 2014b). territories. For example, both Alberta and Brit- In April 2014, COSEWIC reviewed the ish Columbia have released strategic recovery conservation status of caribou in the western documents that suggest a variety of different mountains of Canada (COSEWIC, 2014a), actions aimed at recovering subpopulations in keeping with the 10-year reassessment man- in southern and central portions of the prov- date under SARA (SARA, 2002, s. 24). This inces (e.g., MCTAC, 2002; Alberta Woodland reassessment benefited from an acceleration

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 50 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Figure 1. Approximate historic and current ranges of caribou in the mountain DUs of western Canada (from COSE- WIC 2014a). Map created by Bonnie Fournier (Environment and Natural Resources, Government of the Northwest Territories, 2013).

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 51 of research and monitoring over the past dec- range in Canada (Geist, 2007; Gunn, 2009; ade that yielded new information on popula- Couturier et al., 2009; COSEWIC, 2011a), tion trends and further insights into threats. but is still the most commonly used taxonomy Moreover, it took advantage of the recognition because some aspects do appear to have validity of new conservation units for Rangifer found and no alternative has been identified in a sys- across Canada, a special project undertaken by tematic manner (COSEWIC, 2011a). Previous COSEWIC to define discrete and evolutionar- COSEWIC evaluations used Banfield’s (1961) ily unique “Designatable Units” (COSEWIC, subspecies as the basis for assessment. Caribou 2011a) for caribou throughout the country. in western mountain regions of North America This work used available information to derive were included in woodland subspecies, but the conservation units of the species to orient fu- nationally significant populations (Northern ture COSEWIC status assessments and reas- Mountain and Southern Mountain) were fur- sessments, thereby addressing widely accepted ther divided into two western mountain cari- deficiencies in the current taxonomy (see the bou ecotypes based on COSEWIC’s National ‘Classification of Caribou’ below). Aboriginal Ecological Areas with the same names (COSE- knowledge was also collected and summarized WIC, 2002). from First Nations and Métis sources by the The widely-recognized shortcomings of COSEWIC Aboriginal Traditional Knowledge caribou taxonomy have triggered a reliance on (ATK) Subcommittee (COSEWIC, 2014a). ecotypes, based on behaviour and ecology, for The recent reassessment of caribou in the conservation and management purposes. In a western mountains of Canada provided an broad sense, woodland caribou in North Amer- opportunity to evaluate how subpopulations ica are informally recognised as ‘mountain’ or comprising these newly recognised Designat- ‘boreal’ with the designation distinguishing able Units have fared since the implementation between those subpopulations that exhibit sea- of SARA just over a decade ago. Our objec- sonal or annual use of mountainous terrain vs. tives here are to review: 1) the Designatable lowland boreal habitats (Festa-Bianchet et al., Unit structure for western mountain caribou in 2011). In western Canada, this nomenclature Canada, 2) the 2014 COSEWIC assessments largely coincides with the COSEWIC South- of these units, including population numbers ern, Central, and Northern Mountain DUs and trends that served as their basis, and 3) the (mountain caribou) considered here, and the recovery and management actions planned and Boreal DU (boreal caribou). Caribou subpop- implemented to date. We conclude with a for- ulations in BC are classified by the Province ward-looking perspective on the conservation into three formally-designated ecotypes accord- outlook for these populations. ing to behaviour and habitat use, with mountain subpopulations belonging to ‘Northern’ Taxonomy and conservation units of west- or ‘Mountain’, and the remainder as ‘Boreal’ ern mountain caribou (Government of British Columbia, 2014). The Prevailing taxonomy (Banfield, 1961) recog- BC Northern ecotype corresponds with the nizes four native extant and one extinct cari- Northern and Central Mountain DUs and the bou subspecies in North America, based pri- Mountain with the Southern Mountain DU marily on skull measurements and pelage, but (Stevenson & Hatler, 1985; Heard & Vagt, also antler shape and hoof shape. It is widely 1998). Similarly, ‘mountain’ caribou in Alberta considered to be outdated and insufficient for are distinguished from their ‘boreal’ counter- capturing the variability of caribou across their parts by feeding primarily on terrestrial lichens

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 52 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 and spend at least part of their annual cycle in Using COSEWIC DU criteria for discreteness the mountains (ASRD & ACA, 2010). and significance (COSEWIC, 2011c), western mountain caribou were separated into three COSEWIC Designatable Units units: Northern Mountain caribou of Yukon, SARA recognizes that entities below the species Northwest Territories and northern and cen- level require conservation, and provides COSE- tral British Columbia (DU7), Central Moun- WIC with the mandate to assess them (SARA, tain caribou of east-central British Columbia 2002, s. 15). Accordingly, COSEWIC’s DU and west-central Alberta (DU8), and Southern concept (formalized in 2009) acknowledges Mountain caribou of southeastern British Co- that there are spatially, ecologically, or geneti- lumbia (DU9) (COSEWIC, 2011a). cally discrete and evolutionarily significant Individual subpopulations that comprise units that are irreplaceable components of bio- each of the three DUs are generally discrete diversity (COSEWIC, 2011c). Discreteness from one another, including those recognized may refer to distinctiveness in genetic charac- as members of other DUs (see COSEWIC, teristics or inherited traits, habitat discontinu- 2011a). The Southern Mountain DU and ity, or ecological isolation. Significance is also Central Mountain DUs are discrete from other included in the definition of DU as a reflection neighbouring DUs in that phylogenetically, of the opinion that isolation alone is insuffi- these caribou have both northern (Beringian- cient for designation. Evolutionary significance Eurasian) and southern (North American) may apply when there is: 1) deep phylogenetic lineages (Dueck, 1998; McDevitt et al., 2009, divergence (e.g., glacial races), 2) evidence that Yannic et al., 2014). Caribou sampled in the the population persists in a unique ecological Northern Mountain DU all come from the setting that has likely given rise to local adapta- Beringian-Eurasian lineage (Dueck, 1998; Zitt- tions, especially those related to fitness, or 3) lau 2004). where there is only one natural surviving occur- The new Southern Mountain DU, restricted rence in a particular ecological setting. to southeastern British Columbia and northern In previous COSEWIC assessments (COSE- Idaho (Figure 2), is now comprised of 15 ex- WIC, 2002; 2004) prior to the passage of tant subpopulations, all of which belonged to SARA and use of Designatable Units, caribou the previous Southern Mountain population. in Canada were organized into eight “Nation- Caribou from this DU have a distinct behav- ally Significant Populations”, not including the iour related to their use of habitats found in barren-ground subpopulations, which have steep mountains with deep snowfall (accumu- not been assessed (Festa-Bianchet et al., 2011; lated snowpack of 2-5 m). These extreme snow COSEWIC, 2011a). In preparation for nation- conditions have led to a foraging strategy that al-scale assessments and reassessments of this is unique among cervids, that is, the exclusive wildlife species initiated in 2012, COSEWIC reliance on arboreal lichens for 3-4 months of undertook a 2-year exercise to evaluate DUs for the year (Rominger et al., 1991; Terry et al., caribou in Canada using the new DU guidelines 2000). Caribou of the Southern Mountain DU (COSEWIC, 2011a). The process considered differ from Central and Northern Mountain established taxonomy, phylogenetics, genetics, DU caribou based on inherited traits for be- morphology, life history, ecology, and behav- havioural strategies and habitat selection that iour of the species, as well as biogeographical have resulted from the steep terrain and deep information such as range disjunction and the snow (COSEWIC, 2011a). Hence, this group eco-geography in which the species is found. of caribou differs markedly from all other cari-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 54 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 bou, as they have persisted in an ecological set- this new DU structure and that of the previous ting unique to the species that has given rise to assessment (COSEWIC 2002). One change local adaptations. resulted from the reclassification of terrestrial The Central Mountain DU includes ten ex- lichen feeding/shallow snow caribou that were tant subpopulations of caribou in east-central previously part of the Southern Mountain British Columbia and west-central Alberta population. The new Southern Mountain DU located in and near to the northern Rocky is restricted to central and southeastern BC Mountains. There are 45 Northern Moun- (Figure 2) and includes only the deep snow/ tain DU subpopulations ranging from west- arboreal lichen feeding ecotype. In contrast, all central and northern British Columbia to the shallow snow/terrestrial lichen feeding caribou northern mountains of Yukon and southern were reassigned to either the Central Mountain Northwest Territories (Figure 2; Environment or Northern Mountain DUs. The second major Canada, 2014). Subpopulations in the south- difference is that the new Northern Mountain ern part of the Northern Mountain DU have DU includes nine subpopulations in central relatively discrete ranges, while range overlap is British Columbia from the former Southern more pronounced farther north. Animals from Mountain population of Woodland Caribou these two DUs share similar winter feeding be- (COSEWIC 2002), which is currently listed haviours and seasonal movement patterns, but under SARA as threatened and the subject of a they differ phylogenetically and are isolated by recently-released federal recovery strategy (En- the Peace River (see COSEWIC, 2014a). Evi- vironment Canada, 2014). dence from McDevitt et al.(2009) was sugges- tive of a ‘hybrid swarm’ of two caribou lineages Population abundance and trends within the ice free corridor that appeared along Survey - methods and data availability the eastern front of the Rockies producing a The IUCN/COSEWIC criteria most relevant unique, mixed gene pool at the end of the Wis- for this assessment (for A and C; Mace et al., consin glaciations ca. 14 000 years ago (Cen- 2008; COSEWIC, 2011b) rely on population tral Mountain DU). Although some evidence estimates and trends over time. The 2002 and indicates genetic relatedness between Northern 2014 status assessments (COSEWIC, 2004; Mountain DU subpopulations in west-central 2014a) and supporting literature explain the British Columbia and those in the Central methods, including survey frequency, used to Mountain DU, the majority of sampled sub- estimate the minimum or estimated number of populations of Northern Mountain DU cari- caribou in each subpopulation as well as trends bou differ genetically (Serrouyaet al., 2012). in absolute or relative abundance. In summary, All caribou in nine sampled subpopulations the population estimates are challenging to obtain Northern Mountain DU belong to the north- for these animals as they reside in remote ar- ern clade (Dueck, 1998; Zittlau, 2004; Weck- eas, occupy large ranges at low densities, and worth et al., 2012), but only two of 25 sub- vegetation overstory across forested habitats populations in northern British Columbia have makes observation difficult. Estimates for some been sampled, leaving a large gap in phyloge- subpopulations may be based on information netic information. Further work needs to be derived from expert opinion or on sightings conducted to assess phylogenetics and genetic of caribou during surveys conducted for other population structure of the Northern Moun- species (e.g., Thiessen, 2009). For subpopula- tain DU in particular. tions where late-winter distribution occurs in There are two major differences between high-elevation alpine/subalpine habitat (mostly

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 55 in the Southern and Central Mountain DUs), been conducted only since the mid-2000s (Seip relatively unbiased minimum counts are report- & Jones, 2014). The Jasper National Park sub- ed (e.g., Seip & Jones, 2014). In other cases, populations (Tonquin, Maligne, Brazeau) are however, population estimates are imprecise or surveyed annually during the fall. In addition, do not include a measure of sampling or pro- population trend, mortality rates of radio-col- cess variance (Tables 1-3; COSEWIC, 2014a). lared caribou and late-winter calf recruitment As with all COSEWIC assessments of wildlife rates have been tracked for all subpopulations species with appropriate data (COSEWIC, other than Scott (BC) (ASRD & ACA 2010; 2011b), the number of mature individuals, Seip & Jones, 2014; Alberta Environment and either estimated or counted, was used as an Sustainable Resource Development, unpub- approximate estimate of population size or lished data). By comparison, surveys are incom- percentage change in population size over two plete or infrequent for the majority of the sub- or three generations -- the IUCN timeframes populations of the Northern Mountain DU. over which declines are measured (Mace et al., Twenty-nine of the 45 estimates are older than 2008). It is important to note that the quanti- 5 years, or were based solely on expert opinion, tative criteria used in the COSEWIC status as- and may not reflect the current population size. sessments (COSEWIC, 2011b) are dependent Several other population estimates are based on on thresholds in total number or percentage caribou counted during surveys for other spe- change in mature individuals. From this par- cies. For 18 of the 45 subpopulations, only one ticular perspective, precision and uncertainty estimate is available and some early surveys did becomes most important to consider when not always include all of the range and so are estimates approach a set threshold for designa- not comparable to more recent estimates. Only tions (endangered, threatened, and of special nine of 45 subpopulations have been surveyed concern). more than three times in the past 27 years. Survey frequency has varied among the subpopulations for all DUs (Tables 1-3). In the Population trends Southern Mountain DU, the earliest available Tables 1-3 summarize available subpopulation surveys date back to the late 1980s for some size and trend data for the Southern Mountain, subpopulations or portions of those subpopu- Central Mountain, and Northern Mountain lations (e.g., Barkerville, Wells Gray [south], DUs over the approximate three-generation Groundhog, Quesnel Highlands portion of (27 year) span used for the 2014 COSEWIC the Wells Gray [north] subpopulation) (Seip, assessment (COSEWIC 2014a). Where more 1990; Hatter, 2006; McLellan et al., 2006; than one survey estimate within three genera- Freeman, 2012). During the 1990s, at least tions was available for a subpopulation, we cal- two surveys were conducted for most subpopu- culated a measure of population change. Few lations (Hatter, 2006) and surveys were carried subpopulations had surveys as early as 1987. out in most years for Barkerville, Wells Gray For those that did not, we used the most recent (north), Central Purcells, South Purcells, and survey estimate and the highest earliest survey South Selkirk (Wakkinen, 2003; Kinley, 2007; estimate to represent three-generation change, Freeman, 2012). Since 2002, most subpopula- and did not extrapolate further. For subpopu- tions in the Southern Mountain DU have been lations with one or no survey estimates, when surveyed approximately every 2 years. available, population change was inferred from In the Central Mountain DU, surveys for mortality rates of radio-collared caribou and most British Columbia subpopulations have late winter calf recruitment (e.g., Hervieux et

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 56 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 al., 2013). For subpopulations characterised ability analysis (PVA) for ten subpopulations of by few and/or unreliable survey estimates, or Southern Mountain DU caribou. All ten were where the most recent survey took place five predicted to decline to extinction within <200 or more years ago, trends could not be deter- years and all but two subpopulations had a cu- mined. We calculated the estimated popula- mulative probability of extinction of >20% (24- tion trend for each DU since the last COSE- 100%) within 45 years (5 generations). Increas- WIC assessment by comparing total number of es in the amount of young forest have resulted mature individuals in 2014 to those reported in more rapid predicted extinction rates in all by COSEWIC (2002), taking into account populations. Hatter (2006) conducted a PVA changes in DU boundaries. for all extant subpopulations in this DU and showed that time to quasi-extinction (N<20 Southern Mountain caribou DU animals) was < 50 years for 10 of 15 subpopu- The 2014 estimate for the Southern Mountain lations. The probability of quasi-extinction in DU population was 1,354 mature individuals 20 years was >20% for 12 of 15 subpopulations (Table 1). The three-generation decline rate for and >50% for 13, but Hatter (2006) cautioned the overall population was at least 46%. Only that confidence limits indicated a low level of two subpopulations had more than 250 mature certainty for predictions for five of the sub- individuals, nine numbered fewer than 50, six populations with a high probability of extinc- of these fewer than 15. Some former larger sub- tion. By contrast, the largest subpopulations, populations had split into several due to lack of North Cariboo Mountains and Hart Ranges, dispersal within ranges (Wittmer et al., 2005). were identified in both studies as having a very Two additional subpopulations were recently low probability of extinction in this time peri- extirpated: the George Mountain subpopula- od. However, since 2006, both subpopulations tion in 2003 and the Central Purcells subpopu- have declined, with the Hart Ranges popula- lation in 2005 (Table 1). tion declining 35% (COSEWIC 2014a). All subpopulations in the revised South- ern Mountain DU belonged to the former Central Mountain caribou DU Southern Mountain population of Woodland The 2014 estimate for the Central Mountain Caribou (Environment Canada, 2014). The DU was 470 mature individuals (Table 2). Nine corresponding subpopulations were estimated of ten extant subpopulations each contain fewer at 1,850 mature individuals in 2002 (COSE- than 100 mature individuals, four among them WIC, 2002), indicating a 27% decline. The fewer than 50. The long-term trend of the Scott only increasing subpopulation (Barkerville) has subpopulation in BC, however, is unknown. In likely benefitted from a recent wolf steriliza- addition, the Banff subpopulation was extir- tion and removal program (Roorda & Wright, pated in 2009 (Hebblewhite et al., 2010), and 2012), although there are <100 mature indi- the Burnt Pine subpopulation was confirmed viduals. Some subpopulations have been sub- functionally extirpated in 2014 (Seip & Jones, jected to intensive management measures since 2014). The estimated overall decline in the Cen- 2002 (see below). tral Mountain DU population was at least 64% Because IUCN criteria also take into ac- during the last three generations. All subpopu- count projected declines into the future (Mace lations in the Central Mountain DU belonged et al., 2008; COSEWIC, 2011a), recent popu- to the former Southern Mountain population lation viability analyses were informative. Witt- of Woodland Caribou (Environment Canada, mer et al. (2010) developed a population vi- 2014). The corresponding subpopulations were

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 57 estimated at 1,293 mature individuals in 2002 nine subpopulations at the southern part of the (COSEWIC, 2002). The decrease in numbers DU, all of which belong to the former South- observed during surveys is supported by con- ern Mountain population of Woodland Cari- sistently high adult mortality and low calf re- bou (Environment Canada, 2014) have expe- cruitment (ASRD & ACA, 2010; Hervieux et rienced an overall decline of 34% since 2002, al., 2013; Seip & Jones, 2014). from 4,030 to 2,673 mature individuals. (Table 3; COSEWIC 2014a). Northern Mountain caribou DU About 50,000 to 55,000 caribou occurred in 2014 COSEWIC assessments of western the Northern Mountain DU in 2014, of which mountain caribou 43,187 to 47,496 were estimated to be ma- In April 2014, the Central and Southern Moun- ture individuals (Table 3). These animals ac- tain Caribou DUs were assessed by COSEWIC counted for about 95% of western mountain as endangered (COSEWIC, 2014a). In both caribou in Canada. Over half (26 of 45) the cases, the IUCN Red List criteria (Mace et al., subpopulations contained more than 500 ma- 2008) for high decline rate (A) and small and ture individuals, while 13 subpopulations had declining populations (C) were invoked be- fewer than 250. Nine of the 15 subpopulations cause these DUs have experienced pronounced that consisted of >1,000 mature individuals are population reductions within the last three located in Yukon and Northwest Territories. generations and most subpopulations are cur- Combined, the Bonnet Plume and Redstone rently small in size. subpopulations, the two largest in the DU, Criterion A is measured as a percentage of comprised >15,000 animals, or 26-29% of the loss of mature individuals over time windows Northern Mountain DU (Table 3). in the past, future, or a combination of the past The four subpopulations that comprised < 50 and future (Mace et al., 2008). The decline mature individuals are located in the southern of 64% over the past three generations in the part of the DU in west-central British Colum- Central Mountain population exceeds the cri- bia (Charlotte Alplands, Rainbows, Telkwa) terion of 50% decline for endangered, in cases and north-eastern British Columbia (Finlay). where the causes of the declines have not ceased Trend data were limited for subpopulations and may not be reversible (COSEWIC 2011b). in this DU, with long-term (three-generation) Although the calculated >45% decline for the trend known for only 16 of 45 subpopulations Southern Mountain population did not exceed (Table 3). Recent surveys indicate that all five the IUCN threshold (50%) for past declines, subpopulations in west-central British Colum- it qualified as endangered under this criterion bia (Telkwa, Tweedsmuir, Itcha-Ilgachuz, Rain- based on inferred reduction of >50% within bows, Charlotte Alplands) are currently declin- the next three generations based on PVA (Hat- ing (COSEWIC, 2014a). ter, 2006; Wittmer et al., 2010). The 2002 COSEWIC assessment estimated The focus of IUCN Criterion C is on popu- the number of mature individuals in the for- lations that are numerically small and in con- mer Northern Mountain population as 43,950 tinuing decline (Mace et al., 2008; COSEWIC, (COSEWIC, 2002), suggesting an overall sta- 2011b). Both Central and Southern Mountain ble situation for those 36 subpopulations, al- caribou are endangered under this criterion, as beit with considerable uncertainty because of each population numbered fewer than 2,500 limited survey data (Environment Canada, and has experienced an estimated continuing 2012; COSEWIC, 2014a). In contrast, the two-generation decline that exceeded the 20%

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 58 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 threshold (at least 62% for Central and 40% roads, trails, geophysical exploration lines, for Southern). Furthermore, in the case of Cen- pipelines, and utility rights-of-way can exacer- tral Mountain caribou there was an apparent bate susceptibility to predation, and therefore continuing decline in number of mature indi- alter the movements, distributions, and popu- viduals, while no subpopulation was estimated lation dynamics of caribou. These features fa- to contain more than 250 individuals (COSE- cilitate increased predator mobility, hunting, WIC, 2014a). vehicle collisions, disturbance, and directly or Northern Mountain Caribou did not meet indirectly result in habitat reduction and frag- quantitative thresholds for endangered or mentation (Dyer et al., 2002; Seip et al. 2007, threatened when considering overall popu- van Oort et al., 2010; Williamson-Ehlers, lation size or decline, but were assessed as of 2012; Apps et al., 2013). Predation is often the special concern due to the deteriorating status primary reason for caribou declines, directly re- of a number of subpopulations and increasing lated to increased prey populations that show a magnitude and scope of threats throughout numerical and distributional response to early the DU (COSEWIC, 2014a). All known sta- seral forest and linear features that result from ble or increasing subpopulations are located in cumulative development activities (Serrouya the northern part of the range, whereas nine in et al., 2011; Apps et al., 2013; Ehlers et al., the southern part of the range had declined by 2014). Human developments associated with 34% since the last assessment. However, most timber harvest, oil and natural gas extraction, subpopulations in this DU receive little to no wind energy, and mining have a large cumula- monitoring attention, and many 2014 esti- tive footprint, reducing the amount of habitat mates were based on survey data older than 5 for caribou and increasing the area of early- years. The status of northern subpopulations successional forests favoured by other ungulate may be compromised in the future because of species and the predators of caribou (Nielsen et increasing threats, particularly land-use change al., 2005; Nitschke, 2008; Williamson-Ehlers, resulting from industrial development, and ex- 2012). Although forest harvesting and mineral tent and frequency of forest fires and insect out- and hydrocarbon exploration and development breaks related to climate change (e.g., moun- do not generally result in substantial direct tain pine beetle) (Environment Canada, 2012; mortality of mountain caribou, habitat changes COSEWIC 2014a). Habitat loss and increased arising from these activities and associated in- predation levels can be expected to influence frastructure affect the abundance, habitat use, the distribution and abundance of subpopula- and movements of both predators and alternate tions in a similar fashion to that which has tak- prey (Festa-Bianchet et al., 2011; Serrouya et en place in the Central and Southern Mountain al., 2011). Recent large natural disturbances by DUs (Apps & McLellan, 2006, Wittmer et al., fire and forest insects may render already lim- 2007; DeCesare et al., 2011; Hervieux et al., ited habitat unavailable for decades, thereby re- 2013). ducing already fragmented ranges. For example, after over 50 years of relatively little fire activ- Prevailing and future threats ity on the Tweedsmuir-Entiako caribou range, Threats to woodland caribou in Canada, in- a wildfire in 2014 affected over 130,000 ha of cluding western mountain caribou, have been winter and spring migration range (R. Krause, well documented (Festa-Bianchet et al., 2011; BC Ministry of Forests, Lands and Natural Re- COSEWIC, 2014a). Recent studies have dem- source Operations, pers. comm.). onstrated that linear features resulting from In the Northern Mountain DU, human dis-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 59 turbances and habitat loss (including functional Ehlers et al., 2012; Johnson et al., 2015). Cari- habitat loss) have resulted from the cumulative bou in the Southern Mountain DU are subject effects of forest harvesting, mineral exploration to altered predator-prey dynamics due to habi- and development and associated access, motor- tat change resulting from forest harvesting in ized and non-motorized recreational activities, adjacent valley bottoms, snowmobiling, heli- changes in forest structure due to mountain skiing, impacts from climate change, and Al- pine beetle (Dendroctonus ponderosae) infesta- lee effects that have led to a high likelihood of tions and/or associated salvage logging, and extirpation due to random environmental and impacts from climate change (Environment demographic events (Apps & McLellan, 2006; Canada, 2012; COSEWIC, 2014a). Direct Wittmer et al., 2007; 2013). impacts to southern subpopulations in the DU are already evident, whereas those in the north- Management and recovery actions ern part of the DU may be affected similarly Efforts aimed at recovering or managing de- if the multiple proposed mineral and hydro- clining western mountain caribou since the carbon exploration and development projects, 1980s have focused on habitat protection, windfarms, and associated infrastructure are population management, and mitigation of developed in north-central and northeastern individual development projects as the indus- BC (COSEWIC, 2014a). For example, in trial footprint continues to increase across the north-western BC, there are known large min- distribution of all three DUs. In 2007, the eral deposits stimulating exploration activities Government of British Columbia announced a and mine development in the Skeena region. series of habitat protection measures as part of The 344-km Northwest Transmission line was a Mountain Caribou Recovery Implementation completed in 2014 to supply power to planned Plan (BC Ministry of Environment, 2015). industrial developments and remote commu- Specifically, 2.2 million ha of forested lands nities in the area (BC Hydro, 2015). The new in the Southern Mountain DU were included power supply is likely to increase the feasibility in protected areas or designated as Ungulate of potential projects in and adjacent to caribou Winter Ranges or Wildlife Habitat Areas under ranges in north-western BC. the provincial Forest and Range Practices Act, The primary threats to caribou in the Central whereby mountain caribou habitat require- Mountain DU include altered predator-prey ments receive special consideration when plan- dynamics due to habitat loss and disturbances ning and implementing forest harvesting and from multiple industrial activities including other industrial (e.g., road building) activities forest harvesting, mining of coal, and the ex- (Environment Canada, 2014; BC Ministry of ploration and development of oil and gas re- Environment, 2015). Approximately 1 million serves. Additional factors include deaths from ha were closed to motorized vehicles (primar- vehicle collisions, disturbance from motorized ily to restrict snowmobiling; Seip et al., 2007). recreation (e.g., all-terrain vehicles, snowmo- Ungulate Winter Ranges and Wildlife Habi- biling), facilitated access to caribou winter tat Areas generally provide for no or modified range for predators resulting from increased forest harvesting and include primarily high linear corridors and packed trails or ploughed elevation habitat in the Central and Southern roads in winter, impacts from climate change, Mountains, but also low elevation areas in the and stochastic environmental events associ- Northern Mountains. They also provide some ated with small population sizes (DeCesare et restrictions on mineral exploration and guided al., 2011; Hervieux et al., 2013; Williamson- adventure tourism activities during the calving

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 60 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 season. General Wildlife Measures for those ar- ki, 2014). From 1984 to 1991, 52 caribou eas vary with respect to the proportion of area from the Itcha-Ilgachuz subpopulation were excluded from forest harvesting, and the levels transplanted to the unoccupied Charlotte Alp- and methods of forest harvesting in modified lands range (Young et al., 2001). That subpop- harvest areas (COSEWIC, 2014a). ulation appeared to remain stable until about The South Peace Northern Caribou Imple- 1999, but then declined (Youds et al., 2011). mentation Plan (BC Ministry of Environment, The only transplant of western mountain cari- 2013) provided for protection of ≥90% of bou over the past decade occurred in March identified high-elevation winter ranges across 2012, when 19 caribou were brought from the the Central Mountain and a portion of the Level-Kawdy subpopulation in the Northern Northern Mountain DUs. This includes the Mountain DU to the Purcells South and Pur- Graham, Moberly, Scott, Burnt Pine, and Nar- cells Central ranges in the Southern Mountain raway subpopulations in British Columbia. It DU. Seventeen died within 13 months due to also specifies protection of ≥80% of identified predation by wolves or cougars (n=8), acci- high-elevation winter ranges on the Quintette dents (n=3), malnutrition (n=1), or unknown range, but provides no indication of how the causes (n=5); the fate of the remaining two is protected portions of any of the range will be unknown due to GPS-collar malfunction (L. distributed geographically. In the Southern de Groot, BC Ministry of Forests, Lands and Mountain DU, caribou primarily use high-ele- Natural Resource Operations, pers. comm.). vation ranges, and recovery efforts have focussed Although wolf reduction and/or steriliza- on protecting most of those ranges from forest tion programs often enjoy initial success, as harvesting. However, forest harvesting has con- measured by enhanced caribou survival or re- tinued outside of those ranges in adjacent valley cruitment (e.g., Farnell & McDonald, 1988; bottoms, resulting in increased predation risk Bergerud & Elliott, 1998; Hegel & Russell, for caribou (Apps et al., 2013). Similarly, for 2010), the relatively rare opportunities for caribou in both the Central Mountain DU and longer-term monitoring have demonstrated the southern part of the Northern Mountain that such interventions, once ended, do not DU, continuing declines in caribou numbers always have sustained long-term benefits for is highly correlated to loss of high-quality habi- prey species affected by apparent competition tat and industrial disturbances at low elevations (Wittmer et al., 2013). Over the past decade, (Johnson et al., 2015). predator control efforts have continued, albeit Intensive management of caribou subpopu- constrained by social acceptability (Serrouya et lations including translocations, predator con- al., 2011). As part of the Mountain Caribou trol, prey control, and captive breeding and Recovery Implementation Plan in the South- rearing initiatives, have been deployed since the ern Mountain DU, trapping and hunting sea- mid-1980s (e.g., Compton et al., 1995; Young sons for wolves and cougars were adjusted in et al., 2001; Zittlau, 2004; Cichowski, 2014; 2007 to encourage removal of those predators COSEWIC, 2014a). Initial results can appear near caribou habitat (Mountain Caribou Re- promising but then often are not sustained. For covery Implementation Plan Progress Board, example, the Telkwa subpopulation in west- 2012). Until 2014, the only wolf removal or central British Columbia increased after the sterilization program In the Southern Moun- transplants of 32 caribou from 1997-1999 to at tain DU was on the Barkerville and Wells Gray least 144 total caribou in 2006 before declining (north) subpopulation ranges, where wolves to the current estimate of 19 animals (Cichows- were removed and sterilized leading to densi-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 61 ties of 3.2-3.4 wolves/1000 km2 across about lumbia Ministry of Forests, Lands and Natural 60% of the study area; the Barkerville caribou Resource Operations, pers. comm.). subpopulation increased and the Wells Gray Captive breeding has the strong endorse- (north) subpopulation remained stable, but ment from the Mountain Caribou Recovery calf recruitment remained variable (Roorda & Implementation Plan Progress Board as a means Wright, 2012). to quickly increase mountain caribou numbers In the Central Mountain DU, a 7-year wolf in some key core areas, and there is continued control effort targeting the Little Smokey range, interest by the BC government to augment a boreal caribou subpopulation (Hervieux et imperiled populations (C. Ritchie, BC Min- al., 2014), likely affected the A La Peche Cen- istry of Forests, Lands and Natural Resource tral Mountain caribou subpopulation as well Operations, pers. comm.). A captive-rearing because it shares the same winter range. In Jan- program was conducted for the Chisana sub- uary, 2015, the BC Ministry of Forests, Lands population in Yukon in the Northern Moun- and Natural Resource Operations announced tain DU during 2003-2006 (Chisana Cari- two targeted wolf removal efforts “to save caribou Recovery Team, 2010). In that program, bou herds under threat from wolf predation” in between 20 and 50 adult female caribou were the South Selkirk subpopulation range (South- captured annually in March and held in large ern Mountain DU) and the Quintette, Mober- enclosures (pens) until mid-June to increase ly, Scott and Kennedy-Siding) ranges (Central early calf survival. During the 4-year period, Mountain DU) (BC MFLNRO, 2015). A pro- calf survival until mid-June (time of release) vincial management plan for grey wolf released averaged 93% for captive-reared calves vs. 33% by the Government of BC in April 2014 (BC for calves born in the wild (Chisana Caribou MFLNRO, 2014:17), states that wolf con- Recovery Team, 2010). Survival of calves af- trol “to reduce predation risk on endangered ter release until mid-October was greater for caribou” has been a “provincial priority” since calves born in the pen (70%) than for calves 2001. Bag limits for wolf hunting have been born in the wild (52%). These results suggested removed in specified management units in an that captive rearing could be an effective tool effort to reduce predation on caribou. for small populations that are limited by poor Two moose population reductions have re- calf recruitment (Chisana Caribou Recovery cently been conducted in the Southern Moun- Team, 2010). Captive-rearing projects are cur- tain DU. Liberalized hunting resulted in a 71% rently being conducted (2014) for the Moberly reduction in moose numbers and about a 50% subpopulation in the Central Mountain DU reduction in wolf numbers on three ranges in (10 females captured), and for the Columbia the southern portion of the Southern Moun- North subpopulation (10 females captured) tain DU; the Columbia North population ex- in the Southern Mountain DU (S. McNay, perienced a modest increase while the two small Wildlife Infometrics Inc., pers. comm.; R. Ser- populations (Columbia South, Frisby-Boulder) rouya, Columbia Mountains Caribou Project, decreased regardless (Serrouya et al., 2011). In British Columbia, pers. comm.). In 2011, a the northern portion of the Southern Moun- partnership between Parks Canada, the British tain DU (Parsnip portion of the Hart Ranges), Columbia Government, and the Calgary Zoo moose numbers declined, possibly as a result of was created to implement a captive-breeding increased hunting, but over six years, neither program that would take breeding stock from wolf nor caribou numbers responded measur- British Columbia, and augment or reintroduce ably (Steenweg, 2011; D. Heard, British Co- animals in the four national parks and in BC

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 62 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 (Parks Canada, 2011b). No further details have Recovery Strategy for the Southern Mountain been publicly released since then but in late “nationally significant population” assessed by 2014, Calgary Zoo made a decision not to pro- COSEWIC (2002) was finalised at about the ceed (Ellis, 2014). same time as the most recent COSEWIC status The latest Alberta status report (ASRD & review (COSEWIC, 2014a). Although COSE- ACA, 2010) described various provincial re- WIC (2014a) brought forward changes to both covery planning efforts for both mountain and the DU structure and status of many subpopu- boreal ecotypes since 1986. Not until 2005 was lations (as presented above) that are well-aligned a recovery plan (Alberta Woodland Caribou with provincially-recognized ecotypes (Govern- Recovery Team, 2005) approved by the Alberta ment of British Columbia, 2014), experience government, although this was “qualified” in demonstrates that it may take some time be- that the recommendation for a moratorium on fore legal listing under SARA occurs and these the allocation of new resource extraction rights modifications are reflected in the SARA Reg- until range-specific management plans were in istry and subject to relevant regulations. The place was not accepted by the government of recently completed SARA recovery strategy Alberta (ASRD & ACA, 2010). No habitat has (Environment Canada, 2014) did, however, been protected on Alberta provincial lands in seek to clarify this confusing mismatch by ac- the Central Mountain DU for the purposes of knowledging COSEWIC’s new DU structure. caribou protection over the past decade; ongo- That strategy document also partially identified ing industrial development activities are man- critical habitat specific to the subpopulations aged through an inconsistently-applied patch- of the previously-defined (COSEWIC, 2002) work of caribou-related operating guidelines Southern Mountain population. focused on minimizing the size and duration of individual projects (ASRD & ACA, 2010). Conclusions and recommendations Oil leases continue to be sold within Alberta 2014 marked the third time COSEWIC has Central Mountain Caribou ranges, as recently reviewed the status of caribou in the western as March 2015 (Weber, 2015). Parks Canada mountains of Canada (in addition to 1984 has also produced a strategy (Parks Canada, and 2002). These status evaluations have docu- 2011a) to guide conservation efforts, which are mented profound range loss, pronounced and primarily focused on measures such as seasonal ongoing population declines, unsustainable closures of winter habitat, and management of predation rates, and continuing loss in area and elk populations, vehicle traffic control meas- connectivity of functional habitat, resulting in ures, and recreation in the four national parks small and isolated subpopulations in southern located in the Southern and Central Mountain and central British Columbia and Alberta. At DUs. Predator-prey relationships in these latter the same time, there are mounting concerns protected areas are heavily influenced by land for the welfare of subpopulations in northern use practices or human settlements character- British Columbia, Yukon and western North- ized by the surrounding landscapes. west Territories, which face escalating industrial Scientific assessments of Canadian wildlife development, even in currently remote regions by COSEWIC represent only the first stage (Hegel & Russell, 2013; COSEWIC, 2014a). in SARA listing and recovery processes. As- Increased understanding of the distribution, sessment is followed by the separate steps of ecology, and genetic variation of these western listing decisions and then recovery planning subpopulations has allowed COSEWIC to ap- and actions (Mooers et al., 2010). The SARA ply the Designatable Unit concept to this most

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 63 recent assessment (COSEWIC, 2011a). This exercise resulted in significant modifications to 1) Commit to reducing human intrusion into the boundaries of previously recognized North- caribou ranges, restoring habitat over the long ern and Southern Mountain “nationally signifi- term, and conducting short-term predator control cant populations” (COSEWIC, 2002). COSE- for small and/or declining subpopulations. WIC also introduced a third unit (Central All three components must be conducted si- Mountain), representing subpopulations on multaneously for successful recovery of west- the eastern flanks of the Rocky Mountains that ern mountain caribou. Implementation of the were previously considered Southern Moun- current recovery and management plans and tain caribou (Environment Canada, 2014). perhaps more drastic actions will undoubtedly Although it may be some time before they are result in trade-offs between the persistence of legally recognized, these boundary changes subpopulations of caribou, economic activity, have brought federal recovery units into better and societal expectations for conservation. If alignment with those recognized by provinces restraint of the human footprint is not consid- and territories, particularly BC. ered, then the prospects for preventing extirpa- In spite of considerable management at- tion of declining subpopulation through reli- tention to declining populations, available ance on mitigation of individual development high-quality monitoring data provide a clear projects will be increasingly limited. indication that recovery actions since the passage of SARA have been generally unsuccess- 2) For management and recovery actions, es- ful for caribou in the western mountains of pecially with respect to planned movements of Canada. In some areas, such as the Southern animals to supplement subpopulations, consider Mountain DU, large areas of important range carefully COSEWIC’s new DU structure for cari- have been protected from forest harvesting, but bou, which explicitly recognizes the evolutionary herds are still declining, with many reaching significance of discrete conservation units of the very low numbers. Recent actions focused on species in Canada. proximate causes of decline (e.g., predator con- Translocation efforts have involved, on occa- trol or moose reduction) may have helped to sion, a transfer of animals from one DU to stabilise some subpopulations (e.g., Columbia another, and are being increasingly adopted or North and Barkerville), but these efforts have considered. Increasingly a component of strate- not been accompanied by habitat recovery at gies aimed at maintaining or recovering small the scale necessary to enable overall popula- subpopulations of caribou in all three DUs tion recovery (Johnson et al., 2015). Alberta, (DeCesare et al., 2011; Environment Canada, in particular, has relied on mitigation measures 2014), have either met with failure, as meas- to ameliorate site-level impacts of new and past ured by death, lack of reproduction by intro- resource development projects. Containment duced individuals, or the results are difficult to of the human footprint across the range of disentangle from the effects of other recovery these mountain caribou DUs, however, is not measures applied simultaneously. Transloca- usually regarded as an option, in light of the tion projects can also serve to increase threats economic significance of resource development to caribou subpopulations through 1) the in- to provincial economies. troduction of novel genetic material that could Based on declines, future developments and cause outbreeding depression and reduce local current recovery effects, we offer the following adaptations, 2) removal of individuals from recommendations: source subpopulations that may in some cir-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 64 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 cumstances exacerbate extinction risk related to approach to limiting or mitigating land-use small source population sizes, 3) unanticipated changes and cumulative impacts that have disease transfer between environments that demonstrable negative impacts on caribou. characterize caribou ecotypes, or 4) low sur- Given the limited scope of SARA, recovery vival of individuals transplanted from one DU and management of western mountain caribou into another if the basis for DU designation is subpopulations will necessitate coordination local adaptations to the ecological setting. The within and between jurisdictions at appropri- ecological and behavioural characteristics that ate scales, including the effective protection of differentiate the three mountain caribou DUs critical habitat. The prevailing practice of piece- (COSEWIC, 2011a), make the prospects for meal project-by-project decision making does rescue unlikely through translocation from one not consider how development should proceed DU to another, particularly to the Southern at a regional scale and collectively engenders a Mountain DU. Experience suggests that the reactive approach. success of most translocations will be compromised if the causes of the original decline are Acknowledgements not addressed (St-Laurent & Dussault, 2012). Thanks to Troy Hegel and members of the or- ganizing committee of the 14th North Ameri- 3) Carry out regular surveys to monitor the con- can Caribou Workshop (Whitehorse, May dition of Northern Mountain caribou subpopu- 2014) for providing the impetus for this special lations and immediately implement preventative issue of Rangifer. Funding for the preparation measures on ranges that show signs of population of the COSEWIC status report (COSEWIC, declines or acceleration of threats. 2014a) was provided by Environment Canada. Although the designation of special concern We wish to acknowledge the other members of for the Northern Mountain population confers the COSEWIC Terrestrial Mammals Subcom- few obligations under SARA, the current con- mittee, Dan Benoit and the ATK Subcommit- servation status of the subpopulations in this tee, and numerous other reviewers who provid- DU illustrates well the importance of the third ed insightful comments and improvements to stated purpose of the Act “to manage species of the many drafts of the status report. Thanks to special concern to prevent them from becom- members of the full committee of COSEWIC ing endangered or threatened” (SARA, 2002). who conducted the assessment at the Spring In light of the worrisome signs already exhib- 2014 COSEWIC Species Assessment Meeting ited by southern subpopulations in this DU, in Halifax. Bonnie Fournier (Environment and intensifying natural resource development and Natural Resources, Government of the North- increasing natural disturbance in the region west Territories) produced the maps, and Jenny make it necessary to be vigilant and ready to Wu of the COSEWIC Secretariat provided in- respond. valuable assistance during the two-year process. We are grateful to the following individuals 4) Undertake a proactive, planned approach coor- who helped secure population and trend infor- dinated across jurisdictions to address the spatial mation in the status report: J. Adamczewski, N. extent and resource valuation essential to conserv- Baccante, N. Dodd, L. de Groot, A. Goddard, ing landscape processes. D. Heard, T. Hegel, D. Hervieux, N. Larter, Caribou conservation depends on the main- S. McNay, L. Neufeld, R. Pither, C. Ritchie, tenance of landscape-scale processes expressed J. Scheck, D. Seip, R. Serrouya, J. Surgenor, across extremely broad areas and a proactive C. Thiessen, G. Watts, and M. Williams. G.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 65 Kuzyk, T. Hegel, and an anonymous reviewer BC Ministry of Forests, Lands and Natural provided insightful comments that improved Resource Operations (BC MFLNRO). the manuscript. 2014. Management Plan for the Grey Wolf (Canis lupus) in British Columbia. BC Min- istry of Forests, Lands and Natural Resource Operations, Victoria, BC. 48 pp. http:// References www.env.gov.bc.ca/fw/wildlife/manage- Alberta Woodland Caribou Recovery Team. ment-issues/docs/grey_wolf_management_ 2005. Alberta woodland caribou recovery plan plan.pdf. 2004/05 – 2013/14. Alberta Sustainable Re- BC (MFLNRO). 2015. Government acting source Development, Fish and Wildlife Di- to save endangered caribou. News release. vision. Alberta Species at Risk Recovery Plan http://www.newsroom.gov.bc.ca/2015/01/ No. 4. Edmonton, AB. 48pp. government-acting-to-save-endangered-car- Apps, C., McLellan, B., Kinley, T., Serrouya, ibou.html. January 15, 2015. Accessed 16 R., Seip, D. & Wittmer, H. 2013. Spatial July 2015. factors related to mortality and population Bergerud, A.T. & Elliott, J.P. 1998. Wolf decline of endangered mountain caribou. predation in a multiple-ungulate system in — Journal of Wildlife Management 77: 1409- Northern British Columbia. — Canadian 1419. http://dx.doi.org/10.1002/jwmg.601 Journal of Zoology 76: 1551-1569. http:// ASRD & ACA (Alberta Sustainable Resource dx.doi.org/10.1139/z98-083 Development and Alberta Conservation Brown, W.K., Kansas, J.L. & Thomas, D.C. Association). 2010. Status of the Woodland 1994. The Greater Jasper Ecosystem Caribou Caribou (Rangifer tarandus caribou) in Al- Research Project, Seven chapters in final re- berta: Update 2010. Alberta Sustainable port prepared for Parks Canada, Calgary, Resource Development. Wildlife Status Re- and World Wildlife Fund, Toronto, by port No. 30 (Update 2010). Edmonton, AB. TAEM and Sentar Consultants, Calgary. AB. 88pp. 298 pp. Banfield, A.W.F. 1961. A revision of the rein- Chisana Caribou Herd Working Group. deer and caribou, genus Rangifer. — Na- 2012. Management plan for the Chisana cari- tional Museum of Canada, Bulletin No. 177. bou herd: 2010-2015. Government of Yu- Queen’s Printer, Ottawa, ON. 137pp. kon, Department of Environment, White- BC Hydro. 2015. Northwest Transmission horse, YT. 48pp. Line Project. Available at: https://www.bch- Chisana Caribou Recovery Team. 2010. Re- ydro.com/energy-in-bc/projects/ntl.html. covery of the Chisana Caribou Herd in the Accessed 16 July 2015. Alaska/Yukon Borderlands: Captive-Rearing BC Ministry of Environment. 2013. Imple- Trials. Yukon Fish and Wildlife Branch Re- mentation plan for the ongoing manage- port TR-10-02. ment of South Peace Northern Caribou Cichowski, D. 2014. Telkwa Caribou Popu- (Rangifer tarandus caribou pop.15) in British lation Status and Background Information Columbia. Victoria, BC. 16pp. Summary. Prepared for BC Ministry of For- BC Ministry of Environment. 2015. Moun- ests, Lands and Natural Resource Opera- tain Caribou Recovery. tions, Smithers, BC. 104pp. http://www.env.gov.bc.ca/wld/speciescon�- Compton, B.B., Zager, P. & Servheen, G. servation/mc/. Accessed 16 July 2015. 1995. Survival and mortality of translocated

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 66 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 woodland caribou. Wildlife Society Bulletin March 8-9, 2003, pp. 59-70. 23 (3): 490-496. DeCesare, N., Whittington, J., Hebblewhite, COSEWIC (Committee on the Status of M., Robinson, H., Bradley, M., Neufeld, L. Endangered Wildlife in Canada). 2002. & Musiani, M. 2011. The role of transloca- COSEWIC assessment and update status re- tion in recovery of woodland caribou popu- port on the Woodland caribou Rangifer taran- lations. — Conservation Biology 25: 365-373. dus caribou in Canada. Ottawa, ON. 98pp. Dueck, G.S. 1998. Genetic Relationships and COSEWIC. 2004. COSEWIC assessment and Phylogeography of Woodland and Barren- update Status Report on the Peary caribou ground Caribou. M.Sc. thesis, University of Rangifer tarandus pearyi and the barren- Alberta, Edmonton. 131pp. ground caribou Rangifer tarandus groen- Dyer, S.J., O’Neill, J.P., Wasel, S.M. & Bou- landicus (Dolphin and Union population) in tin, S. 2002. Quantifying barrier effects of Canada. Ottawa, ON. 91pp. roads and seismic lines on movements of COSEWIC. 2011a. Designatable Units for Car- female woodland caribou in northeastern ibou (Rangifer tarandus) in Canada. Com- Alberta. — Canadian Journal of Zoology 80: mittee on the Status of Endangered Wildlife 839–845. http://dx.doi.org/10.1139/z02- in Canada. Ottawa, ON. 88pp. http://www. 060 cosewic.gc.ca/eng/sct12/COSEWIC_Cari- Dzus, E. 2001. Status of the Woodland Cari- bou_DU_Report_23Dec2011.pdf. Ac- bou (Rangifer tarandus caribou) in Alberta. cessed 16 July 2015. Wildlife Status Report No. 30. Environment COSEWIC. 2011b. COSEWIC’s Assessment Fisheries and Wildlife Management Divi- Process and Criteria. http://www.cose- sion and Alberta Conservation Association. wic.gc.ca/pdf/Assessment_process_and_ 47pp. criteria_e.pdf. Accessed 16 July 2015. Ehlers, L.P.W., Johnson, C.J. & Seip, D.R. COSEWIC. 2011c. Guidelines for Recogniz- 2014. Movement ecology of wolves across an ing Designatable Units. http://www.cosewic. industrial landscape supporting threatened gc.ca/eng/sct2/sct2_5_e.cfm. Accessed 16 populations of woodland caribou. — Land- July 2015. scape Ecology 29: 451-465. http://dx.doi. COSEWIC. 2014a. COSEWIC assessment and org/10.1007/s10980-013-9976-8 status report on the Caribou Rangifer tarandus Ellis, C. 2014. Funding threatens caribou in Canada. Committee on the Status of En- program. Rocky Mountain Outlook, Nov. dangered Wildlife in Canada. Ottawa, ON. 6, 2014. http://www.rmoutlook.com/arti- xxii + 113pp. cle/20141106/RMO0801/310069967/-1/ COSEWIC. 2014b. About COSEWIC. http:// RMO08/funding-threatens-caribou-pro�- www.cosewic.gc.ca/eng/sct6/sct6_3_e. gram. Accessed 16 July 2015. cfm#hist. Accessed 16 July 2015. Environment Canada. 2012. Management Couturier, S., Hout, J., Côte, S.D., van Plan for the Northern Mountain Population Ginhoven, Q., Otto, R. & Jean, D. 2009. of Woodland Caribou (Rangifer tarandus cari- Populations, metapopulations, ecotypes and bou) in Canada. Species at Risk Act Man- subspecies of caribou in Quèbec agement Plan Series. Environment Canada, -Labrador: an exploratory discussion. — Ottawa, ON. vii + 79pp. In: McFarlane, K., Gunn, A. & Strobeck, C. Environment Canada. 2014. Recovery Strategy (Eds.). Proceedings from the Caribou Genetics for the Woodland Caribou, Southern Moun- and Relationships Workshop. Edmonton, AB, tain population (Rangifer tarandus caribou)

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 67 in Canada. Species at Risk Act Recovery Hebblewhite, M., White, C. & Musiani, M. Strategy Series. Environment Canada, Ot- 2010. Revisiting extinction in National tawa, ON. viii + 103pp. Parks: mountain caribou in Banff. — Con- Farnell, R., & McDonald, J. 1988. The demog- servation Biology 24: 341-344. http://dx.doi. raphy of Yukon’s Finlayson caribou herd 1982- org/10.1111/j.1523-1739.2009.01343.x 1987. Yukon Fish and Wildlife Branch Report Hegel, T.M. & Russell, K. 2010. Aishihik and TR-872. Whitehorse, Yukon. 54pp. Kluane northern mountain caribou herds cen- Festa-Bianchet, M., Ray, J.C., Boutin, S., sus, 2009. Fish and Wildlife Branch, Yukon Cote, S.D. & Gunn, A. 2011. Conserva- Environment. 14pp. tion of caribou (Rangifer tarandus) in Can- Hegel, T.M. & Russell, K. 2013. Status of ada: an uncertain future. — Canadian Jour- northern mountain caribou (Rangifer taran- nal of Zoology 89: 419-434. http://dx.doi. dus caribou) in Yukon, Canada. — Rangifer, org/10.1139/z11-025 Special Issue 13 (21): 59-70. http://dx.doi. Freeman, N. 2012. 2012 mountain caribou org/10.7557/2.33.2.2528 census of the Wells Gray North and Barkerville Hervieux, D., Hebblewhite, M., Stepnisky, sub-populations, Cariboo Region. Ministry of D., Bacon, M. & Boutin, S. 2014. Man- Environment, Williams Lake, BC. 11pp. aging wolves (Canis lupus) to recover threat- Geist, V. 2007. Defining subspecies, invalid ened woodland caribou (Rangifer tarandus taxonomic tools, and the fate of the wood- caribou) in Alberta. — Canadian Journal land caribou. — Rangifer Special Issue 17: 25- of Zoology 92: 1029-1037. http://dx.doi. 28. http://dx.doi.org/10.7557/2.27.4.315 org/10.1139/cjz-2014-0142 Government of British Columbia. 2014. Hervieux, D., Hebblewhite, M., DeCesare, Caribou Distribution in British Columbia by N.J., Russell, M., Smith, K., Robertson, S. Ecotype. http://www.env.gov.bc.ca/wld/spe- & Boutin, S. 2013. Widespread declines in ciesconservation/caribou_by_ecotype.html. woodland caribou (Rangifer tarandus cari- Accessed 16 July 2015. bou) continue in Alberta. — Canadian Jour- Government of Canada. 2014. Species at Risk nal of Zoology 91: 872-882. http://dx.doi. Public Registry. http://www.registrelep-sarar- org/10.1139/cjz-2013-0123 egistry.gc.ca. Johnson, C.J., Ehlers, L.P.W. & Seip, D.R. Gunn, A. 2009. Caribou morphology and 2015. Witnessing extinction – Cumulative taxonomy with emphasis on Canada’s Arc- impacts across landscapes and the future tic Islands. — In: McFarlane, K., Gunn, A. loss of an evolutionary significant unit of & Strobeck, C. (eds.). Proceedings from the woodland caribou in Canada. — Biological Caribou Genetics and Relationships Workshop. Conservation 186: 176-186. http://dx.doi. Edmonton, Alberta, March 8-9, 2003, pp. org/10.1016/j.biocon.2015.03.012 23-42. Kinley, T. 2007. 2007 Population survey for the Hatter, I. 2006. Mountain caribou 2006 survey Purcells-Central and Purcells-South subpopu- results, subpopulation trends and extinction lations of mountain caribou. Prepared for B.C. risk. Draft for Technical Review. BC Minis- Ministry of Environment, Nelson, B.C. and try of Environment. Victoria, BC. 19pp. Tembec Industries, Cranbrook, B.C. 11pp. Heard, D. and Vagt, K. 1998. Caribou in Brit- Mace, G.M., Collar, N.J., Gaston, K.J., Hil- ish Columbia: A 1996 status report. — ton-Taylor, C., Akçakaya, H.R., Leader- Rangifer, Special Issue No. 10: 117-123. Williams, N., Milner-Gulland, E.J. & http://dx.doi.org/10.7557/2.18.5.1548 Stuart, S.N. 2008. Quantification of ex-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 68 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 tinction risk: IUCN’s system for classifying — Ecography 28: 197-208. http://dx.doi. threatened species. — Conservation Biology org/10.1111/j.0906-7590.2005.04002.x 22: 1424-1442. http://dx.doi.org/10.1111/ Nitschke, C.R. 2008. The cumulative effects j.1523-1739.2008.01044.x of resource development on biodiversity McDevitt, A.D., Mariani, S., Hebblewhite, and ecological integrity in the Peace-Mober- M., DeCesare, N., Morgantini, L., Seip, ly region of Northeast British Columbia, D., Weckworth, B. & Musiani, M. 2009. Canada. — Biodiversity and Conservation Survival in the Rockies of an endangered 17:1715-1740. http://dx.doi.org/10.1007/ hybrid swarm from diverged caribou (Rangi- s10531-008-9376-6 fer tarandus) lineages. — Molecular Ecology Parks Canada. 2011a. Conservation Strategy 18: 665–679. http://dx.doi.org/10.1111/ for Southern Mountain Caribou in Canada’s j.1365-294X.2008.04050.x National Parks. Parks Canada, Ottawa, ON. McLellan, B., Serrouya, R. & Flaa, J. 2006. 29pp. Population censuses of caribou in the North Parks Canada. 2011b. Government of Can- Columbia Mountains. B.C. Ministry of For- ada Announces Woodland Caribou Cap- ests and Canadian Parks Service, Revelstoke, tive Breeding Partnering Arrangement be- British Columbia. 13pp. tween Parks Canada, BC Government and MCTAC (Mountain Caribou Technical Ad- Calgary Zoo. News Release November 25, visory Committee). 2002. A Strategy for the 2011. http://www.pc.gc.ca/APPS/CP-NR/ Recovery of Mountain Caribou in British Co- release_e.asp?id=1806&andor1=nr. Ac�- lumbia. B.C. Ministry of Water, Lands, and cessed 16 July 2015. Air Protection, Biodiversity Branch, Victoria, Rominger, E.M., Oldemeyer, J.L. & Rob- BC. 73 pp. http://www.env.gov.bc.ca/wld/ bins, C.T. 1991. Foraging dynamics and documents/mtcaribou_rcvrystrat02.pdf. Ac- woodland caribou: A winter management cessed 16 July 2015. conundrum. — Rangifer, Special Issue 7: Mooers, A.O., Doak, D.F., Findlay, S.C., 123. http://dx.doi.org/10.7557/2.11.4.1001 Green, D.M., Grouios, C., Manne, L.L., Roorda, L. & Wright, R. 2012. Caribou re- Rashvand, A., Rudd, M.A. & Whit- covery in the Quesnel Highlands: Predator ton, J. 2010. Science, policy, and spe- Management Progress Report. Prepared for cies at risk in Canada. — BioScience 60 the Ministry of Forests, Lands and Natural (11): 843-849. http://dx.doi.org/10.1525/ Resource Operation, Cariboo Region, Wil- bio.2010.60.10.11 liams Lake, BC. 23pp. Mountain Caribou Recovery Implementa- Species at Risk Act (SARA). 2002. Bill C-5, tion Plan Progress Board. 2012. Annual An act respecting the protection of wildlife Report on Activities and Accomplishments of species at risk in Canada. 25 August 2010. the Mountain Caribou Recovery 2011-12. http://laws.justice.gc.ca/PDF/Statute/S/S- 27pp. http://www.env.gov.bc.ca/wld/spe- 15.3.pdf. Accessed 16 July 2015. http://laws. ciesconservation/mc/files/2011_2012%20 justice.gc.ca/PDF/Statute/S/S-15.3.pdf. Ac- Annual%20Report.pdf. Accessed 15 July cessed 16 July 2015 2015. Seip, D. 1990. Ecology of woodland caribou Nielsen, S.E., Johnson, C.J., Heard, D.C. & in Wells Gray Provincial Park. BC Ministry Boyce, M.S. 2005. Can models of presence- of Environment Wildlife Bulletin No. B-68, absence be used to scale abundance? Two case Victoria, BC. 43pp. studies considering extremes in life history. Seip, D., Johnson, C. & Watts, G. 2007. Dis-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 69 placement of mountain caribou from winter Special Issue 20: 127-138. http://dx.doi. habitat by snowmobiles. — Journal of Wild- org/10.7557/2.32.2.2261 life Management 71 (5): 1539-1544. http:// Terry, E., McLellan, B., & Watts, G. 2000. dx.doi.org/10.2193/2006-387 Winter habitat ecology of mountain caribou Seip, D. & Jones, E. 2014. Population status of in relation to forest management. — Journal caribou herds in the Central Mountain Desig- of Applied Ecology 37: 589-602. http://dx.doi. natable Unit within British Columbia, 2014. org/10.1046/j.1365-2664.2000.00523.x BC Ministry of Environment, Prince George, Thiessen, C. 2009. 2009 Stone’s Sheep/Caribou BC. 33pp. inventory – MU 7-52 (revised). BC Ministry Serrouya, R., McLellan, B.N., Boutin, S., Seip, of Environment, Fort St. John, BC. 22pp. D.R. & Nielsen, S.E. 2011. Developing a van Oort, H., McLellan, B.N. & Serrouya, R. population target for an overabundant un- 2010. Fragmentation, dispersal and metap- gulate for ecosystem restoration. — Journal opulation function in remnant populations of Applied Ecology 48: 935-942. http://dx.doi. of endangered mountain caribou. — Animal org/10.1111/j.1365-2664.2011.01998.x Conservation 14: 215–224. http://dx.doi. Serrouya, R., Paetkau, D., McLellan, B.N., org/10.1111/j.1469-1795.2010.00423.x Boutin, S., Campbell, M. & Jenkins, Wakkinen, W. 2003. 2003 Selkirk ecosystem D. 2012. Population size and major val- woodland caribou census. Idaho Fish and leys explain microsatellite variation better Game unpublished report. 2pp. http://www. than taxonomic units for caribou in western env.gov.bc.ca/wildlife/wsi/reports/4332_ Canada. — Molecular Ecology 21: 2588- WSI_4332_RPT1.PDF. Accessed 16 July 2601. http://dx.doi.org/10.1111/j.1365- 2015. 294X.2012.05570.x Weber, B. 2015. Alberta plans huge energy Spalding, D.J. 2000. The early history of wood- lease sale on range used by caribou. Cana- land caribou (Rangifer tarandus caribou) in dian Press, March 5, 2015. Available from: British Columbia. B.C. Ministry of the Envi- http://www.huffingtonpost.ca/2015/03/05/ ronment, Lands and Parks, Wildlife Branch, alberta-plans-huge-energy_n_6809908. Victoria. — Wildlife Bulletin No. 100. 61pp. html. Accessed 16 July 2015. Steenweg, R. 2011. Interactions of Wolves, Weckworth, B.V., Musiani, M., McDevitt, A., Mountain Caribou and an Increased Moose- Hebblewhite, M. & Mariani, S. 2012. Re- Hunting Quota - Primary-Prey Management construction of caribou evolutionary history as an Approach to Caribou Recovery. MSc. in western North America and its implica- Thesis. University of Northern British Co- tions for conservation. — Molecular Ecology lumbia, Prince George, British Columbia, 21: 3610-3624. http://dx.doi.org/10.1111/ Canada. 144pp. j.1365-294X.2012.05621.x Stevenson, S. & Hatler, D. 1985. Woodland Williamson-Ehlers, E.P. 2012. Impacts of in- caribou and their habitat in southern and dustrial developments on the distribution and central British Columbia. Vol. 1. B.C. Min- movement ecology of wolves (Canis lupus) and istry of Forestry, Victoria, BC. Land Man- woodland caribou (Rangifer tarandus caribou) agement Report No. 23. 354pp. in the South Peace region of British Columbia. St-Laurent, M.H. & Dussault, C. 2012. The MSc Thesis. University of Northern Brit- reintroduction of boreal caribou as a con- ish Columbia, Prince George, BC. Canada. servation strategy: A long-term assessment 163pp. at the southern range limit. — Rangifer Wittmer, H., McLellan, B., Serrouya, R. &

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 70 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Apps, C. 2007. Changes in landscape Committee. Williams Lake, BC. 97pp. composition influence the decline of a Young, J. & Freeman, N. 2001. Summary of threatened woodland caribou population. mountain caribou surveys within the Quesnel — Journal of Applied Ecology 76:568- Highland and Cariboo Mountains, Cariboo 579. http://dx.doi.org/10.1111/j.1365- Region, up to and including 2001. BC Minis- 2656.2007.01220.x try of Water, Land and Air Protection, Wil- Wittmer, H., McLellan, B.N., Seip, D., liams Lake, BC. 33pp. plus appendices. Young, J., Kinley, T., Watts, G. & Hamil- Young, J. & Roorda, L. 1999. Towards inte- ton, D. 2005. Population dynamics of the grated management solutions: the Itcha-Ilga- endangered mountain ecotype of woodland chuz caribou project radio-telemetry year four caribou (Rangifer tarandus caribou) in Brit- progress report 1995-99. BC Ministry of En- ish Columbia, Canada. — Canadian Jour- vironment and Lands, Cariboo Region, Wil- nal of Zoology 83: 407-418. http://dx.doi. liams Lake, BC. 51pp. org/10.1139/z05-034 Zittlau, K.A. 2004. Population Genetic Analyses Wittmer, H.U., Ahrens, R.N.M. & McLellan, of North American Caribou (Rangifer taran- B.N. 2010. Viability of mountain caribou in dus). PhD Thesis, University of Alberta, British Columbia, Canada: Effects of habitat Canada. 374pp. change and population density. — Biologi- cal Conservation 143: 86-93. http://dx.doi. org/10.1016/j.biocon.2009.09.007 Wittmer, H.U., Serrouya, R., Elbroch, L.M. & Marshall, A.J. 2013. Conservation strategies for species affected by apparent competition. — Conservation Biology 27: 254-260. http://dx.doi.org/10.1111/cobi.12005 Yannic, G., Pellissier, L., Ortego, J., Lecomte, N., Couturier, S., Cuyler, C., Dussault, C., Hundertmark, K.J., Irvine, R.J., Jenkins, D.A., Kolpashikov, L., Mager, K., Musiani, M., Parker, K.L., Roed, K.H., Sipko, T., Porisson, S.G., Weckworth, B.V., Guisan, A., Bernatchez, L. & Cote, S.D. 2014. Ge- netic diversity in caribou linked to past and future climate change. — Nature Climate Change 4 (2): 131-137. Youds, J., Armleder, H., Young, J., Freeman, N., Roorda, L., Pelchat, M., Ranking, J., Bjorkman, C., Waterhouse, M., Wright, R., McLeod, J. & Peel, D. 2011. Cariboo- Chilcotin Land Use Plan Northern Caribou Strategy Review: Update #1. Prepared for the Cariboo Regional management Team, Min- istry of Forests, Lands and Natural Resource Operations, and the Cariboo Manager’s

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 71 0 2 2 4 6 0 11 20 22 54 12 78 45 157 341 202 398 1354 Mature Estimates 0 2 3 4 7 0 Population 22 23 64 13 13 90 47 183 392 222 459 1544 Total 2014 Estimate 4 95 -62 -65 -72 -91 -93 -50 -69 -94 -36 -88 -43 -28 -38 -35 -46 -100 -100 Long-term % change 0 2 2 4 6 0 11 20 22 54 12 76 45 157 298 202 381 1292 Mature 2 0 2 3 4 7 0 22 23 64 13 13 88 47 183 343 222 439 1473 Total 6 2014 Estimate Year 2011 2011 2014 2014 2005 2012 2008 2013 2013 2013 2013 2013 2012 2014 2004 2013 2014 Survey Estimates 7 3 8 53 63 19 28 39 89 39 73 22 23 192 100 247 522 280 590 2387 Mature 63 69 22 31 30 12 43 46 81 24 211 114 280 109 631 299 716 2781 Total 5 8 5 Initial Estimate Year 199 1994 1996 1999 1995 1994 1994 1994 1997 1990 1995 1988 1999 1999 2006 1995 1993 8 5 9 8 4 5 6 8 6 8 3 6 4 7 4 11 11 2014 # full 2002- surveys 9 4 6 3 3 2 1 3 3 3 5 0 2 2 2 0 14 2001 # full 1987- surveys 1 SM SM SM SM SM SM SM SM SM SM SM SM SM SM SM SM SM 2002 NSP COSEWIC Assessment Subpopulation South Selkirks (BC) Purcells South (BC) Purcells Central (BC) Nakusp (BC) Duncan (BC) Central Rockies (BC) Monashee (BC) Frisby Boulder (BC) Columbia South (BC) Columbia North (BC) Groundhog (BC) Gray (BC) Wells Barkerville (BC) North Cariboo Mountains (BC) Narrow Lake (BC) George Mountain (BC) Hart Ranges (BC) Total Table 1. Subpopulation and trend estimates for 15 caribou subpopulations in the Southern Mountain Designatable Unit. Designatable Mountain estimates for 15 caribou subpopulations in the Southern and trend 1. Subpopulation Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 72 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 - - -

Nationally significant population (see full explanation in manuscript text). Nationally Censuses of DU9 caribou are conducted using standardized methods and searching predetermined survey areas. Various techniques have been used to estimate num techniques have Various survey predetermined areas. methods and searching conducted using standardized Censuses of DU9 caribou are This survey/estimatereliable surveyyears). is the oldest conducted with the highest count of animals during last 3 generations (27 the initial and 2014 survey between estimates. Long-term % change is based on the difference This survey incomplete but it had the highest number of caribou counted in that time period. was considered the two subpopulations during surveyare between so all caribou counted Duncan subpopulation but it was not possible to differentiate This survey includes the individuals was calculated based on the proportion of adults in all surveys so the number of mature for the subpopula for this year available composition data were No individuals was calculated based on the was 2002 so the number of mature composition was available the only year for this year; available composition data were No Footnotes to Table 1. Table to Footnotes 1 2 the present, caribou are surveys no radio-collared where ber of caribou in the survey (survey estimate) and for the whole population (population estimate). For area area sometimes found outside the survey caribou are available, caribou are surveys radio-collared where survey to the population estimate. For estimate is usually equivalent for assessing the population estimate is presented this table, the most recent incorporated into the population estimate but not survey these are estimate. In area; the estimate was calcu surveys a survey where estimate was not provided, of caribou in DU9, but % change is based on survey population size estimates. For current (1999) to total caribou & Roorda Young (1990) and factor of 0.83 used for DU9 caribou surveys sightability correction based on Seip applying a standardized lated by seen plus tracks. 3 4 5 6 subpopulation. included in the Nakusp 7 tion that included composition data. 8 Lake subpopulation, which included composition data. proportion of adults in all surveysaverage Narrow for the nearby

Initial Estimate Year 2007 1995 2007 1996 2008 2008 1999 1999 1989 1986 0 4 7 5 3 0 0 0 6 12 2014 # full 2002- surveys

14 0 3 0 4 0 0 0 0 6 0 2001 # full 1987- surveys 1 SM SM SM SM SM SM SM SM SM SM NSP Assessment 2002 COSEWIC

12 11 (BC/ 9 (AB/BC) 10 (BC) 4 Subpopulation Scott Moberly (BC) Kennedy Siding (BC) Burnt Pine (BC) Quintette (BC) Narraway AB) Redrock/Prairie Creek La Peche A (AB) Jasper (AB) Banff (AB) Total Nationally significant population (see full explanation in manuscript text). Nationally This survey/estimatereliable surveyyears). is the oldest conducted with the highest count of animals during last 3 generations (27 the initial and 2014 survey between estimates. Long-term % change is based on the difference portion survey side of the Scott range; estimate for western of the range was based on anecdotal information and expert has been conducted for the west No 25-35 (with a mid winter range during the fall, minimum population is 22 with an estimated ranging from on observation on the low-elevation Based proportion of adults in all surveys individuals was calculated based on the average for the so the number of mature for this year available composition data were No Table 2. Subpopulation and trend estimates for 13 caribou subpopulations in the Central Mountain Designatable Unit. Designatable estimates for 13 caribou subpopulations in the Central Mountain and trend 2. Subpopulation Table 1 2 3 4 The number population trend. estimate for the purpose of assessing overall opinion and has not been updated since 2007, so the same estimate was used for current applying the proportion of adults during the two surveys by side of the Scott range was derived in the east side of Scott range to individuals for the west of mature side estimate. The east of the Scott range was surveyed in 2007 and 2014. the west 5 by applying the proportion of adults seen (21/22) to the total population estimate 30. (2014)). The number of adults was estimated & Jones Seip point of 30 from 6 subpopulation that included composition data. nearest

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 74 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 - =0.904; λ =0.983; 2011: λ (1999-2012) = 0.71. The population was estimated to be λ (1999-2012) = 0.27. λ 2010). =0.835). Cumulative =0.835). Cumulative λ =0.749). Cumulative =0.749). Cumulative λ et al. =0.880; 2012: λ =0.921; 2012: 1994). λ et al. 2010. Data based on annual maximum counts and telemetry 2010. Data individuals was studies. The number of mature =0.836; 2011: et al. λ =0.866; 2011: λ (2008-2012) = 0.55. λ =0.811). Cumulative =0.811). Cumulative λ Although no structured surveys were conducted prior to 2004 (J. Whittington, Parks Canada, pers. comm. 2015), the 1986 estimate is based on a minimum count conducted prior to 2004 (J. surveysAlthough no structured were Whittington, Parks Population estimates based on 2009 population estimate of 212 caribou and mature estimate of 178 caribou (ASRD&ACA 2010) and then extrapolated back to estimate of 178 caribou (ASRD&ACA estimates based on 2009 population estimate of 212 caribou and mature Population 2010) and then extrapolated back estimate of 116 caribou (ASRD&ACA estimates based on 2009 population estimate of 135 caribou and mature Population combined to assess long-term trend. were but Jasper, 3 subpopulations in are There 1 in Hebblewhite Figure from interpreted Data Although 1 bull caribou was seen during the March 2014 survey, Seip & Jones (2014) concluded that the population was functionally extirpated. & Jones Seip 2014 survey, Although 1 bull caribou was seen during the March by applying the % 2014). The number of adults was estimated & Jones population estimated at 98-113 (midpoint=106) (Seip Total a partial from survey. Extrapolated portion, based on 2009 population estimate of 100 2014) and for the remaining & Jones portion (Seip Estimate based on minimum count for Bearhole/Redwillow Footnotes to Table 2 continued. Table to Footnotes 7 8 of adults (82%) in the partial survey to the total estimate. 9 (2010) and extrapolated ASRD&ACA 2010) and then extrapolated back to 2008 using annual lambdas from estimate of 90 caribou (ASRD&ACA caribou and mature (unpublished data; 2010: Development Resource Alberta and Sustainable to 2012 using annual lambdas from Ministry of Environment 2012: 10 Re Alberta and Sustainable (2010) and extrapolated to 2012 using annual lambdas from Ministry of Environment ASRD&ACA 1999 using annual lambdas from (unpublished data; 2010: Development source 11 Alberta and Sustainable (2010) and extrapolated to 2012 using annual lambdas from Ministry of Environment ASRD&ACA to 1999 using annual lambdas from (unpublished data; 2010: Development Resource 1988 (Brown 175-200 based on a count of 162 in October 12 13 Canada unpublished data) to the total number. (Parks applying the proportion of adults in caribou sightings that year by derived 14 annual maximum counts and/or telemetry obtained from studies (Hebblewhite

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 75 5 5 801 413 388 631 163 1853 1886 1813 1065 4200 7300- 10000 Mature 900 450 450 701 181 Estimates Population Population 1180 2200 5000 2105 2044 2014 Estimate Total >10000 7 4 2 30 10 -20 Long- Long- term % change 5 5 801 413 388 631 163 1065 1853 1886 1813 4200 7300- 10000 Mature 900 450 450 701 181 1180 2200 5000 2105 2044 Total >10000 2014 Estimate 2006 2001 1982 2012 2009 2008 1993 2013 2009 2009 2012 Year Survey Estimates 607 204 1337 1399 Mature 720 235 1432 1500 Total 10 Initial Estimate ------Year 2001 2003 1981 2003 1 0 0 0 1 1 0 9 2 1 2014 # full 2002- surveys

3 8 11 1 0 0 1 1 0 1 4 1 1 2 2004 # full 1987- surveys 1 NM NM NM NM NM NM NM NM NM NM NM 2002 NSP COSEWIC Assessment 4 6 9 Subpopulation Territories Yukon/Northwest Northern Hart River Clear Creek Bonnet Plume Redstone South Nahanni Coal River La Biche Yukon Southwest Chisana Kluane Aishihik Klaza Table 3. Subpopulation and trend estimates for 45 caribou subpopulations in the Northern Mountain Designatable Unit. Designatable Mountain estimates for 45 caribou subpopulations in the Northern and trend 3. Subpopulation Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 76 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 289 270 415 876 176 748 674 2907 2657 1240 514-857 Mature 316 300 521 200 850 775 Estimates 600- Population 3758 1000 3077 1491 1000 2014 Estimate Total 2 12 13 -2 - 116 127 13 Long- term % change 289 270 415 876 176 748 674 2907 2657 1240 Mature 514-857 316 300 521 200 850 775 600- 3758 1000 3077 1491 1000 Total 2014 Estimate 1993 1991 1991 2000 2002 2007 1998 2003 2008 2007 2007 Year Survey Estimates 329 312 679 2350 Mature 424 403 809 3067 Total Initial Estimate ------Year 1986 1998 1997 1999 0 0 0 0 1 1 0 1 2 2 1 2014 # full 2002- surveys 1 1 1 1 0 4 3 0 1 1 1 2004 # full 1987- surveys 1 NM NM NM NM NM NM NM NM NM NM NM 2002 NSP COSEWIC Assessment 14 14 Subpopulation Central Yukon Ethel Lake Moose Lake River Tay Tatchun Pelly Herds Finlayson Lake Wolf Yukon Southern Lakes Laberge Ibex Carcross Atlin Table 3 continued. Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 77 19 140 140 838 220 199 672- 1145 1342 1239 2258 1095 85-340 515-686 680-850 Mature 26 151 151 265 237 Estimates 800- 800- Population 1600 1000 1538 2861 1300 1000 1275 2014 Estimate Total 600-800 100-400 9 2 15 -89 Long- term % change 19 611 140 140 954 220 199 1145 1239 2258 Mature 515-686 680-850 26 800 151 151 738 265 237 600- 800- 1133 1000 1538 2861 1275 Total 2014 Estimate 18 18 2011 2007 1999 2000 1998 2006 2008 1994 2007 2007 2002 1993 Year 2000 2001 Survey Estimates 122 559 170 Mature 141 658 193 Total Initial Estimate ------Year 2005 2001 1994 1 0 0 0 1 0 0 3 1 1 0 0 1 0 2014 # full 2002- surveys 15 17 0 1 1 0 0 1 0 1 1 1 1 1 2 2 2004 # full 1987- surveys 1 NM NM NM NM NM NM NM NM NM NM NM NM NM NM 2002 NSP COSEWIC Assessment 14 - 14 16 14 14 Subpopulation Northwest BC Swan Lake Little Ranche ria Horseranch Level Kawdy Edziza Tsenaglode Spatsizi Northeast BC Liard Plateau Rabbit Muskwa Gataga Frog Finlay Pink Mountain Table 3 continued. Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 78 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 98 19 43 19 637 404 298 248 1220 Mature 25 50 23 708 475 341 122 300 Estimates Population 1685 2014 Estimate Total 2 19 20 21 22 -9 - - -64 -53 -50 34 -47 Long- term % change 98 12 43 19 637 404 298 248 905 Mature 16 50 23 708 475 341 122 300 1350 Total 2014 Estimate 2009 2009 2010 2004 2013 2002 2014 2008 2001 Year Survey Estimates 33 92 38 324 433 675 Mature 40 53 361 471 933 103 Total

Initial Estimate - - - Year 1996 1987 1987 1987 1987 1993 2 4 6 1 1 7 1 0 13 2014 # full 2002- surveys 1 1 1 1 2 7 6 10 14 2004 # full 1987- surveys 1 SM SM SM SM SM SM SM SM SM 2002 NSP COSEWIC Assessment Subpopulation North-central BC Graham Chase Wolverine Takla BC West-central Telkwa Tweedsmuir Itcha-Ilgachuz Rainbows Charlotte Alplands Although a survey survey was conducted in 1997, the population estimate not comparable to most recent because the survey was not exactly the same. area Nationally significant population (see full explanation in manuscript text). Nationally the initial and 2014 survey between estimates. Long-term % change is based on the difference Although a survey survey was conducted in 1978, the population estimate not comparable to most recent because the survey was not exactly the same. area Estimate based on expert opinion. hunter observations % adults from 1991 to 2010 (Larter from by applying the average 2012). individuals was derived The number of mature was not included as a conducted survey. This estimate is based on an opportunistic survey ground and therefore comparable to the 2001 survey individuals in 2009 for an area area. individuals in 2001 and 1465 mature based on 1337 mature % change in population size not prior to 2003 are 2012). Data Group Working 1989 and 2004 (Chisana 1987 to 2011, except composition surveys been conducted annually from have Fall 1 2 3 4 5 6 7 8 9 based on formal estimates of the 2003, population estimates are based on an interpolation of composition data. Since because population estimates were considered estimates. comparable to pre-2003 not directly and are size subpopulation’s Table 3 continued. Table

of 0.947 for 12 years where data were available during the period 1985/86 to 2008/09 (Cichowski & MacLean 2005, & MacLean during the period 1985/86 to 2008/09 (Cichowski available data were where of 0.947 for 12 years

Although the change from 1987 to 2012 was a net increase of 34%, the population increased from 675 mature individuals in 1987 to 2161 2003, and then de 675 mature from of 34%, the population increased 1987 to 2012 was a net increase Although the change from The 2003 survey was a fall composition survey so it was not a formal population estimate. not survey not comparable to the most recent because the survey were areas conducted in 1989 and 2000, the population estimates were Although surveys were individuals in 1986 to 4474 1990, and then 2350 mature from of 13%, the population increased 1986 to 2007 was a net increase Although the change from population estimate so the 3 generation % change was not calculated. recent No Yukon/BC border Plateau subpopulations straddle the Horseranch and Liard Swan Lake, Little Rancheria, Atlin, The ranges of the Carcross, not exactly the same. not comparable because the survey were areas conducted in 1988 and 1999, the population estimate were Although surveys were by assuming 85% adults. individuals was derived The population estimate is based on expert opinion; the number of mature not survey not comparable to the most recent because the survey were areas conducted in 1996 and 2000, the population estimates were Although surveys were a minimum number available to determine whether this was a full count or is insufficient information partial count so this estimate should be considered There surveys. between % change not possible to assess since the total survey varied Absolute area survey not comparable to the most recent because the survey area/ conducted in 1993, 2002, 2007 and 2008, the population estimates were Although surveys were The decline is supportedby an average 10 11 exactly the same. 12 its peak in 1990 to 2007). 2007 (a 41% decline from to 2657 by decreased 13 14 15 16 17 exactly the same. 18 until a full survey when a full surveypresent can be conducted, and should not used to assess trend is conducted. 19 20 not exactly the same. survey method were 21 2010). Cichowski 22 its peak in 2003 to 2014). 2014 (a 58% decline from to 905 by creased Footnotes to Table 3 continued. Table to Footnotes

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 80 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

Peary caribou distribution within the Bathurst Island Complex relative to the boundary proposed for Qausuittuq National Park, Nunavut

Kim G. Poole1, Anne Gunn2, Jack Wierzchowski3 & Morgan Anderson4

1 Aurora Wildlife Research, 1918 Shannon Point Rd., Nelson, British Columbia, V1L 6K1, Canada (Corresponding author: [email protected]). 2 368 Roland Road, Salt Spring Island, British Columbia, V8K 1V1, Canada. 3 Geomar Consulting Inc., Stuben, Austria. 4 Department of Environment, Government of Nunavut, P.O. Box 209, Igloolik, Nunavut, X0A 0L0, Canada.

Abstract: How caribou (Rangifer tarandus), including Peary caribou (R. t. pearyi), use their annual ranges varies with changes in abundance. While fidelity to some seasonal ranges is persistent, use of other areas changes. Consequently, understanding changes in seasonal distribution is useful for designing boundaries of protected areas for caribou conservation. A case in point is the proposed Qausuittuq (Northern Bathurst Island) National Park for Bathurst Island and its satellite islands in the High Arctic of Canada. Since 1961, Peary caribou have been through three periods of high and low abundance. We examined caribou distribution and composition mapped during nine systematic aerial surveys (1961–2013), unsystematic helicopter surveys (1989–98), and limited radio-collaring from 1994–97 and 2003–06. While migration patterns changed and use of southern Bathurst Island decreased during lows in abundance, use of satellite islands, especially Cameron Island for winter range, persisted during both highs and lows in abundance. The northeast coast of Bathurst Island was used to a greater extent during the rut and during summer at low abundance. We suggest that Park boundaries which include Cameron Island and the northeast coast of Bathurst Island will be more effective in contributing to the persistence of Peary caribou on the Bathurst Island Complex.

Key words: boundaries; distribution; Peary caribou; protected area; Rangifer tarandus pearyi.

Rangifer, 35, Special Issue No. 23, 2015: 81-98 DOI 10.7557/2.35.2.3635

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 81 Introduction areas and seasons, but a key habitat require- Peary caribou Rangifer tarandus pearyi regularly ment is terrain and vegetation features that of- occur only on Canada’s Arctic islands (Fig. 1) fer choices as caribou adjust their foraging to and have been nationally recognized as Endan- changing snow conditions. Upland habitats gered since 1999 based on declines and fluctua- with shallow snow cover are selected during tions in sub-populations (COSEWIC 2004). winter in many areas. Calving areas generally On Bathurst Island and its satellite islands, provide snow-free or shallow snow-covered Peary caribou abundance has been through sites. Habitat requirements during the snow- three periods of high (early 1960s, early 1990s free season relate to maximizing protein intake and 2013) and low abundance (early to mid- and the most nutritious forage. Specific fall and 1970s and mid- to late 1990s) (Miller & Barry, rutting areas occur, but these habitat require- 2009:Table 1; Jenkins et al., 2011; Anderson, ments are poorly known. Annual migrations 2014). The die-offs in the early 1970s and mid- and range sizes vary among years, but can range 1990s coincided with winters characterised up to 500 km and 4,000 km2, respectively by early and unusually high snowfall, freezing (COSEWIC, 2004; Jenkins et al., 2011). rain, and warmer temperatures (Miller & Bar- Seasonal distribution and migration strate- ry, 2009). gies vary with changes in abundance and over Parks Canada selected the proposed Qau- the longer-term, with climate (Species at Risk, suittuq National Park on northern Bathurst 2012). In 2012 Parks Canada requested we as- Island and its satellite islands (Fig. 1) in 1996 sess Peary caribou distribution and movements to be representative of the Western High Arc- relative to the proposed Qausuittuq National tic Natural Region (Parks Canada, 2012). Na- Park boundaries (Gunn et al., 2012). Updated tional parks have a goal to maintain ecological with data from 2013 (Anderson, 2014), we ex- integrity over the long term, which requires amined the ability of the proposed boundaries that parks encompass the habitat and connec- to protect caribou throughout their population tivity needs for viable wildlife populations. To cycles – especially during the more important maintain ecological integrity, national parks calving, post-calving, rutting and winter sea- need to accommodate the natural range of vari- sons – and how information on distribution ability (Landres et al., 1999) which, for Peary can contribute to decisions about the proposed caribou, is marked by pronounced changes in Park boundary. We assumed that if a seasonal numbers driven by abundance of forage or spo- range was used by a substantial or dispropor- radic, unpredictable abiotic variables (Species tionate proportion of the population during at Risk, 2012). Those changes in abundance any period of cyclic high or low abundance, are reflected in shifts in distribution and migra- then to be effective a national park should en- tory strategies at least on Bathurst and its sat- compass that range. Our objectives were to 1) ellite islands. Thus, for migratory species such determine the relative distribution of Peary car- as caribou and their predators, park boundaries ibou based on aerial surveys for Bathurst Island need to capture sufficient migratory corridors and the Governor General Islands, 2) deter- within natural ranges of variability. Elsewhere mine the distribution of Peary caribou relative in North America, the role of national parks for to the boundaries of the proposed Qausuittuq migratory mammals is well-recognized (Berger, National Park, 3) determine at the individual 2004). caribou scale, seasonal ranges and movements Habitat requirements of Peary caribou (sum- relative to the proposed boundaries of the Park, marized in Species at Risk, 2012) vary among and 4) summarize the adequacy and effective-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 82 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Figure 1. Study area for Bathurst Island, satellite islands and water bodies, NU.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 83 ness of the proposed Park boundary relative to sessment Committee, which rated high poten- known Peary caribou distribution and habitat tial for lead zinc mineralization on the north- requirements (COSEWIC, 2004). east coast of Bathurst Island, and petroleum potential on southwest Cameron Island (Parks Methods Canada, 2012). Southern Cameron Island is Study area the southern extension of the Sverdrup Basin Bathurst Island is a relatively large island which has petroleum potential, where Bent (16 080 km2) cut by deep inlets and bays Horn was a single producing well abandoned into several large peninsulas; 77% of the land due to falling pressure in 1996. The highly frac- mass is within 10 km of the coast line. Miller tured field limited production but was included (1998) described the ‘Bathurst Island Complex’ in a 2012–13 call for exploration bids in antici- (28 000 km2) as the approximately 30 islands pation that new techniques might be applicable clustered around and including Bathurst Is- to further development (AANDC, 2012). land. In addition to Bathurst Island, we examined Peary caribou use of the five large islands Data lying along the northwest coast (Governor Most data on caribou were collected before General group of islands: Cameron 1,059 km2, the lands were withdrawn and, therefore, stud- Vanier 1,126 km2, Massey 432 km2, Marc ies were not specifically designed to examine 56 km2, Alexander 484 km2) and island group- boundaries. The historic data had variable spa- ings off the north coast (Helena 220 km2 and tial and temporal resolution, and most were not the surrounding six small islands; Fig. 1). available digitally. The vegetation in this area is mostly High Arctic semi-desert (Gould et al., 2003; Miller 1. Aerial surveys of Bathurst and satellite islands & Barry, 2009) with a sparse to moderate cover Nine complete systematic aerial strip transect of cushion forbs, prostrate dwarf shrubs, sedges surveys were conducted between 1961 and and grasses. The climate is a short plant grow- 2013 (Tener, 1963; Miler et al., 1977; Miller, ing season marked by variability in the dates 1989; Ferguson, 1991; Gunn & Dragon, 2002; in June and August when green-up starts and Jenkins et al., 2011; Anderson, 2014). Data ends, respectively. from systematic surveys –mostly text descrip- tions and the raw data (observation sheets, Proposed Qausuittuq National Park boundaries maps) – were unavailable for all surveys except The current 2002 federal boundary proposal 1988, 1997, 1998 and 2013. S. Barry and F.L. covers Bathurst Island north of the Polar Bear Miller (Canadian Wildlife Service (CWS), re- Pass National Wildlife Area (PBPNWA) except tired) provided scanned images of the original for the northeast coast, and includes all the maps for Tener (1963), which included tran- Governor General Islands except Cameron Is- sects, caribou group locations and numbers. land, and Helena Island and surrounding small The data in the published reports allowed us to islands (except Seymour Island) (Fig. 1; Parks describe the proportional distribution of Peary Canada, 2012). Polar Bear Pass National Wild- caribou by satellite island and the western and life Area is an east-west oriented wetland desig- eastern halves of northern Bathurst Island. To nated in 1990 (surface and subsurface rights to further examine the distribution relative to the exploration and development are withdrawn). proposed Park boundaries, we screen digitized The Park proposal reflects the recommendations report figures. We obtained digital locations for of the Senior Mineral Energy & Resource As- 1997 and 1998 caribou surveys from GNWT-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 84 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 WMIS and for the 2001 survey we scanned polygons from 1 year compared to 2 or 3 years, maps from Jenkins et al. (2011). but this comparison broadly supports our point Between 1989 and 1998, the CWS con- of larger ranges in unfavourable years. ducted unsystematic helicopter surveys to de- In 2003, M. Ferguson (Government of Nu- scribe the relative distribution and sex and age navut) fitted seven Peary caribou cows with composition in June, July and August (Miller, satellite transmitters, which provided data from 1991; 1992; 1993; 1994; 1995a; 1997; 1998; 29 April 2003 to 18 May 2006. Jenkins & Le- Miller & Gunn, 2003). Bathurst Island was di- comte (2012) initially reported the collared vided into 12–13 search zones, and a helicopter caribou use of sea-ice, calving and wintering was used to fly a low-level, unstructured path areas. Using the same collar data, we examined through each zone. Live animals and carcasses calving, summer and wintering areas relative to were counted, and composition data were col- the proposed Park boundaries. The satellite collected. We summarize the proportionate use lars had corresponding <150 m, <350 m, and of the areas as described in Miller’s reports as <1,000 m error with satellite collar locations of the observations were neither mapped nor geo- location quality classes 3 (79% of locations), 2 referenced. (18%), and 1 (2%), respectively, and a 2-day duty cycle from 10 April to 10 July and a 5-day 2. Satellite telemetry duty cycle for the remainder of the year. Miller (1997; 1998; n.d.) captured and fitted From the 2003–06 collar dataset we devel- satellite and VHF collars on adult cow and oped annual and seasonal (summer and winter) bull caribou in 1993 and 1994 on Bathurst ranges using 90% fixed kernels (Worton, 1989; Island and the Governor General Islands. We Seaman & Powell, 1996) using the Home scanned and digitized Miller’s (n.d.) maps to Range Extension (Rodgers & Carr, 1998) for describe collared caribou home ranges relative ArcView, with unit variance standardization, to the proposed Park boundaries. We depicted a user-defined smoothing factor of 0.60, and home ranges using a minimum convex poly- raster resolution set to 120. These polygons gon (MCP) (Mohr, 1947) since collar location depicted annual range use with two caveats: a) sample sizes were variable and digital locations mid-April to mid-July is over-represented on an were not available. Miller (2002) and Miller annual basis, and b) the collars have an associ- & Barry (2003) described the seasonal move- ated error that may be as much as 1,000 m. For ments of five satellite-collared cows and one summer range, we used 1 July to 15 September, bull for July 1993 and 1994 which was a fa- and for winter range we used 1 November to vourable winter with dry shallow snow. Dur- 15 April. ing early winters of 1994-1996 (1 September To estimate the 1993–97 and 2003–06 col- to 30 November) snowfall was high (>1.5 SD lared cow use of the proposed Park, we meas- above the 55 yr mean), suggesting these were ured the proportion of each polygon or kernel unfavourable winters. In July 1994, four of within 1) the current proposed Park boundary, the five cows and the bull were re-collared as and 2) the current proposed boundary plus the well as two more bulls. After inspection of the northeast section of Bathurst Island east of the MCPs revealed the ranges in the unfavorable boundary (and north of PBPNWA) and Cam- years were larger, we examined differences us- eron Island. ing a t-test (Two-Sample Assuming Unequal For each year (1 May to 30 April) and for the Variances and one-tail test). We acknowledge a cumulative dataset 2003–06, we calculated the possible confounding effect of comparing range number of days each caribou spent within the

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 85 proposed Park, northeast Bathurst Island out- antlers, and we assumed where they died was side of the Park boundaries, Cameron Island, close to or within a rutting area. We also exam- and south Bathurst Island. Days were calculat- ined the 2003–06 collar data and assumed that ed from sequential collar locations; movements rut occurred during the last 10 days of October. that crossed boundaries between areas were We described these locations relative to the pro- weighted by the length of segment in each area. posed Park boundaries. We examined the 2003–06 collar locations to describe the timing and direction of movement 4. Carcass locations by individual cows across the current proposed In July 1998, an unsystematic aerial survey re- Park boundaries: the area east of the northeast corded the distribution of carcasses from the boundary and movements to or from Cameron 1995–97 die-off on Bathurst Island and the Island. We generated sequential line segments Governor General Islands (Miller & Gunn, by caribou collar number and date, extracted 2003). We acquired the carcass database (lo- those segments that crossed either the Cameron cations, sex and age classes) through NWT- Island coast or the northeast boundary, calcu- WMIS. We plotted the locations of the carcass- lated segment lengths within and outside both es and summed their distribution relative to the areas, and assigned the crossing date based on proposed Park boundaries. In addition, Miller the length proportions. (1998) reported on carcasses found during an We determined the annual locations of unsystematic aerial surveys in July 1996. calving sites from the 2003–06 collar movement rates and examination in GIS (Fancy & Results Whitten, 1991; Kelleyhouse, 2001; Gunn et Relative distribution of caribou from systematic al., 2008; comparable to average daily displace- aerial surveys ment in Miller & Barry, 2003). Clusters of lo- Between 1961 and 2013 there were nine sys- cations between late May and early July that tematic surveys for Peary caribou on Bathurst demonstrated the lowest daily movement rates Island and seven of those surveys included the and a degree of localization were averaged using Governor General Islands (these islands were the Animal Movement extension to ArcView not covered in August 1974 and August 1981). (Hooge & Eichenlaub, 2000) to determine The two winter surveys (March–April 1973 the approximate annual calving site. We then and 1974) had a relatively high proportion of compared the distribution of the calving sites caribou on southern Bathurst Island, and dur- to Miller’s (2001) generalised map of known ing the two spring (latter half of May) surveys calving areas. caribou were less concentrated in the northeast stratum and more evenly distributed among 3. Rutting areas strata (Table 1). During May 2013 proportion- We obtained locations for about 160 shed bull ately more caribou groups were observed on the antlers and bull carcasses from the 1995–97 die- Governor General Islands (only two groups on off (GNWT-WMIS) from a July 1998 survey, Cameron Island). where Zittlau et al. (1999; unpubl. data) had For the systematic surveys flown in sum- collected samples for mtDNA analyses from mer 1961, 1974, 1981, 1988 and 1997, more the shed antler and carcass locations (Gunn than 35% of the groups of caribou were on the & Miller, 2003). Prime bull caribou tend to northeast stratum, especially in 1974 and 1981 shed fairly quickly after the rut. Many of the when over 70% were observed in this stratum carcasses were prime bulls with fully developed (Table 1). The proposed eastern boundary

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 86 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 of the Park runs north-south and divides the 2003). The percentage distribution among the northeast stratum into two segments (Fig. 1). three strata changed during calving and post- The percentage of groups that would have been calving (χ2 = 179, 6 df, P < 0.001) with an in- outside the Park varied among surveys (Table crease in use of the northwest stratum in July 1). In 1981, over 45% of the caribou groups in compared to June (Table 2). The pattern of the northeast stratum would have been outside distribution changed with very high (88%) use the proposed Park boundary eastern but in the of the northeast in July 1998 (after the ~90% other years, the percentage excluded from the decline) and low use of the other areas. Park was generally 10–15%. Describing the relative distribution of Peary Satellite telemetry caribou among the islands off the Governor From July 1993 to July 1994, two of the four General Islands is hindered as very few caribou collared caribou on northern Bathurst Island groups were recorded in several surveys. Overall had home ranges that were within the 2002 between 1961 and 2001, about 70% of the car- Park boundaries. Two of the cows used the east ibou groups within the Governor General Is- coast and crossed the northeast boundary. The lands occurred on the two largest islands (Vani- two caribou on the Governor General Islands er and Cameron), although in May 2013 only moved between those islands including Cam- 4% of groups occurred on Cameron Island and eron Island and Bathurst Island. However, the the remainder were relatively evenly distributed pattern changed during the following 2 years on the other Governor General Islands. with unfavourable winters, as the caribou made wide-ranging movements within and Relative distribution of caribou from unsystematic off Bathurst Island. One cow in October 1995 surveys moved 110 km across sea ice to Lougheed Is- Unsystematic aerial surveys were flown in land and then a further 110 km to Borden Is- June–August 1989–95 (Miller 1991; 1993; land where she died in December 1995. All six 1994; 1995a; 1997) and 1998 (Miller & Gunn, caribou made movements across the 2002 Park

Table 1. Proportion of caribou groups relative to survey strata (Fig. 1) for Bathurst Island and the Gov- ernor General Islands, and for within and outside the proposed Park boundary in the northeast stratum, derived from figures in the survey reports, NU, 1961–2013.

Proportion of groups Proportion of groups in NE II Survey NW I NE II S III GG Is. Within Park Outside Park Jun-Jul 61 24 48 1 27 33 15 Mar-Apr 73 18 22 51 8 6 15 Mar 74 6 23 57 13 15 9 Aug 74 14 71 14 0 57 14 Aug 81 12 77 12 0 29 47 Jul 88 18 36 10 35 --a --a Jul 97 29 43 14 14 29 14 May 01 38 21 29 13 13 8 May 13 32 20 15 25 6 14

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 87 Table 2. Proportion of caribou observed corrected for effort (flying time) relative to the three survey strata on Bathurst Island during unsystematic surveys in June, July and August 1989-1998, Nunavut (Miller 1991; 1993; 1994; 1995a; 1997; Miller & Gunn, 2003). Proportion of caribou by survey Strata Jun 89 Jul 89 Jun 91 Jul 91 Jun 92 Jul 92 Aug 93 Jun 95 Jul 95 Jul 98 NW I 8 42 10 28 16 39 19 42 36 5 NE II 67 38 60 60 48 51 69 46 56 88 S III 25 20 29 12 37 10 12 13 8 6 boundaries and three of the six used Cameron the percent of home ranges within the Park Island. By 1997, all the satellite-collared cari- dropped to 31% (± 8.3%) and 42% (± 7.1%) if bou were dead. northeast Bathurst was included. The percent- The size of the ranges (100% MCP) varied age use of Cameron Island scarcely changed, (t = 2.1, df = 3, P = 0.06) between the 1993–94 but the use of outlying areas increased sharply (rated as a favourable winter) and 1994–97 during unfavourable years (from 13 to 51%). rated as unfavourable winters (Fig. 2). In the None of seven 2003-06 caribou home ranges favourable year, four females had a home range was restricted to the proposed Park and only four times smaller (2,118 ± 210 km2) than the one cow’s home range on southern Bathurst home range area for five females during the Island (summer and winter) was outside the unfavorable years (8,899 ± 3,191 km2). For 2002 boundaries, northeast Bathurst and Cam- the three individual females collared for more eron Island. For the other six cows, the mean than 1 year, home ranges in the favorable year percentage of home ranges in the Park was 40% (2,004 km2) were five times smaller than dur- (± 4.0% SE) which only increased to 43% (± ing the unfavorable years (10 983 km2). Home 4.6%) if the area of northeast Bathurst Island ranges were larger for the single male during was included and 54% (± 5.5%) if Cameron the favourable year (6,132 km2) and also for Island was included. the two males during the unfavourable years The summer ranges show more concentrat- (5,604 and 7,780 km2). ed use for northern Bathurst Island extending The seven female caribou tracked from southwest into the PBPNWA (Fig. 3). Of the 2003–06 (Jenkins & Lecomte, 2012) had a Governor General Islands, only Vanier Island mean annual range (90% fixed kernel) of 3,994 received use in summer although calving oc- ± 747.4 km2 (SE). Seasonal ranges did not show curred on Alexander Island. The winter ranges any consistent trends in size among 2003–06. for the six cows (Fig. 4) showed a contrasting Although none of the 1993–97 caribou pattern to summer. Four of the six cows win- home ranges were restricted to within the 2002 tered on Cameron Island for two or three win- boundaries, four of the six caribou in a favour- ters (Fig. 4). able year used the Park but also the segment of The seven collared caribou monitored dur- northeast Bathurst Island outside the Park. The ing 2003–06 spent an average of 41% of their other two caribou also used the Park, Cameron time within the proposed Park boundaries, Island and outlying area. The mean percentage 30% on Cameron Island, 3% on the northeast of home ranges in the Park in a favourable year coast, and 26% on south Bathurst Island. The was 55% (± 7.2% SE) which rose to 83% (± seven animals made limited use of northeast 11.7%) if the northeast coast of Bathurst Island Bathurst Island. Differences among years were was included. However, in unfavourable years relatively minor (e.g., time within the proposed

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 88 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Figure 2. Minimum convex polygons (100%) for 1993–94 and 1994–97 satellite-collared caribou relative to favourable (green polygons) and unfavourable (red polygons) winters (see text); Bathurst Island Complex, NU (derived from F. L. Miller. unpubl.).

Park boundaries for 2003–04, 2004–05, and half occurring from late July to late September. 2005–06 were 48%, 36%, and 40%, respec- Three of four collared cows in 1994 calved tively) and non-significant χ( 2 = 7.9, df = 6, P within the 2002 boundaries and the fourth cow = 0.24). calved on the east coast (Miller & Barry, 2003). Two cows from the 2003–06 telemetry data- The cows collared 2003–06 had a total of 21 set crossed the northeast boundary a total of 16 suspected calving sites and 80% were within times and five cows crossed a total of 31 times the 2002 boundaries. One calving site was out- to or from Cameron Island. Most crossings to side the northeast boundary and one cow had Cameron Island occurred in September and three calving sites on south Bathurst Island. October (median 7 October, 80% occurred be- One cow calved for 3 years on Alexander Island tween 15 September and 25 October; n = 15). which is the only one of the Governor General Most crossings from Cameron Island occurred Islands used by the seven cows for calving. in April (median 20 April, 81% occurred between 21 March and 20 May; n = 16). Move- Rutting areas ments to and from northeast Bathurst Island Based on cast bull antlers and 1993–97 telem- occurred throughout the year, with roughly etry, five rut areas (Miller, 2001) were mapped

Figure 3. Summer individual female Peary caribou ranges for 2003–06 (unpubl. GN data).

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 90 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 nearly twice their expected rate in the carcass Bathurst Island. In 1998, about 25% and 30% sample, while cows and juvenile/ yearling males of all carcasses were on Cameron Island and and females were underrepresented. Miller southern Bathurst Island, respectively (Miller (1998) reported 146 dead caribou (estimated & Gunn, 2003). 1,143 ± 164) found during unsystematic aerial surveys in July 1996 with a significantly greater Discussion than by chance distribution on Cameron and We suggest that based on information collected Vanier islands compared to Bathurst Island from 1961 to 2013 the proposed 2002 bound- and fewer carcasses than expected on southern aries for Qausuittuq National Park on the

Figure 4. Winter individual female Peary caribou ranges for 2003–06 (unpubl. GN data).

Figure 5. Rutting areas as obtained from 2003–06 collared cow Peary caribou. Shaded “rut areas” are from Miller (2001).

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 92 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 the caribou were collared were well-dispersed limited for the Bathurst Island Complex as across Bathurst Island and included two of the aerial surveys were not timed for the pre-rut or northwestern satellite islands (Miller, 1997, rut. Mapping shed antlers and telemetry identi- 1998; Jenkins & Lecomte, 2012). fied Cameron Island, as well as the north coasts Data indicate the 2002 boundaries of the of Vanier Island and Graham Moore Bay and Park are effective for protection of calving areas. the southern coast of Bathurst Island as rutting Miller (2002) mapped calving areas which are areas. The 2002 Park boundaries only include largely within the 2002 boundaries except the two of Miller’s (2002) five areas and notably ex- strip of coastal calving along the north coast of cludes Cameron Island. Bathurst Island. The 1994 calving locations of Areas outside of the currently proposed the five satellite collared cows were within those Park boundary are important to the ecology areas identified by Miller (n.d.). However, while of Peary caribou in the Bathurst Island Com- 44% of calving sites on northern Bathurst for plex. During a series of severe, unfavourable 2003–06 were within the areas mapped by winters (1994–97), collared caribou increased Miller (2002), the other 56% were either on the size of their annual ranges with increased north Bathurst Island (within the 2002 bound- movements, and five individual home ranges aries) or along the Park’s south boundary. Those expanded beyond the Park’s boundaries. Map- calving sites had high between-year fidelity. It is ping from Miller (2001) placed part of calving difficult to assess whether the 2003–06 calving and rutting areas on northeastern Bathurst Is- sites are a partial shift in calving distribution land. Relatively low use of northeast Bathurst following the 97% decline in abundance, indi- was detected in 2001 and 2013 (17–21%), but vidual variability or small sample size. the surveys occurred in May and possibly cari- The summer distribution of Peary caribou bou were moving from winter ranges. on the Bathurst Island Complex based on sys- The likelihood of seasonal movements by tematic aerial surveys (1961–1997) revealed caribou across the southern Park boundary is that 71–88% of Peary caribou used northern based on late winter (March-April) systemat- Bathurst Island and the southern islands of the ic aerial surveys when about half the caribou Governor General Islands. Unsystematic sur- groups were on southern Bathurst Island. How- veys also showed high use (80–94% of caribou ever, this is based on only two systematic sur- observed) of northern Bathurst Island during veys of in the early 1970s after a die-off (Miller July–August 1989–1998 during a shift from et al., 1977). The use of southern Bathurst Is- high to low abundance. Based on systematic land was expected as Inuit observations (Riewe, surveys the proportion of caribou groups in 1976) led Miller et al., (1977) and Ferguson the northeast stratum located outside of the (1991) to suggest that southeastern Bathurst proposed Park boundary varied from 10–45%, Island was the wintering range and northern with the highest use during a late summer sur- Bathurst was the summer range; more recent vey in 1981 when caribou densities were low. local knowledge also showed winter use of Use of the area outside of the proposed Park southern Bathurst Island (Taylor, 2005) . Fur- boundary is supported by the 1993–97 and ther support for this seasonal pattern was the 2003–06 collar data. Four of six home rang- directional movements of large groups moving es (1993–94) and two of seven home ranges south through Polar Bear Pass during the first (2003–06) involved movements across the two weeks of September 1970 and 1971 (Gray, northeast Park boundary. 1998). At the individual scale, two of seven The information on the rut distribution is home ranges of satellite collared cows were on

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 93 southern Bathurst Island and while their move- to a moderate degree by caribou in the past ments brought them into the PBPNWA, those (COSEWIC, 2004) – and Cameron Island has movements did not include the Park. a higher proportion of prostrate dwarf shrub- The Park boundary should include Cam- graminoid tundra. eron Island. There are three lines of evidence Currently, there is not enough information that support the importance of Cameron Island to assess the boundaries relative to any popu- as a rut area and winter range. Firstly, Miller lation structure and longer-term viability of (2002) identified Cameron Island as a rutting Peary caribou on the Bathurst Island Complex, area based on sightings of cast male antlers; although fine-scale spatial and temporal genetic the 2003–06 collar data support this observa- structure is likely (for example, Nussey et al., tion. Secondly, at the individual scale, five of 2005). Zittlau et al. (1999; unpubl. data) did 12 satellite-collared caribou during 1993–96 not find differences based on nuclear DNA used Cameron Island in their annual range in- between cast antlers and carcasses found on cluding winter; Fig. 2). Miller (2002) reported Cameron Island, the east central coast and the that between August 1993 and July 1994, a southwest coast of Bathurst Island. A possible satellite-collared cow and a bull spent 24–46% model is female philopatry and male-mediated of the year (during winter) on Cameron Is- gene flow as the males moved between the rut land. Miller (n.d.) commented that while there aggregations of the females from two neigh- were caribou with their annual range among bouring herds (Roffleret al., 2012). the islands of the Governor General Islands, Over the longer-term, the effectiveness of there were also caribou from Bathurst Island the Park boundaries in maintaining ecological that moved to Cameron Island for the rut and integrity including the population viability of winter. This is supported by both the annual Peary caribou will depend on many factors in- and winter ranges for five of seven 2003–06 cluding the severity of any future declines and satellite collared cows; use of Cameron Island consequent population bottlenecks. Genetic averaged 11% (± 2.3%) by these five individu- variation for Peary caribou sampled in 1998 als. The mapped winter range for the 2003–06 across the Bathurst Complex is lower than oth- cows shows extensive use of Cameron Island er caribou, possibly from a genetic bottleneck (Fig. 4). Lastly, a high number (25%) of car- resulting from the 1973–74 die-off (Zittlau casses were recorded on Cameron Island after et al., 1999; unpubl. data). Consequences of the 1994–97 severe winters. Considering that low genetic variation, such as in-breeding de- Cameron Island comprises only 3.8% of the pression, are usually considered less likely than Bathurst Island Complex land area, these three demographic risk, but the relevance of this to lines of evidence show disproportionately high Peary caribou is uncertain given the extent of, use of the island in during the rut and winter, for example, the 1994–97 declines from an es- especially in years with extreme severe winters. timated 3,000 caribou to less than 100, which The distribution of vegetation complexes suggests that evolutionary selection is extreme (Gould et al., 2003) within the Bathurst Is- and the survivors may be a particular sub-set land Complex suggest that adding northeast of the population (see Sinclair et al., 2003). Bathurst Island and Cameron Island would It is unknown over the longer term how the provide a greater diversity than the current pro- boundaries of a national park could affect the posed boundaries (Gunn et al., 2012). These likelihood of dispersal and the scale necessary two areas contribute a greater proportion of to minimize population fragmentation. cushion forb barrens – a vegetation type used A likely significant factor in assessing bound-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 94 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 aries for a national park will be climate change. Acknowledgements Understanding influences of a changing cli- Funding for this review was provided by Parks mate on Peary caribou distribution is complex Canada. We thank Christian Bucher, Parks with many interacting changes. Later forma- Canada, for the opportunity to conduct this re- tion and earlier break up in the extent of land view, and for patiently answering questions and fast ice is already measurable in the western providing information. As always, we appreci- Queen Elizabeth Islands including the Bathurst ate the assistance of Bonnie Fournier and Adri- Island Complex (Galley et al., 2012) which will an D’Hont, GNWT ENR, with data acquisi- have implications for inter-island movements. tion and handling. Frank L. Miller and Sam Although Peary caribou swim between the Barry helped us by sharing the original tran- closer islands (Miller, 1995b), changes in the sect data from Tener’s (1963) 1961 survey. We timing of land fast ice and the greater distances thank Peter Hale, Nicolas Lecomte and Debbie between Bathurst and Cameron islands, for ex- Jenkins for access to GN telemetry data. Two ample, may have an impact. anonymous reviewers provided excellent com- Our assessment from 1961–2013 sampled ments that improved an earlier draft of the the known range of natural variability for dis- manuscript. tribution and abundance on the Bathurst Is- land Complex. We found that the 2002 Park References boundaries are crossed in the northeast, and AANDC (Aboriginal Affairs and North- in the northwest to Cameron Island, by the ern Development Canada). 2012. Pro- seasonal movements of a relatively large proposed Bent Horn significant discovery portion of Peary caribou, even though Peary licence. http://www.aadnc-aandc.gc.ca/ caribou have been through three peaks of abun- eng/1360881767008/1360881807647 dance and two die-offs and associated periods Anderson, M. 2014. Distribution and abun- of low abundance. The 2002 boundaries of the dance of Peary caribou (Rangifer tarandus Park are more effective for protection of calv- pearyii) and muskoxen (Ovibos moschatus) ing but the distribution of caribou suggest that on the Bathurst Island Group, May 2013. northeast Bathurst Island, including outside of — Status Report, Nunavut Department of the proposed boundaries, are important during Environment, Wildlife Research Section, Ig- summer, both Cameron Island and northeast loolik, NU. 39pp. Bathurst Island are rutting areas with relatively Berger, J. 2004. The last mile: how to sustain high use, and Cameron Island is important long-distance migration in mammals. — Con- during winter. Our analyses of the available servation Biology 18: 320–331. http://dx.doi. information indicate that Qausuittuq National org/10.1111/j.1523-1739.2004.00548.x Park boundaries which include Cameron Island COSEWIC. 2004. COSEWIC assessment and the northeast coast of Bathurst Island will and update status report on the Peary cari- be more effective in contributing to the persis- bou Rangifer tarandus pearyi and the barren- tence of Peary caribou on the Bathurst Island ground caribou Rangifer tarandus groen- Complex during most seasons and at differing landicus (Dolphin and Union population) population levels. in Canada. — Committee on the Status of Endangered Wildlife in Canada. Ottawa. x + 91 pp. Fancy, S.G. & Whitten, K.R. 1991. Selection of calving sites by Porcupine herd caribou.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 95 — Canadian Journal of Zoology 69: 1736– movement extension to ArcView. Version 1743. http://dx.doi.org/10.1139/z91-242 2.0. — Alaska Biological Science Center, Ferguson, M.A.D. 1991. Peary caribou and U.S. Geological Survey, Anchorage, Alaska, muskoxen on Bathurst Island, Northwest USA. Territories, from 1961-1981. — Northwest Jenkins, D.A., Campbell, M., Hope, G., Territories Department of Renewable Re- Goorts, J. & McLoughlin, P. 2011. Recent sources Yellowknife, Northwest Territories. trends in abundance of Peary caribou (Rang- File Report No. 88, 64pp. ifer tarandus pearyi) and muskoxen (Ovibos Galley, R.J., Else, B.G.T., Howell, S.E.L., Lu- moschatus) in the Canadian Arctic archi- kovich, J.V. & Barber, D.G. 2012. Landfast pelago, Nunavut. — Department of Envi- sea ice conditions in the Canadian Arctic: ronment, Government of Nunavut, Wildlife 1983–2009. — Arctic 65: 133–144. http:// Report No. 1, Pond Inlet, Nunavut. 184pp. dx.doi.org/10.14430/arctic4195 Jenkins, D. & Lecomte, N. 2012. All about Gould, W.A., Raynolds, M. & Walker, D.A. ice: Peary caribou movements in the Ba- 2003. Vegetation, plant biomass, and net thurst Islands complex. — Department of primary productivity patterns in the Canadi- Environment, Government of Nunavut, an Arctic. — Journal of Geophysical Research unpublished report, Pond Inlet, Nunavut. 108: 1–14. 22pp. Gray, D. R. 1998. The Peary caribou of Polar Kelleyhouse, R.A. 2001. Calving ground habi- Bear Pass. — Unpublished report prepared tat selection: Teskekpuk Lake and Western for Department of Indian Affairs and North- Arctic caribou herds. — MS thesis, Univer- ern Development. 22pp. http://dx.doi. sity of Alaska, Fairbanks, Alaska. 140pp. org/10.1029/2001JD000948 Landres, P.B., Morgan, P. & Swanson, F.J. Gunn, A. & Dragon, J. 2002. Peary caribou 1999. Overview of the use of natural vari- and muskox abundance and distribution on ability concepts in managing ecological sys- the western Queen Elizabeth Islands, North- tems. — Ecological Applications 94: 1179– west Territories and Nunavut June-July 1188. 1997. — Northwest Territories Department Miller, F.L. 1989. Reevaluation of the status of of Resources, Wildlife and Economic Devel- Peary caribou and muskox populations with- opment File Report No. 130. 93pp. in the Bathurst Island complex, Northwest Gunn, A., Poole, K.G. & Wierzchowski, J. Territories, July 1988. — Technical Report 2008. A geostatistical analysis for the pat- Series No. 78. Canadian Wildlife Service, terns of caribou occupancy on the Bathurst Prairie & Northern Region, Edmonton, Al- calving grounds 1966-2007. — Unpub- berta. 86pp. lished report submitted to Indian & North- Miller, F.L. 1991. Peary caribou calving and ern Affairs Canada. Yellowknife, NT. 51pp. postcalving periods, Bathurst Island com- Gunn, A., Poole, K.G. & Wierzchowski, J. plex, Northwest Territories, 1989. — Tech- 2012. Peary caribou distribution within nical Report Series No. 118. Canadian the Bathurst Island Complex relative to the Wildlife Service, Prairie & Northern Re- Federal Government boundary proposed gion, Edmonton, Alberta. 72pp. for Qausuittuq National Park, northern Ba- Miller, F.L. 1992. Peary caribou calving and thurst Island, Nunavut. — Unpublished re- postcalving periods, Bathurst Island com- port for Parks Canada, Ottawa, ON. 58pp. plex, Northwest Territories, Canada, 1990. Hooge, P.N. & Eichenlaub, B. 2000. Animal — Technical Report Series No. 151. Cana-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 96 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 dian Wildlife Service, Prairie & Northern Miller, F.L. n.d. Peary caribou range use within Region, Edmonton, Alberta. 87pp. the Bathurst Island complex, South-central Miller, F.L. 1993. Peary caribou calving and Queen Elizabeth Islands, Northwest Terri- postcalving periods, Bathurst Island com- tories. — Unpublished report – preliminary plex, Northwest Territories, Canada, 1991. draft working copy, Canadian Wildlife Ser- — Technical Report Series No. 166. Cana- vice, Edmonton, AB. dian Wildlife Service, Prairie & Northern Miller, F.L. & Barry, S.J. 2003. Single- Region, Edmonton, Alberta. 99pp. island home range use by four female Miller, F.L. 1994. Peary caribou calving and Peary caribou, Bathurst Island, Canadian postcalving periods, Bathurst Island com- High Arctic, 1993-94. — Rangifer Spe- plex, Northwest Territories, 1992. — Tech- cial Issue No. 14: 267–281. http://dx.doi. nical Report Series No. 186. Canadian org/10.7557/2.23.5.1710 Wildlife Service, Prairie & Northern Re- Miller, F.L. & Barry, S.J. 2009. Long-term gion, Edmonton, Alberta. 99pp. control of Peary caribou numbers by un- Miller, F.L. 1995a. Peary caribou conservation predictable, exceptionally severe snow or ice studies, Bathurst Island complex, Northwest conditions in a non-equilibrium grazing sys- Territories, July-August 1993. — Technical tem. — Arctic 62: 175–189. http://dx.doi. Report Series No. 230. Canadian Wildlife org/10.14430/arctic130 Service, Prairie & Northern Region, Ed- Miller, F.L. & Gunn, A. 2003. Catastrophic monton, Alberta. 76pp. Die off of Peary Caribou on the Western Miller, F.L. 1995b. Inter-island water crossings Queen Elizabeth Islands, Canadian High by Peary caribou, south-central Queen Eliz- Arctic. — Arctic 56: 381–390. http://dx.doi. abeth Islands. Arctic 48:8–12. http://dx.doi. org/10.14430/arctic635 org/10.14430/arctic1219 Miller, F.L., Russell, R.H. & Gunn, A. 1977. Miller, F.L. 1997. Peary caribou conservation Distributions, movements and numbers of studies, Bathurst Island complex, Northwest Peary caribou and muskoxen on western Territories, April-August 1994 and June-July Queen Elizabeth Islands, Northwest Territo- 1995. — Technical Report Series No. 295. ries, 1972-74. — Canadian Wildlife Service Canadian Wildlife Service, Prairie & North- Report Series No. 40, Edmonton, Alberta. ern Region, Edmonton, Alberta, 155pp. 55pp. Miller, F.L. 1998. Status of Peary caribou and Mohr, C.O. 1947. Table of equivalent popula- muskox populations within the Bathurst tions of North American small mammals. — Island complex, south-central Queen Eliza- American Midland Naturalist 37: 223–249. beth Islands, Northwest Territories, July http://dx.doi.org/10.2307/2421652 1996. — Canadian Wildlife Service Techni- Nussey, D.H., Coltman, D.W., Coulson, T., cal Report Series No. 317. 147pp. Kruuk, L.E.B., Donald, A., Morris, S.J., Miller, F.L. 2001. Peary caribou – a case for a Clutton-Brock, T.H. & Pemberton, J. national park on northern Bathurst Island. 2005. Rapidly declining fine-scale spatial ge- — Unpublished report prepared for the Ca- netic structure in female red deer. — Molec- nadian Nature Federation. 38pp. ular Ecology 14: 3395–3405. http://dx.doi. Miller, F.L. 2002. Multi-island seasonal home org/10.1111/j.1365-294X.2005.02692.x range use by two Peary caribou, Canadian Parks Canada. 2012. Qausuittuq National High Arctic, 1993-94. — Arctic 55: 133– Park proposal Bathurst Island, Nunavut pro- 142. http://dx.doi.org/10.14430/arctic697 ject proposal. — Prepared by Parks Canada

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 98 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

Biodiversity offsets and caribou conservation in Alberta: opportunities and challenges

Christine B. Robichaud1 & Kyle H. Knopff2

1 Golder Associates Ltd., 16820 107 Avenue, Edmonton, Alberta, Canada T5P 4C3 (Corresponding author: [email protected]). 2 Golder Associates Ltd., 102, 2535 - 3rd Avenue S. E., Calgary, Alberta, Canada T2A 7W5.

Abstract: The federal recovery strategy for boreal woodland caribou Rangifer( tarandus caribou) sets a goal of self-sustaining populations for all caribou ranges across Canada. All caribou herds in Alberta are currently designated as not self-sustaining and the recovery strategy requires an action plan to achieve self-sustaining status. At the same time, continued natural resource extraction in caribou ranges may be worth hundreds of billions of dollars. Some regulatory bodies have recognized an opportunity for biodiversity offsets to help meet the caribou recovery strategy’s goals while still permitting economic benefits of development. In this review, we evaluate offset opportunities for caribou in Alberta and practical impediments for implementation. We conclude that a number of actions to offset impacts of development and achieve no net loss or net positive impact for caribou are theoretically feasible (i.e., if implemented they should work), including habitat restoration and manipulations of the large mammal predator-prey system. However, implementation challenges are substantial and include a lack of mechanisms for setting aside some resources for long periods of time, public opposition to predator control, and uncertainty associated with loss-gain calculations. A framework and related policy for offsets are currently lacking in Alberta and their development is urgently needed to guide successful design and implementation of offsets for caribou.

Key words: Alberta; biodiversity offsets; conservation; habitat restoration; no net loss; woodland caribou.

Rangifer, 35, Special Issue No. 23, 2015: 99-122 DOI 10.7557/2.35.2.3636

Introduction In an effort to halt escalating global biodiver- and even the corporate sector. Examples of sity loss caused by human development activi- government-driven NNL or NPI requirements ties, new development projects are increasingly are numerous and include the United States required to achieve no net loss (NNL) or net wetland policy (Environmental Law Institute, positive impact (NPI) for biodiversity. Such 2002), Canada’s fish habitat compensation requirements come from a variety of sources, policy (Pearson et al., 2005), and France’s no including governments, lending institutions, net loss policy for biodiversity (Quétier et al.,

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 99 2014). The International Finance Corporation’s mitigated (Kiesecker et al., 2010; Saenz et Performance Standard 6 requires that develop- al., 2013). Offsets should achieve long-term ers demonstrate an approach to achieve NNL conservation outcomes (IFC, 2012b), lasting for biodiversity in natural habitats and NPI in at least as long as the impacts from the pro- critical habitat prior to obtaining a loan from ject (Bull et al., 2013a). Offsets are typically the World Bank (IFC, 2012a), and many other achieved through three primary types of ac- large lending institutions have similar require- tions: financial mechanisms, protecting exist- ments. Several resource extraction and manu- ing biodiversity that might otherwise be lost, facturing companies, including large mining and enhancing existing biodiversity through companies such as Teck and Rio Tinto, have management actions (Poulton, 2014). adopted voluntary corporate NNL or NPI poli- Biodiversity offsets have received much at- cies with respect to the impacts of their opera- tention over the last decade and are increasingly tions on the environment; by 2013 at least 32 advocated as a conservation tool (Bull et al., companies had made public commitments to 2013a). However, with the exception of com- NNL or NPI (Rainey et al., 2014). pensation for fish habitat, offsets have not been Biodiversity offsets are a key mechanism for widely applied in Canada. This is now chang- achieving NNL or NPI and addressing some of ing with increasing pressure for application of the world’s most pressing conservation prob- offsets in Alberta (Dyeret al., 2008; ABCOG, lems. Biodiversity offsets are defined by the 2009), and regulatory requirements to main- Business and Biodiversity Offsets Programme tain critical habitat as defined under Canada’s (BBOP) as “measurable conservation outcomes Species at Risk Act. resulting from actions designed to compensate for Conservation of boreal woodland caribou significant residual adverse biodiversity impacts ([Rangifer tarandus caribou], hereafter caribou), arising from project development after appropri- which are a species listed federally as ‘Threat- ate prevention and mitigation measures have been ened’ on Schedule 1 of the Species at Risk Act taken” (BBOP, 2012a). Offsets can be applied (Species at Risk Public Registry, 2014) and pro- to biodiversity as a whole, but are frequently vincially as “At Risk” (AESRD, 2010), is among applied to individual biodiversity elements, the most pressing issues for which offsets are be- such as a habitat type or even individual spe- ing considered in Alberta (Habib et al., 2013). cies, depending on what biodiversity elements Most of Alberta’s caribou populations are de- are significantly impacted (IFC, 2012a; Bull clining rapidly (Hervieux et al., 2013), generat- et al., 2013b). Actions to achieve offsets occur ing international attention and numerous ap- along a continuum of compensation for adverse peals to limit development in Alberta’s caribou impacts, but compensation must reach a mini- ranges (e.g., ALT, 2009; Boutin, 2010; Ethical mum of NNL before a true offset is achieved Consumer, 2010). On the other hand, devel- (BBOP, 2012a). opment in Alberta’s caribou ranges contributes Offsets should be employed only as the final significantly to the Canadian economy and op- step of the standard mitigation hierarchy for portunity costs associated with protecting all of development projects, which includes avoid- Alberta’s caribou ranges (i.e., avoiding all future ing adverse effects where possible, minimizing impacts) have been estimated to be in excess of adverse effects to the extent feasible, restoring 100 billion dollars (Schneider et al., 2010). biodiversity to the extent practicable through In theory, offsets designed to achieve NNL restoration, and finally compensating for any or NPI for caribou could simultaneously sup- residual effects that could not be otherwise port both development and caribou conserva-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 100 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 tion; however, such offsets may prove difficult decline (Hervieux et al., 2013). To achieve to achieve in practice. Demonstrations of NNL measurable conservation outcomes of NNL or or NPI for biodiversity as a consequence of better for a development project in a cost effec- large development projects are few, and off- tive way, offsets must focus efforts on address- set implementation rarely meets the concep- ing the most important of these causes. tual principles applied during planning (Fox Available evidence overwhelmingly indicates & Nino-Murcia, 2005; Bull et al., 2013a). In that predation is the primary proximate factor Canada, for example, 67% of fish habitat com- limiting caribou populations (Bergerud, 1988; pensation failed to meet objectives and resulted Stuart-Smith et al., 1997; James & Stuart- in net losses of habitat area (Quigley & Harper, Smith, 2000; Boutin et al., 2012; Hervieux et 2006a). Moreover, government NNL or NPI al., 2013; Latham et al., 2013). Increased preda- policies may not be successful because organi- tion can largely be attributed to a phenomenon zations to govern such policy or institutional known as apparent competition (Holt, 1977) frameworks to evaluate and enforce them have where growing number of predators, such as not been created, and/or because the appropri- wolves (Canis lupus), increase in tandem with ate legal framework permitting implementa- the number of primary ungulate prey, such as tion of offset requirements is not in place (Qué- white-tailed deer (Odocoileus virginianus), to tier et al., 2014). the detriment of secondary prey species, such as The purpose of this review is to better un- caribou. White-tailed deer have increased sub- derstand the opportunities and challenges as- stantially in northeast Alberta in recent decades sociated with developing a caribou offset strat- (Dawe, 2011; Latham et al., 2011a), resulting egy in Alberta and improve the potential for in a near doubling of the wolf population in the successful offset applications. We have organ- west side of the Athabasca River caribou range ized our review into sections presenting: 1) the from approximately 6 wolves/1000 km2 in the causes of caribou decline in Alberta, 2) a review mid 1990’s to more than 11 wolves/1000 km2 of recent regulatory mechanisms and decisions in 2005-2009 (Latham et al., 2011a), and prob- recommending or requiring caribou offsets, 3) ably in most other caribou ranges (Hervieux a theoretical discussion of offset opportunities et al., 2013). Other predators such as cougars for caribou, 4) an investigation of the practical (Puma concolor) and black bears (Ursus ameri- challenges associated with implementing cari- canus) may also contribute to caribou declines bou offsets, and finally 5) our conclusions and in some places, and white-tailed deer increases recommendations for a successful offset appli- have been linked to cougar population increase cation. and expansion in Alberta, including into caribou range (Knopffet al., 2014). Causes of caribou decline in Alberta Although predation is the proximate cause In order to identify opportunities to efficiently of decline, changing predator-prey dynamics in and effectively offset adverse effects of a- de Alberta’s caribou ranges are ultimately driven by velopment project to a particular biodiversity broader landscape-level habitat changes caused value, one must first understand the proximate by agriculture, forestry, oil and gas, and climate and ultimate drivers of change for that value, change, which are creating increasingly favora- even when they may not be immediately linked ble conditions for species like white-tailed deer to activities of a project. In the case of caribou (Dawe, 2011; Boutin et al., 2012). Specifically, in Alberta, substantial research over the last 3 anthropogenic development is creating more decades has clearly identified causes of rapid early successional or other habitats with higher

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 101 forage potential for non-caribou ungulates. In has been addressed, actions focused on creat- addition to increasing the number of predators ing new habitats, more forage, or otherwise in- on the landscape, anthropogenic development creasing landscape-level carrying capacity may in caribou range may facilitate increased wolf fail to improve caribou conservation prospects movements into caribou habitats along lin- in the short term. Over the long term, creat- ear features such as roads, pipelines and seis- ing new habitats through restoration activities mic lines, which can increase encounter rates will be essential to address the ultimate cause of between wolves and caribou, thereby increas- caribou decline (Hervieux et al., 2013). ing predation (Latham et al., 2011b; DeC- esare, 2012; McKenzie et al., 2012). Wolf use Regulatory requirements for caribou offsets of linear features to make forays into caribou In 2012, the Government of Canada released habitat may be especially high during summer a federal recovery strategy that aims to achieve when wolves focus hunting efforts on prey such “self-sustaining local populations in all caribou as beavers (Castor canadensis) that tend to be ranges throughout their current distribution in found in the same habitats as caribou (Latham Canada, to the extent possible” (Environment et al., 2013). Canada, 2012a, p. 19). The status of 51 identi- Less well understood are possible changes fied local caribou populations include 26 that in caribou carrying capacity as a result of an- are “not self-sustaining”, 10 that are “as likely thropogenic development. Forage availability as not self-sustaining”, 14 that are “self-sus- is, of course, fundamental to the persistence taining”, and 1 is “unknown”. Where popula- of caribou populations (Darby et al., 1989). tions are not self-sustaining, the federal strategy Some studies have found that caribou avoid an- dictates that recovery actions be implemented. thropogenic disturbances such as seismic lines, All of the populations in Alberta (n=12) are roads, or forestry cutblocks (Dyer et al., 2001), considered “not self-sustaining” (Environment and cutblocks in particular may reduce avail- Canada, 2012a). able forage over long periods of time (Herbert To achieve acceptable probability of a self- & Weladji, 2013). Other developments such sustaining caribou population (i.e., 60%), the as oil sands mines completely remove caribou recovery strategy sets a target of at least 65% foraging opportunities over large areas. How- undisturbed habitat within each caribou range. ever, forage quantity and quality probably The proportion of undisturbed habitat within does not limit non-migratory caribou popula- Alberta’s 12 caribou ranges varies between tions (McLellan et al., 2012). Indeed, because 5% and 43% (Environment Canada, 2012a). caribou may avoid areas of high forage qual- Because critical habitat for caribou has been ity that also have high predation risk (Briand identified and a disturbance threshold within et al., 2009), addressing the predation problem critical habitat has been set, developments can could both improve survival and recruitment theoretically be stopped under Canada’s Species and provide recovering caribou populations at Risk Act should the development compro- with additional access to high quality forage mise the ability of a range to maintain or be resources. restored to 65% undisturbed habitat. Consequently, to achieve measureable con- The recovery strategy is based on a habitat servation benefits for caribou, caribou offsets disturbance threshold because the probability focused on managing large mammal predator- of a population being self-sustaining is linked prey dynamics in caribou range may prove to to the proportion of disturbed habitat con- be most effective. Until the predation problem tained within its range (Environment Canada,

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 102 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 2012a). Environment Canada (2011) con- a caribou offset: 1) protecting existing caribou cluded that the percentage of range disturbed, habitat that might otherwise be lost, 2) restor- defined as all anthropogenic disturbances plus ing disturbed caribou habitat, 3) manipulating a 500 m buffer and all areas burned in the last the predator-prey system to reduce predation 40 years, best explained the variation in calf rates, and 4) in lieu fees. recruitment across 24 ranges, which probably Regardless of which type of action or com- also reflects the extent to which the predator- bination of actions is used, at least NNL must prey system has changed. The 500 m buffer be demonstrated to achieve an offset. Demon- around anthropogenic disturbances not only strating NNL or NPI entails some method of captured the effects of habitat loss but also measuring losses as a result of development and those related to fragmentation and spatial con- gains as a result of conservation actions (Qué- figuration of disturbances (Environment Can- tier & Lavorel, 2011; BBOP, 2012b). The ideal ada, 2011). Hervieux et al.’s (2013) evaluation measure, or currency, to use for offsets focused of caribou demographics in Alberta supported on a single species, such as caribou, is the num- Environment Canada’s habitat-based approach ber of individuals in the population (Doherty et linking range condition and population viabil- al., 2010), or a surrogate that accurately reflects ity. Habitat creation through caribou offsets this. Gains measured using the selected curren- for each newly approved project is one way to cy must demonstrate additionality relative to a work towards achieving recovery strategy habi- counterfactual in an amount equal to or greater tat intactness goals (i.e., if NPI is achieved), or than the losses incurred from the project. Ad- at least to prevent further reductions of undis- ditionality means that the conservation actions turbed habitat within critical habitat ranges al- undertaken as part of a development project ready below 65% undisturbed habitat (i.e., if must be over and above actions planned with- offsets achieve NNL). out the project (BBOP, 2013). A counterfactual Some regulatory bodies within Canada have is a measurement of what might have occurred begun to request or require caribou offsets as without implementation of the conservation part of review panel recommendations or ap- action (Bull et al., 2014). proval conditions for new development pro- Protecting existing caribou habitat that jects in caribou ranges. Recently issued project would otherwise be impacted achieves what is approvals recommending or requiring offsets known as an averted loss offset. Averted loss off- for caribou are summarized in Table 1. Guid- sets can be problematic because they would still ance from the National Energy Board on off- result in a net decline in caribou populations set plan design has evolved with each project relative to existing conditions (Maron et al., decision as more detailed requirements have 2012). However, averted loss offsets still pro- been included in project decisions over time. duce an advantage compared to a case where Projects reviewed by joint federal and provin- all development were to proceed, and such off- cial panels have offset considerations limited to sets can be acceptable where background rate recommendations, as opposed to conditions. of habitat loss is high and protective legislation and mandatory compensation policies are Offset opportunities not in place (Gibbons & Lindenmayer, 2007; In theory, actions to achieve a caribou offset Maron et al., 2012). can take a variety of forms. Based on our re- The most common method of achieving an view of the causes of caribou decline, we inves- averted loss offset is to apply some mechanism tigated four types of actions that might achieve of permanent protection to land that is other-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 104 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 NEB, Source 2013 2013 CEAA, AER & - - habitat restoration and offset measures implemented as part of birds that are wetland-reliant or species at risk. indirectly-disturbed caribou habitat […] […] The Panel further recommends that during development of […] the CHRP and Offset Measures Plan. the CHRP the biodiversity management framework, consideration be given to principles such as no net loss of caribou habitat, limiting linear Northern Gateway must file with the NEB for approval […] an Northern Gateway must file with the NEB for approval […] a pro Offset Requirements mitigate project effects. The potential use of conservation offsets mitigate project effects. project effects to wetland-reliant species at risk and migratory Offset Measures Plan for each affected caribou range to offset should include a consideration of the need to compensate for gram for monitoring and verifying the effectiveness of caribou disturbances in critical caribou habitat, and restoration of histori all residual Project-related effects resulting from directly- and and Alberta consider the need for conservation offsets to further and cal and present caribou ranges. The Panel recommends that […] the Governments of Canada Location NE Alberta / British Columbia Alberta - - - Regulator National En Regulator Energy ronmental panel) (joint review dian Envi and Cana ergy Board Alberta Assessment Agency - - /day dry 3 m bitumen. infrastructure. Increase produc tion capacity by 15,900 tion capacity Project Description mining areas, and associated processing facilities, utilities and pipeline spanning 1,178 km and Construction of a new oil product open pit oil sands extraction mine. of way along the entire route. associated facilities. Project appli cation identified a 1 km wide right Amendment to expand an existing Addition to include additional - NGTL = NOVA Gas Transmission Ltd.; NEB = National Energy Board; CHRP = caribou habitat restoration plan; NE = north-east; NW = north-west. CHRP = caribou habitat restoration Board; Energy Ltd.; NEB = National Transmission Gas NGTL = NOVA Project Mine Expansion Northern Gate Project Project Energy Jackpine way Pipelines Inc. Northern Shell Canada Gateway Pipeline Condition 7(b)(vi) refers to a quantitative and qualitative assessment of the area of caribou habitat within the Chinchaga caribou range that was directly and indirectly and indirectly of caribou habitat within the Chinchaga range that was directly assessment of the area and qualitative to a quantitative Condition 7(b)(vi) refers the Caribou Protection Plan and Caribou Habitat Restoration Plan, as well as identify the remaining residual effects. residual as identify the remaining as well Plan, Restoration and Caribou Habitat Plan the Caribou Protection Notes: Notes: disturbed as a result of construction of the Project. The assessment shall identify and assess the caribou habitat to be mitigated for as a result of the implementation The assessment shall identify and assess the caribou habitat to be mitigated for as a of construction of the Project. disturbed as a result 1 Table 1 continued. Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 105 wise under legitimate threat of disturbance. For trails, seismic lines, and abandoned roads is a caribou in Alberta this type of offset could be logical first step given their prominence on Al- achieved by protecting land that occurs in cari- berta’s landscape and potential importance for bou range and would otherwise be developed. caribou predator-prey relationships (Latham A theoretical example of an averted loss offset et al., 2011b), but reclaiming any disturbance would be an oil sands development identifying in caribou range would likely count towards and protecting an area of caribou habitat that an offset. To meet the requirement of addi- otherwise would be disturbed by an approved tionality, restoration activities must target dis- forestry operation and that supports an equal turbances outside the proponent’s approved or greater number of caribou than the land dis- project footprint that have not recovered either turbed by the oil sands development. due to environmental conditions (e.g., cold, Active management interventions are re- wet soils) and/or historical clearing and restora- quired to achieve caribou offsets that provide tion practices, such as mulched seismic lines, NNL or NPI relative to baseline conditions. admixing of soils during facilities construction, Caribou populations are linked to their habitat low-blading during access clearing, and seeding (Environment Canada, 2012a), much of which of grasses on reclaimed areas. has been previously disturbed and has not been Offsets achieved by active restoration of pre- reclaimed. Restoration efforts to offset the im- viously disturbed areas away from the project pacts of a new development project on caribou can be implemented directly by project propo- populations can target historic disturbances. nents, or provided through a conservation bank This includes reclaiming historic disturbance managed by a third party. A conservation bank on public lands, but may be especially effective is an offset generated by a third party that de- in areas that were previously disturbed but now velops and controls the offset and subsequent- reside in newly created conservation areas pro- ly sells it, in whole or in part, to developers. tected under provincial land-use plans, such as Conservation banks provide an opportunity to the Lower Athabasca Regional Plan (Govern- combine habitat protection (i.e., averted loss), ment of Alberta, 2012). restoration (e.g., reforestation), and enhance- Given that anthropogenic disturbances and ment (e.g., reduced white-tailed deer density), areas deforested by wildfires within the last 40 and most importantly, would ensure the crea- years cover 57 to 95% of caribou ranges in Al- tion of offset credits before development oc- berta (Environment Canada, 2012a), we can curs. Third party offset banking is the preferred confidently assume that there exist many oppor- offset approach under the US federal wetland tunities to reclaim forested habitat preferred by compensatory mitigation system (Gardner et caribou in each of Alberta’s 12 caribou ranges. al., 2009), and could be applied to caribou. Habitat restoration can be implemented using Most offsets focus on conservation actions different techniques depending on disturbance that benefit habitat, and habitat restoration is type and local site conditions. Treatments such likely the only way to address the ultimate caus- as mounding, tree planting, tree/shrub trans- es of caribou population decline and achieve the planting and spreading of coarse woody debris federal recovery strategy goal of self-sustaining can accelerate natural reforestation or encour- caribou herds (Hervieux et al., 2013). Howev- age reforestation on sites that otherwise might er, because predation is the central proximate remain a shrubland or grassland (Coupal & cause of caribou decline, predator management Bentham, 2014). may be required to stop caribou declines in the Reclaiming historic linear features such as short term (Boutin et al., 2012; Hervieux et al.,

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 106 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 2013; Hervieux et al., 2014), and offsets focus- of the adverse impacts they have on caribou. ing on actions to reduce predation may have This type of management action would need some of the strongest immediate benefits for to be implemented by the appropriate govern- caribou. ment body (i.e., Fish and Wildlife Division) as Although not a traditional habitat-based project proponent and third parties do not have offset, actively managing interactions between jurisdiction over wildlife management. caribou and their predators addresses the most Increasing wolf populations in northern Al- immediate threat to caribou. Similar non-tra- berta appear to be driven by invading white- ditional offset actions have been proposed else- tailed deer (Latham et al., 2011a). Consequent- where. For example, impacts of unintentional ly, direct control measures also could be applied bycatch of seabirds from the fishing industry to white-tailed deer, which should elicit a nu- may not be best addressed by focusing on the merical response in the wolf population and industry itself; instead knowing that a much presumably reduce the predation pressure on greater source of seabird mortality results from caribou (Serrouya, 2013; Wittmer et al., 2013; nest predation from invasive predators provides Serrouya et al., 2015). However, we caution opportunities to deliver offsets using predator against the use of primary prey reductions as control (Pascoe et al., 2011). Another proposed an offset mechanism in isolation of other man- application of non-traditional offset is to fund agement actions such as wolf control. Without anti-poaching efforts to improve conservation simultaneously controlling wolves, reducing prospects for saiga antelope (Saiga tatarica) in non-caribou ungulate populations could cause Uzbekistan and Kazakhstan, where traditional wolves to switch to killing more caribou before protected area offsets cannot achieve NNL be- their numbers fall, exacerbating instead of re- cause saiga migrate over large areas and the pri- lieving predation pressure on caribou (Wittmer mary cause of decline is poaching, not habitat et al., 2013; Serrouya et al., 2015). Similar to loss (Bull et al., 2013b). wolf control, reductions in white-tailed deer Managing interactions between caribou and would require implementation by government; their predators can take a variety of forms. One however, hunters could play a key role if the option is to take action to directly reduce wolf approach involves liberalizing harvest quotas. populations. Wolf control measures, including Another method of changing predation aerial gunning and poisoning, have been im- rates on caribou is the use of predator fencing. plemented by the Government of Alberta in Caribou calf survival is typically low, with the the Little Smoky Caribou Range since winter highest rate of mortality occurring in the first 2005-2006 (ASRD and ACA, 2010; Hervieux month after birth (e.g., Stuart-Smith et al., et al., 2014). These measures appear to have 1997; Mahoney & Virgl, 2003; Gustine et al., been effective; the Little Smoky caribou popu- 2006). Juvenile recruitment rates are important lation growth rate increased and the population determinants of population dynamics (Gaillard has stabilized (Hervieux et al., 2013; Hervieux et al., 1998); in Alberta caribou recruitment is et al., 2014). very low due to high predation, ranging from Environment Canada (2012a) and the Gov- 0.100 to 0.206 calves/cow (Hervieux et al., ernment of Alberta (2011) highlight that main- 2013). Improving calf survival could therefore tenance and recovery of caribou is unlikely to constitute an offset action. By corralling female succeed without the implementation short- caribou into maternity pens, equipped with term predator management. By contributing to predator-proof fencing, while they give birth such efforts, developers can partially offset some and for the first few months of the newborn

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 107 calf’s life, recruitment may improve. Mater- implementing the offset. In lieu fees can work nity pens have been previously implemented to achieve a caribou offset, as defined by BBOP in Alberta (Smith & Pittaway, 2011), the Yu- (2012a), only if they fund actions that achieve kon (CCRT, 2010), and in British Columbia, a NNL or better conservation outcome. Con- where pens have been developed through part- sequently, these financial mechanisms require nership among industry, First Nations, and the the regulator, or a third party, to implement ac- provincial government (Hume, 2014). Indus- tions already described. Although in lieu fees try contribution to such a project may contrib- have not been formally identified as an option ute to or meet offset requirements for a new for caribou in Alberta, the Government of Brit- development, depending on the magnitude of ish Columbia has proposed payment of a pre- the development and of the benefit from the determined fee per hectare of caribou habitat maternity pen. Although not without risk, disturbed as an offset mechanism for future predator fencing also might be extended be- development projects (MFLNRO, 2012). The yond maternity pens to encompass larger areas proposed amount paid increases from $1250/ that could support the entire caribou lifecycle, ha to $10 000/ha as one moves from ‘low val- similar to conservation fencing implemented in ue’ caribou habitat to ‘very high value’ caribou other parts of the world (Hayward & Kerley, habitat (MFLNRO, 2012). 2009). The recovery goal for caribou targets self- Implementation challenges sustaining populations in each caribou range. Achieving NNL or NPI for caribou through Active management of predator-prey interac- careful application of the mitigation hierarchy, tions by controlling predators and non-caribou including offsets, represents an ambitious and ungulates and constructing fences may sustain laudable environmental goal and, in theory, a caribou population artificially, but when ac- there are several actions that might be imple- tive management ceases, caribou decline to- mented to achieve this for caribou. However, as wards extirpation may begin anew. In such good as offset theory may be, implementing off- cases, caribou populations would not meet the sets in practice has proven challenging. Measur- self-sustaining requirement of the federal re- able conservation outcomes that achieve NNL covery strategy. Predator control can be part of and NPI have rarely been demonstrated (Quig- an offset strategy, but is an interim solution to ley & Harper, 2005 and 2006b; Burgin, 2010). a problem that requires substantial change in Indeed, the concept of biodiversity offsets has habitat such that the carrying capacity for non- recently been criticized by academics and non- caribou ungulates and the predators they feed government organizations (NGOs) for achiev- is significantly reduced. Reclaiming disturbed ing the opposite of what it intends; instead of habitats to mature forests that support caribou biodiversity conservation, offsets sometimes but contain minimal forage for other ungulates create a “license to trash” because developers would be required with predator control. receive approvals for their developments based Lastly, in lieu fees represent payments set by on a promise to offset that cannot be realized or a regulator and made by a developer accord- for which actions are not appropriately imple- ing to a predetermined fee schedule to finance mented (e.g., ten Kate et al., 2004; Matthews actions that lead to an offset. Such payments & Endress, 2008; Walker et al., 2009; Burgin, are convenient for developers because costs of 2010). the offset are clearly defined up front and the Failure to implement appropriate action to developer is not responsible for designing or achieve an offset can have a variety of causes,

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 108 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 and effective implementation of caribou offsets of caribou offsets in Alberta have been either in Alberta requires that these are overcome. voluntary or individual offset plans required Potential problems include inconsistent in- by regulators (Poulton, 2014). Although these terpretations of NNL (Gardner et al., 2013); one-off project-specific offsets can effectively lack of information required to clearly assess compensate for impacts to caribou at the lo- and quantify project impacts (Brownlie et al., cal project scale, they are unlikely to contrib- 2013); failure to identify impacts that cannot ute to broader landscape conservation strate- be offset under any circumstance (Norton, gies and outcomes if they are not coordinated 2009; Bull et al., 2013a); inappropriate use with regional plans or initiatives (Kiesecker of metrics or currencies (Quigley & Harper, et al., 2010). On the other hand, if offset re- 2006b; Walker et al., 2009; Doherty et al., quirements are regulated and standardized, 2010; Bull et al., 2013a; Gardner et al., 2013); land managers can more readily incorporate non-compliance with regulations and lack offset actions into broader initiatives (Poul- of enforcement (Quigley & Harper, 2006a; ton, 2014). Previous consideration of offsets Matthews & Endress, 2008; Norton, 2009); in Canada (DFO, 1986; Government of Can- failure to use appropriate offset ratios (Quig- ada, 1991; Lynch-Stewart, 1996; Environment ley & Harper, 2006b; Moilanen et al., 2009); Canada, 2012b; DFO, 2013; Poulton, 2014), implementation without prior evidence of could serve as useful building blocks for policy technical feasibility or effectiveness (Gibbons makers and environmental practitioners tasked & Lindenmayer, 2007); inadequate regulatory with developing effective and efficient offset framework and government oversight (Qué- plans for caribou. tier et al., 2014); and lack of monitoring and Even if a regulatory framework were estab- maintenance (Brown & Lant, 1999; Quigley & lished, implementation of theoretical options Harper, 2006b). for caribou offsets is not straightforward. Fi- Perhaps the most important challenge that nancial mechanisms, for example, typically applies to all offset opportunities summarized include fees paid to support caribou and wolf in this paper is that there currently is no clear monitoring programs, maternal pens, as well as guidance or framework for offset requirements other research and outreach programs (MFLN- in Alberta. Provision of key design elements for RO, 2012). Although worthwhile endeavours, offsets provided through a comprehensive offset monitoring, research, and outreach do not typi- framework and policy would greatly improve cally deliver measurable conservation outcomes the effectiveness of offset implementation (e.g., and therefore do not provide offsets. Although Quétier et al., 2014). Proponents requiring off- the simplest option for developers, the risk as- sets in Alberta currently have little guidance on sociated with financial offsets is that the funds basic standards and performance criteria such do not deliver the direct conservation outcomes as (i) offset currency, (ii) loss-gain calculations, required to achieve NNL or NPI. Adequate (iii) equivalency and ‘trading up’ (e.g., Bull et means of defining on-the-ground benefits of al., 2013a; Habib et al., 2013), (iv) uncertainty actions implemented using funds generated and time lags, (v) duration, and (vi) monitor- from financial offset requirements are needed ing requirements and appropriate indicators. to demonstrate success of this approach, but An offset policy is required to identify what guidance for achieving this for caribou is not actions constitute permissible offsets and to en- currently available. The US regulations govern- sure that offsets are consistently applied across ing compensatory mitigation for wetlands and development projects. Recent applications other aquatic resources can provide useful guid-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 109 ance for a payment program (Gardner et al., bou conservation because such offsets do not 2009). Achieving specific milestones and per- typically provide NNL or NPI relative to exist- formance standards prior to the release of offset ing conditions; hence, they cannot contribute credits is a central consideration of the regula- to the net restoration of habitat required to tion and emphasizes the importance of timely achieve the caribou recovery strategy objective compensatory actions (Gardner et al., 2009). of achieving self-sustaining caribou herds. Averted loss offsets established by protect- Habitat restoration is the best approach to ing caribou habitat that otherwise would be address ultimate causes of caribou decline and disturbed are even less straightforward, to the is currently being implemented in Alberta point of being nearly impossible under cur- (Coupal & Bentham, 2014). Key problems rent provincial legislation. Caribou range in with habitat restoration are the long time-lag Alberta is almost entirely restricted to public before caribou benefit from the action and lands which cannot be purchased. The excep- uncertainty about loss-gain calculations, res- tion is a small tract of private land located in toration success, and the potential for future the Chinchaga caribou range in northwest Al- development programs where habitats have berta (http://thecarbonfarmer.ca). There is no been reclaimed. Boreal forests grow slowly and established conservation banking system in the even with extensive restoration, readjustments province, and even if one existed, private lands to predator-prey systems that are driven by that could be purchased and protected or im- landscape change at broad spatial extents (i.e., proved primarily occur outside of caribou range north-eastern Alberta) probably will take dec- and would not benefit caribou. ades. Despite Environment Canada’s (2012a) Extinguishing development rights on public support for habitat restoration, they fail to land in Alberta also is prohibited and averted provide a formal definition of restored caribou loss offsets cannot be achieved by one industry habitat, pointing to a need to develop targets paying for another not to develop (e.g., an oil and measureable criteria for restoration. Such and gas company cannot purchase development targets and criteria are required to determine rights from a Forestry Management Agreement when an offset is realized. holder). The petroleum and natural gas mineral Some guidance for offset currency is pro- rights leasing system, guided by the Mines and vided by the federal recovery strategy, which Minerals Act, requires all lease holders to ‘prove uses 65% undisturbed habitat within a cari- the mineral resource’ by drilling, production or bou range as a surrogate for achieving a self- technical mapping. Failure to do so can result sustaining population (Environment Canada, in loss of the lease. Some oil and gas produc- 2012a). Surrogate currencies can be useful, but ers in the province have indicated that provin- may have risks associated with them if they are cial regulators do not always consider technical not closely linked to the desired outcome (e.g., mapping an acceptable means of delineating self-sustaining caribou populations). For exam- the resource, which encourages more drilling ple, reclaiming linear features may provide the and therefore more habitat loss (CAPP, 2013). greatest value for a developer’s investment in Hence, setting aside or otherwise not develop- terms of demonstrating an offset using loss-gain ing an oil and gas lease in caribou range as an calculations derived from the federal caribou averted loss offset for impacts elsewhere does recovery strategy habitat models, but this may not appear to be an option. not translate into an equal benefit to caribou. Inability to implement averted loss offsets Consider a 5 m wide seismic line that extends may not be a substantial constraint for cari- over 1000 m with no other disturbances near-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 110 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 by. Because caribou range disturbance metrics areas of core habitat to maximize the benefit to in the federal recovery strategy were calculated caribou. The Alberta Public Lands Act does not by applying a 500 m buffer on either side of include mechanisms for permanent protection that seismic line, successful restoration returns of such restoration efforts; therefore, they are 5000 m2 (i.e., 5 m x 1000 m) of forest, but at risk of being destroyed or compromised by 1 010 000 m2 of caribou habitat (i.e., ([500 m other land-users. This deficiency must be -ad x 1000 m] + [5 m x 1000 m]) x 2). Reclaiming dressed to ensure that offsets are in place for 1000 m of seismic line will likely benefit cari- the duration of project impacts, and preferably bou populations, but perhaps not by a factor of for much longer (Gibbons & Lindenmayer, over 200 for every habitat unit reclaimed. 2007; McKenney & Kiesecker, 2010; Bull et Uncertainty around loss-gain calculations al., 2013a). Protection of restored sites could be and restoration success, along with any time- achieved by creating a new disposition type for lags prior to achieving functional habitat also offsets under the Public Lands Act and would may require offset multipliers that are unachiev- require the development of a regulatory review able, depending on how loss-gain calculations process to approve offset site selection. are implemented. Curran et al., (2014) suggest To be most effective, caribou habitat restora- ratios required to achieve a true NNL offset tion activities should consider future develop- through habitat restoration may be as high as ment footprint and industrial access require- 100:1, which is much higher than typically ap- ments, relative quality of adjacent caribou plied ratios less than 10:1 and probably cost- habitat, recreational use by the public, Abo- prohibitive for most development projects in riginal use, caribou occurrence, and provin- Alberta. A better option to address time lags is cial habitat restoration priorities for caribou. to have an offset policy requiring demonstrated Weighing competing land-use demands is a conservation outcomes ahead of development, challenging process and may require complex which would reduce the required ratio (Gard- and lengthy consultation, but ignoring it may ner et al., 2009; Maron et al., 2012). result in failure. For example, recreational all- There is also the matter of where to imple- terrain vehicles use can significantly hinder ment restoration (or any other offset action). revegetation efforts implemented at high cost In the context of oil and gas development, a (Coupal & Bentham, 2014). developer may choose to implement offsets by Resolving issues pertaining to habitat based conducting on-lease or off-lease habitat res- caribou offsets will be challenging, and even toration. On-lease restoration provides more if achieved, probably cannot be implemented certainty that restoration efforts will not be without simultaneously addressing preda- disturbed by future development because the tion because habitat values for caribou herds developer exerts more control over the land in Alberta are already considered below those base, albeit not full control owing to overlap- required to achieve caribou conservation (En- ping oil and gas tenures issued in stratigraphic vironment Canada, 2012a), because caribou layers and forestry management areas. On the habitat takes a long time to reclaim, and be- other hand, on-lease restoration is likely to have cause caribou are in such steep decline that any limited benefits for caribou during the project’s substantial time-lag may result in conservation operational phase because the restored areas failure (Hervieux et al., 2013). will presumably be located in proximity to ex- Addressing predation in the short-medium isting and/or future disturbances. Off-lease term is necessary, but applying them as offsets restoration provides the opportunity to target is extremely challenging. Management actions

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 111 such as predator and/or prey control through effectively achieving conservation outcomes aerial gunning or poison directly address the is often unclear (Kiesecker et al., 2009). For proximate causes of caribou decline and are caribou offsets, our interpretation of the federal much more likely to facilitate caribou popu- recovery strategy’s goal to maintain or recover lation persistence, but still present numerous all populations within caribou range in Can- challenges (Hervieux et al., 2014). Wolf con- ada (Environment Canada, 2012a) is that any trol and prey control generally generate nega- negative impacts to caribou or caribou habitat tive public perceptions (NRC, 1997; Martín- should be offset within that same range. This ez-Espiñeira, 2006; Van Ballenberghe, 2006), added restriction poses an additional challenge and developers could contribute only through by spatially limiting acceptable offset locations. a program implemented by the province, which The requirement to offset within a given cari- has the responsibility for directly managing bou range precludes the application of the tri- wildlife. Fencing options, especially large-scale age-based approach recommended by Schnei- predator exclusion fences may work, but also der et al., (2010). Given that there are limited will probably be viewed negatively (Pickard, resources to implement recovery efforts and 2007; Scofield et al., 2011) and may have un- that population viability varies among herds, intended consequences (Pople et al., 2000; Schneider et al., (2010) argue that caribou con- Norrdahl et al., 2002; Long & Robley, 2004; servation efforts should focus on probability of Hayward & Kerley, 2009). success and cost as opposed to risk of extirpa- Changes in hunting regulation to encourage tion. Considering the Alberta context where greater harvest rates of moose and deer may be resources have been over-allocated in some more palatable and have been applied previ- caribou ranges, we think it prudent that poli- ously (Serrouya et al., 2015), but this is also not cy makers consider prioritizing caribou offsets an offset that can be actioned by a developer. where there is a greater probability of success. Developers might contribute to these efforts This approach would consider caribou offsets by creating financial incentives for hunters and at the provincial scale with the trade-off of po- trappers to reach prescribed quotas for moose, tentially losing herds in highly impacted areas deer and wolves. Implementation of such in- while increasing the odds of successful conser- centives would likely necessitate lengthy ne- vation of other herds. gotiations with government in order to reach an agreement, they may be viewed negatively Conclusions and recommendations as “bounties”, and their effectiveness relative We have discussed offset opportunities for cari- to aerial control or poisoning is questionable bou and identified practical challenges associ- (Webb et al., 2011). ated with them. The prevalence of implementa- A final challenge to offset implementation tion challenges is not surprising given that the is achieving clarity about where offsets, either science of offsets is still in its early development habitat-based or focused on addressing pre- stage and government policy has not yet devel- dation, might be appropriately undertaken. oped to accommodate it. A common challenge Standard like-for-like approaches indicate that for all caribou offset opportunities in Alberta offset sites should be as close as possible to im- is the lack of framework and policy to guide pact sites to ensure that benefits are realized in consistent and appropriate application of off- the same area (McKenney & Kiesecker, 2010); sets. Successful implementation of caribou off- however, the selection of offset locations that sets will depend in part on the development of best balance proximity to the impact sites with comprehensive offset framework and policy to

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 112 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 guide project proponents and environmental the provision of measurable offsets means that practitioners. Such framework and policy are restoration cannot be the sole solution to cari- urgently needed. bou recovery. Current levels of population de- Regulations and policy need to emphasize cline dictate that restoration should be used that offsets are not a panacea. Offsets are the in conjunction with immediate management last resort in the mitigation hierarchy and are actions (i.e., manipulations of the large mam- not a solution to failing to do a good job of mal predator-prey system) addressing proxi- avoidance, minimization and rehabilitation/ mate causes of caribou declines to ensure their restoration. The mitigation hierarchy has not persistence over the short and medium-term. always been systematically applied in Alberta’s These kinds of programs are more difficult for environmental assessment process (Clare et al., industry to contribute to and may be less com- 2011; Clare & Krogman, 2013). Project devel- monly used as a caribou offset. Where industry opers and environmental practitioners could cannot contribute, such programs must remain benefit from evaluation criteria to help deter- the responsibility of government (Environment mine when one can defensibly move down the Canada, 2012a; Hervieux et al., 2014). mitigation hierarchy (e.g., move from avoid- Given the many challenges of implementing ance to minimization). Clear guidance from caribou offsets, we think there is much value government agencies would provide much in considering Schneider et al.’s (2010) triage needed consistency across projects and could perspective further. Using a provincial scale for prevent bureaucratic slippage (Clare & Krog- caribou offset site selection would facilitate this man, 2013), that is the propensity for broad approach. Although a triage approach where policies to be changed through successive rein- offsets focus on the least affected herds may terpretation, such that the ultimate implemen- mean accepting the loss of some of Alberta’s tation may bear little resemblance to the broad caribou herds, it may also mean that some statements of policy intent (Freudenburg & can be saved while simultaneously developing Gramling, 1994). some of the province’s most valuable resources Habitat restoration is likely the most prom- (Schneider et al., 2010). The current policy of ising caribou offset strategy for industry given exploration and development everywhere all the extensive opportunities for habitat restora- the time and conservation everywhere all the tion, its technical feasibility, and the fact that time may result in both conservation failure maintenance of critical habitat is mandated and higher costs and increased uncertainty for under the federal recovery strategy. Although developers. habitat restoration has been criticized as an in- We recommend that future research focus appropriate offset tool in some cases (e.g., Cur- on evaluating the efficacy of proposed offset ran et al., 2014), we argue that this approach strategies for caribou. Further empirical evi- is applicable to achieve offsets specifically for dence is required to reduce uncertainty and to caribou because habitats can be restored to a help policy makers, regulators, project propo- form that is less likely to support alternate prey. nents, and environmental practitioners make Under the current public lands tenure system, informed decision on offset design and imple- no mechanism is in place to secure restoration mentation. Specifically, a caribou offset frame- efforts on caribou ranges and this deficiency work would benefit greatly from understanding must be addressed before restoration can serve the time interval until restored habitat benefits as a viable offset strategy. caribou by adjusting predator-prey dynamics, Time-lags between restoration actions and or the scale at which restoration of historic dis-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 113 turbance must occur to achieve a measurable source Development. Wildlife Status Report benefit to caribou populations given the scale No. 30 (Update 2010). Edmonton, AB. 88 at which large mammal predator prey systems pp. operate. Similarly, changes in predation rates or BBOP (Business and Biodiversity Offsets caribou populations due to a project are rarely Programme). 2012a. Standard on Biodiver- quantified in environmental assessments, mak- sity Offsets. BBOP, Washington, DC. ISBN ing it difficult to estimate what kind of changes 978-1-932928-44-0. to the predator-prey system may be required BBOP. 2012b. Guidance Notes to the Stand- to offset impacts of a project (e.g., using wolf ard on Biodiversity Offsets. BBOP, Wash- control, large scale predator fencing, or mater- ington, D.C. ISBN 978-1-932928-47-1. nity penning). A better understanding of how BBOP. 2013. To no net loss and beyond: An to apply multiple offset currencies, including overview of the business and biodiversity both habitat and predation rates (sensu Bull et offsets programme (BBOP). Washington, al., 2013a), would be helpful for loss-gain cal- DC. ISBN 978-1-932928-55-6. culations. Bergerud, A.T. 1988. Caribou, wolves and man. —Trends in Ecology and Evolution 3: 68-72. http://dx.doi.org/10.1016/0169- References 5347(88)90019-5 Boutin, S. ABCOG (Alberta Boreal Conservation Off- 2010. Expert reporting on woodland caribou [Rangifer tarandus caribou] in sets Advisory Group). 2009. Regulated the traditional territory of the Beaver Lake conservation offsets with banking: a concep- Cree Nation. 58 pp. tual business model and policy framework. Boutin, S., Boyce, M.S., Hebblewhite, M., Report. Government of Alberta, Edmonton. Hervieux, D., Knopff, K.H., Latham, AER & CEAA (Alberta Energy Regula- M.C., Latham, A.D., Nagy, J., Seip, D. & tor and Canadian Environmental As- Serrouya, R. 2012. Why are caribou declin- sessment Agency). 2013. Decision 2013 ing in the Oil Sands? — Frontiers in Ecol- ABAER 011. Issued July 9, 2013. CEAA ogy and the Environment 10: 65-67. http:// Reference Number 59540. Available at: dx.doi.org/10.1890/12.WB.005 http://www.ceaa-acee.gc.ca/050/details-eng. Briand, Y., Ouellet, J.P., Dussault, C. & St- cfm?evaluation=59540. Accessed May 2014. Laurent, M.H. AESRD (Alberta Environment and Sustain- 2009. Fine-scale habitat selection by female forest-dwelling caribou in able Resource Development). 2010. The managed boreal forest: empirical evidence of General Status of Alberta Wild Species. a seasonal shift between foraging opportuni- 2010. Available at: http://esrd.alberta.ca/ ties and antipredator strategies. — Ecoscience fish-wildlife/species-at-risk/wild-species-sta- 16: 330-340. http://dx.doi.org/10.2980/16- tus-search.aspx. Accessed June 23, 2014 3-3248 ALT (Athabasca Landscape Team). 2009. Brown, P.H. & Lant, C.L. 1999. The effect Athabasca caribou landscape management of wetland mitigation banking on achieve- options report. 75pp. + 3 app. Environmental ASRD & ACA (Alberta Sustainable Resource ment of no-net-loss. — Management Development and Alberta Conservation 23: 333-345. http://dx.doi. org/10.1007/s002679900190 Association). 2010. Status of the Woodland Brownlie, S., King, N. & Treweek, J. 2013. Caribou (Rangifer tarandus caribou) in Al- Biodiversity tradeoffs and offsets in impact berta: Update 2010. Alberta Sustainable Re-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 114 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 assessment and decision making: Can we The case of wetland management in Alberta. stop the loss? — Impact Assessment and Pro- — Society and Natural Resources 26: 672-687. ject Appraisal 31: 24-33. http://dx.doi.org/1 http://dx.doi.org/10.1080/08941920.2013. 0.1080/14615517.2012.736763 779341 Bull, J.W., Suttle, K.B., Gordon, A., Singh, Curran, M., Hellwig, S. & Beck, J. 2014. Is N.J. & Milner-Gulland, E.J. 2013a. Biodi- there any empirical support for biodiversity versity offsets in theory and practice. —Oryx offset policy? —Ecological Applications 24: 47: 369-380. http://dx.doi.org/10.1017/ 617-632. http://dx.doi.org/10.1890/13- S003060531200172X 0243.1 Bull, J.W., Suttle, K.B., Singh, N.J. & Mil- Coupal, B. & Bentham, P. 2014. If you build ner-Gulland, E.J. 2013b. Conservation it, will they come? Caribou habitat restora- when nothing stands still: moving targets tion for pipeline projects. Proceedings of the and biodiversity offsets. —Frontiers in Ecol- 2014 International Pipeline Conference and ogy and Environment 11: 203-210. http:// Exposition. September 29-Octiber 3, 2014. dx.doi.org/10.1890/120020 Calgary, Alberta. http://dx.doi.org/10.1115/ Bull, J.W., Gordon, A., Law, E.A., Suttle, ipc2014-33577 K.B. & Milner-Gulland, E.J. 2014. Impor- Darby, W.R., Timmermann, H.R., Snider, tance of baseline specification in evaluating J.B., Abraham, K.F., Stefanski, R.A. & conservation interventions and achieving no Johnson, C.A. 1989. Woodland caribou in net loss of biodiversity. — Conservation Biol- Ontario: Background to a policy. Ontario ogy 28: 799-809. http://dx.doi.org/10.1111/ Ministry of Natural Resources, Wildlife cobi.12243 Branch, Toronto, Ontario. Burgin, S. 2010. ‘Mitigation banks’ for wet- Dawe, K.L. 2011. Factors driving range ex- land conservation: a major success or an pansion of white-tailed deer, Odocoileus unmitigated disaster? — Wetlands Ecology virginianus, in the boreal forest of northern and Management 18: 49-55. http://dx.doi. Alberta, Canada. Ph.D. thesis, University of org/10.1007/s11273-009-9147-5 Alberta, Edmonton. CAPP (Canadian Association of Petroleum DeCesare, N.J. 2012. Separating spatial search Producers). 2013. CAPP Caribou Energy and efficiency rates as components of preda- Sector Workshop. Meeting Minutes. Octo- tion risk. — Proceedings of the Royal Society of ber 22, 2013. Calgary, Alberta. London B: Biological Sciences 279: 4626-4633. CCRT (Chisana Caribou Recovery Team). http://dx.doi.org/10.1098/rspb.2012.1698 2010. Recovery of the Chisana caribou DFO (Department of Fisheries and Oceans). herd in the Alaska/Yukon borderlands: Cap- 1986. Policy for the management of fish tive rearing trials. Yukon Fish and Wildlife habitat. 28pp. ISBN 0-662-15033-3. Avail- Branch Report TR-10-02. 29 pp. able at: http://www.dfo-mpo.gc.ca/Li- Clare, S., Krogman, N., Foote, L. & Lem- brary/23654.pdf. Accessed June 2014. phers, N. 2011. Where is avoidance in the DFO. 2013. Fisheries productivity investment implementation of wetland law and poli- policy: A proponent’s guide to offsetting. cy? — Wetlands Ecology and Management 19pp. ISBN 978-1-100-22930-0. Available 19: 165-182. http://dx.doi.org/10.1007/ at: http://www.dfo-mpo.gc.ca/pnw-ppe/ s11273-011-9209-3 fpp-ppp/guide-eng.html. Accessed June Clare, S. & Krogman, N. 2013. Bureaucratic 2014. slippage and environmental offset policies: Doherty, K.E., Naugle, D.E. & Evans, J.S.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 115 2010. A currency for offsetting energy de- 1007. http://dx.doi.org/10.1111/j.1523- velopment impacts: Horse trading sage- 1739.2005.00231.x grouse on the open market. — PLoS ONE 5: Freudenburg, W.R. & Gramling, R. 1994. Bu- e10339. http://dx.doi.org/10.1371/journal. reaucratic slippage and failures of agency vi- pone.0010339 ligance: The case study of the environmental Dyer, S.J., O’Neill, J.P., Wasel, S.M. & Bou- studies program. — Social Problems 41: 214- tin, S. 2001. Avoidance of industrial devel- 239. http://dx.doi.org/10.2307/3096931 opment by woodland caribou. — Journal of Gaillard, J.M., Festa-Bianchet, M. & Yoc- Wildlife Management 65: 531-542. http:// coz, N.G. 1998. Population dynamics of dx.doi.org/10.2307/3803106 large herbivores: variable recruitment with Dyer, S., Grant, J., Lesack, T. & Weber, M. constant adult survival. —Trends in Ecol- 2008. Catching up: conservation and biodi- ogy and Evolution 13: 58-63. http://dx.doi. versity offsets in Alberta’s boreal forest. Ca- org/10.1016/S0169-5347(97)01237-8 nadian Boreal Initiative, Ottawa, ON. 34pp. Gardner, R.C., Zedler, J., Redmon, A., Turn- Environment Canada. 2011. Scientific Assess- er, R.E., Johnston, C.A., Alvarez, V.R., ment to Inform the Identification of Criti- Simenstad, C.A., Prestegaard, K.L. & cal Habitat for Woodland Caribou (Rangi- Mitsch, W.J. 2009. Compensating for wet- fer tarandus caribou), Boreal Population, in land losses under the clean water act (redux): Canada: 2011 update. Ottawa, Ontario, Evaluating the federal compensatory mitiga- Canada. 102 pp. + appendices. tion regulation. — Stetson Law Review 38 (2) Environment Canada. 2012a. Recovery Strat- :213-249. egy for the Woodland Caribou (Rangifer Gardner, T.A., Von Hase, A., Browlie, S., Ek- tarandus caribou), Boreal population, in strom, J.M.M., Pilgrim, J.D., Savy, C.E., Canada. Species at Risk Act Recovery Strategy Stephens, R.T.T., Treweek, J., Ussher, G.T., Series. Environment Canada, Ottawa. xi + Ward, G. & ten Kate, K. 2013. Biodiversity 138pp. offsets and the challenge of achieving no net Environment Canada. 2012b. Operation loss. — Conservation Biology 27: 1254-1264. Framework for Use of Conservation Allow- http://dx.doi.org/10.1111/cobi.12118 ances. Available at: http://www.ec.gc.ca/ee- Gibbons, P. & Lindenmayer, D.B. 2007. ea/default.asp?lang=En&n=DAB7DD13-1. Offsets for land clearing: No net loss or the Accessed June 23, 2014. tail wagging the dog? — Ecological Manage- Environmental Law Institute. 2002. Banks ment and Restoration 8: 26-31. http://dx.doi. and fees: the status of off-site wetland miti- org/10.1111/j.1442-8903.2007.00328.x gation in the United States. Environmental Government of Alberta. 2011. A woodland Law Institute: Washington, DC, USA. caribou policy for Alberta. 2pp. ISBN No. Ethical Consumer. 2010. Tar sands threaten 978-0-7785-9478-9. Available at: http:// caribou extinction. Posted 20 July, 2013. esrd.alberta.ca/fish-wildlife/wildlife-man- Available at: http://www.ethicalconsumer. agement/caribou-management/default.aspx. org/latestnews/tabid/62/entryid/423/tar- Accessed June 2014. sands-threaten-caribou-extinction.aspx. Ac�- Government of Alberta. 2012. Lower Athabas- cessed June 2014. ca Regional Plan. 94 pp. ISBN No. 978-1- Fox, J. & Nino-Murcia, A. 2005. Status of 4601. species conservation banking in the United Government of Canada. 1991. The Federal States. — Conservation Biology 19: 996- policy on Wetland Conservation. Canadi-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 116 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 an Wildlife Service, Environment Canada. com/news/british-columbia/maternity- 13pp. ISBN 0-662-18940-X. Available at: ward-for-caribou-in-northern-bc-keeps- http://publications.gc.ca/site/eng/100725/ wolves-at-bay/article18048278/. Accessed publication.html. Accessed June 2014. July 2015. Gustine, D.D., Parker, K.L., Lay, R.J., Gill- Holt, R.D. 1977. Predation, Apparent Com- ingham, M.P. & Heard, D.C. 2006. Calf petition, and the structure of prey commu- survival of woodland caribou in a multi- nities. — Theoretical Population Biology 12: predator ecosystem. — Wildlife Monographs 197-229. http://dx.doi.org/10.1016/0040- 165: 1-32. http://dx.doi.org/10.2193/0084- 5809(77)90042-9 0173(2006)165[1:CSOWCI]2.0.CO;2 IFC (International Finance Corporation). Habib, T.J., Farr, D.R., Schneider, R.R. & 2012a. Performance Standard 6. Biodiversity Boutin, S. 2013. Economic and ecological Conservation and Sustainable Management outcomes of flexible biodiversity offset sys- of Natural Resources. January 1, 2012. 7pp tems. — Conservation Biology 27: 1313-1323. IFC. 2012b. Guidance Note 6. Biodiversity http://dx.doi.org/10.1111/cobi.12098 Conservation and Sustainable Management Hayward, M.W. & Kerley, G.I.H. 2009. of Natural Resources. January 1, 2012. 69 Fencing for conservation: restriction of evo- pp. lutionary potential or a riposte to threaten- James, A.R.C. & Stuart-Smith, A.K. 2000. ing processes? — Biological Conservation Distribution of Caribou and Wolves in Rela- 142: 1-13. http://dx.doi.org/10.1016/j.bio- tion to Linear Corridors. — Journal of Wild- con.2008.09.022 life Management 64: 154-159. http://dx.doi. Herbert, I. & Weladji, R.B. 2013. The use org/10.2307/3802985 of coniferous forests and cutovers by New- Kiesecker, J.M., Copeland, H., Pocewicz, foundland woodland caribou. — Forest Ecol- A., Nibbelink, N., McKenney, B., Dahl- ogy and Management 291: 318-325. http:// ke, J., Holloran, J. & Stroud, D. 2009. dx.doi.org/10.1016/j.foreco.2012.11.011 A framework for implementing biodiver- Hervieux, D., Hebblewhite, M., DeCesare, sity offsets: Selecting sites and determining N.J., Russel, M., Smith, K., Robertsone, scale. — Bioscience 59: 77-84. http://dx.doi. S. & Boutin, S. 2013. Widespread decline org/10.1525/bio.2009.59.1.11 of woodland caribou (Rangifer tarandus cari- Kiesecker, J.M., Copeland, H., Pocewicz, A. bou) continue in Alberta. — Canadian Jour- & McKenney, B. 2010. Development by nal of Zoology 91: 872-882. http://dx.doi. design: Blending landscape-level planning org/10.1139/cjz-2013-0123 with the mitigation hierarchy. — Frontiers Hervieux, D., Hebblewhite, M., Stepnisky, in Ecology and the Environment 8: 261-266. D., Bacon, M. & Boutin, S. 2014. Man- http://dx.doi.org/10.1890/090005 aging wolves (Canis lupus) to recover threat- Knopff, K.H., Webb, N.F. & Boyce, M.S. ened woodland caribou (Rangifer tarandus 2014. Cougar population status and range caribou) in Alberta. — Canadian Journal expansion in Alberta during 1991 and 2010. of Zoology 92: 1029-1037. http://dx.doi. — Wildlife Society Bulletin 38:116-121 org/10.1139/cjz-2014-0142 http://dx.doi.org/10.1002/wsb.369Latham, Hume, M. 2014. Maternity ward for caribou A.D.M., Latham, M.C., McCutchen, N.A., in northern B.C. keeps wolves at bay. — The & Boutin, S. 2011a. Invading white-tailed Globe and Mail. Published 16 April, 2014. deer change wolf–caribou dynamics in Available at: http://www.theglobeandmail. northeastern Alberta. —Journal of Wildlife

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 117 Management 75: 204-212. http://dx.doi. Martínez-Espiñeira, R. 2006. Public attitudes org/10.1002/jwmg.28 towards lethal coyote control. — Human Latham, A.D.M., Latham, M.C., Boyce, Dimensions of Wildlife 11: 89–100. http:// M.S., & Boutin, S. 2011b. Movement re- dx.doi.org/10.1080/10871200600570288 sponses by wolves to industrial linear fea- Matthews, J.W. & Endress, A.G. 2008. Per- tures and their effect on woodland caribou formance criteria, compliance success, and in northeastern Alberta. — Ecological Ap- vegetation development in compensatory plications 21: 2854-2865. http://dx.doi. mitigation wetlands. — Environmental org/10.1890/11-0666.1 Management 41: 130-141. http://dx.doi. Latham, A.D.M., Latham, C., Knopff, K., org/10.1007/s00267-007-9002-5 Hebblewhite, M., & Boutin, S. 2013. McKenney, B.A. & Kiesecker, J.M. 2010. Wolves, white-tailed deer, and beaver: im- Policy development for biodiversity offsets: plications of seasonal prey switching for A review of offset frameworks. — Environ- woodland caribou declines. — Ecography mental Management 45: 165-176. http:// 36: 1-15. http://dx.doi.org/10.1111/j.1600- dx.doi.org/10.1007/s00267-009-9396-3 0587.2013.00035.x McKenzie, H.W., Merrill, E.H., Spiteri, R.J. Long, K. & Robley, A. 2004. Cost effective & Lewis, M.A. 2012. How linear features feral animal exclusion fencing for areas of alter predator movement and the functional high conservation value in Australia. Natu- response. — Interface Focus 2: 205–216. ral Heritage Trust, Department of Sustain- http://dx.doi.org/10.1098/rsfs.2011.0086 ability and Environment, State Government McLellan, M.L., Serrouya, R., McLellan, of Victoria. 61 pp. B.N., Furk, K., Heard, D.C. & Wittmer, Lynch-Stewart, P., Neice, P., Rubec, C. & H.U. 2012. Implications of body condition Kessel-Taylor, I. 1996. The federal policy on the unsustainable predation rates of en- on wetland conservation: Implementation dangered mountain caribou. — Oecologia guide for federal land managers. Canadian 169: 853-860. http://dx.doi.org/10.1007/ Wildlife Service, Environment Canada, s00442-011-2227-2 33pp. ISBN 0-662-24081-2. Available at: MFLNRO. 2012. The draft Peace northern http://nawcc.wetlandnetwork.ca/pubs.html. caribou plan. Ministry of Forests, Lands and Accessed June 2014. Natural Resource Operations, Victoria, Brit- Mahoney, S.P. & Virgl, J.A. 2003. Habitat ish Columbia, Canada. selection and demography of a nonmigra- Moilanen, A., van Teefflen, A.J.A., Ben- tory woodland caribou population in New- Haim, Y. & Ferrier, S. 2009. How much foundland. — Canadian Journal of Zool- compensation is enough? A framework for ogy 81: 321-334. http://dx.doi.org/10.1139/ incorporating uncertainty and time dis- z02-239 counting when calculating offset ratios for Maron, M., Hobbs, R.J., Moilanen, A., Mat- impacted habitat. — Restoration Ecology 17: thews, J.W., Christie, K., Gardner, T.A., 470-478. http://dx.doi.org/10.1111/j.1526- Keith, D.A., Lindenmayer, D.B. & McAlp- 100X.2008.00382.x ine, C.A. 2012. Faustian bargains? Restora- NEB (National Energy Board). 2011. Reason tion realities in the context of biodiversity for Decision, NOVA Gas transmission Ltd. offset policies. — Biological Conservation GH-2-2011. Issued January 2011. Available 155: 141-148. http://dx.doi.org/10.1016/j. at: https://docs.neb-one.gc.ca/ll-eng/llisapi. biocon.2012.06.003 dll?func=ll&objId=665172&objAction=bro

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 118 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 wse. Accessed April 2014. odology and example case studies. — Can. NEB. 2012a. Reason for Decision, NOVA Gas Manuscr. Rep. Fish. Aquat. Sci. 2729: 124 pp. transmission Ltd. GH-2-2011. Issued Feb- Pickard, J. 2007. Predator-proof fences for ruary 2012. Available at: https://docs.neb- biodiversity conservation: some lessons from one.gc.ca/ll-eng/llisapi.dll?func=ll&objId=7 dingo fences. In: Dickman, C.R., Lunney, 93570&objAction=browse. Accessed April D., & Bergin, S. (Eds.), Animals of Arid Aus- 2014. tralia: Out on their own. Royal Zoological So- NEB. 2012b. Reason for Decision, NOVA ciety of New South Wales, Sydney, Australia, Gas transmission Ltd. GH-004-2011. Is- pp. 197-207. http://dx.doi.org/10.7882/ sued July 2012. Available at: http://gateway- FS.2007.050 panel.review-examen.gc.ca/clf-nsi/dcmnt/ Pople, A.R., Grigg, G.C., Cairns, S.C., Beard, rcmndtnsrprt/rcmndtnsrprtvlm2ppndx-eng. L.A. & Alexander, P. 2000. Trends in num- html#ppndx1. Accessed April 2014. bers of red kangaroos and emus on either NEB. 2013. Northern Gateway Project Joint side of the South Australian dingo fence: Review Panel, Appendix 1: The Panel’s evidence for predator regulation? — Wild- Conditions. Available at: http://gateway- life Research 27: 269-276. http://dx.doi. panel.review-examen.gc.ca/clf-nsi/dcmnt/ org/10.1071/WR99030 rcmndtnsrprt/rcmndtnsrprtvlm2ppndx-eng. Poulton, D.W. 2014. Biodiversity offsets: A html#ppndx1. Accessed April 2014. primer for Canada. Prepared for Sustainable Norrdahl, K., Klemola, T., Korpimaki, E. & Prosperity and the Institute of the Environ- Koivula, M. 2002. Strong seasonality may ment. 58pp. attenuate trophic cascades: vertebrate preda- Quétier, F. & Lavorel, S. 2011. Assessing tor exclusion in boreal grassland. — Oikos equivalence in biodiversity offset schemes: 99: 419-430. http://dx.doi.org/10.1034/ Key issues and solutions. — Biological Con- j.1600-0706.2002.12025.x servation 144: 2991-2999. http://dx.doi. Norton, D.A. 2009. Biodiversity offsets: two org/10.1016/j.biocon.2011.09.002 New Zealand case studies and an assessment Quétier, F., Regnery, B. & Levrel, B. 2014. framework. — Environmental Management No net loss of biodiversity or paper offsets? 43: 698-706. http://dx.doi.org/10.1007/ A critical review of the French no net loss s00267-008-9192-5 policy. — Environmental Science and Policy NRC (National Research Council). 1997. 38: 120-131. http://dx.doi.org/10.1016/j. Wolves, bears, and their prey in Alaska: Bio- envsci.2013.11.009 logical and social challenges in wildlife man- Quigley, J.T. & Harper, D.J. 2005. No net agement. National Academy Press, Washing- loss of fish habitat: A review and analysis of ton, D.C., USA. habitat compensation in Canada. — Envi- Pascoe, S., Wilcox, C. & Donlan, C.J. 2011. ronmental Management 36: 343-355. http:// Biodiversity offsets: A cost-effective in- dx.doi.org/10.1007/s00267-004-0114-x terim solution to seabird bycatch in fisher- Quigley, J.T. & Harper, D.J. 2006a. Compli- ies? — PLoS ONE 6: e25762. http://dx.doi. ance with Canada’s fisheries act: a field audit org/10.1371/journal.pone.0025762 of habitat compensation projects. — Envi- Pearson, M.P., Quigley, J.T., Harper, D.J. & ronmental Management 37: 351-366. http:// Galbraith, R.V. 2005. Monitoring and as- dx.doi.org/10.1007/s00267-004-0263-y sessment of fish habitat compensation and Quigley, J.T. & Harper, D.J. 2006b. Effec- stewardship projects: Study design, meth- tiveness of fish habitat compensation in

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 119 Canada in achieving no net loss. — Envi- http://dx.doi.org/10.7557/2.31.2.1994 ronmental Management 37: 351-366. DOI: Species at Risk Public Registry. 2014. Species 10.1007/s00267-004-0263-y. http://dx.doi. at Risk Public Registry, Government of Can- org/10.1007/s00267-004-0263-y ada. Available at: http://www.sararegistry. Rainey, H.J., Pollard, E.H.B., Dutson, G., gc.ca/default_e.cfm. Accessed April, 2014. Ekstrom, J.M.M., Livingstone, S.R., Tem- Stuart-Smith, A.K., Bradshaw, C.J.A., Bou- ple, H.J. & Pilgrim, J.D. 2014. A review tin, S., Hebert, D.M. & Rippin, A.B. 1997. of corporate goals of no net loss and net Woodland caribou relative to landscape pat- positive impact on biodiversity. — Oryx terns in northeastern Alberta. — Journal of 49: 232-238. http://dx.doi.org/10.1017/ Wildlife Management 61: 622-633. http:// S0030605313001476 dx.doi.org/10.2307/3802170 Saenz, S., Walschburger, T., Gonzalez, J.C., ten Kate, K., Bishop, J. & Bayon, R. 2004. Leon, J., McKenney, B. & Kiesecker, J. Biodiversity offsets: Views, experience, and 2013. Development by design in Colom- the business case. IUCN, Gland, Switzer- bia: making mitigation decisions consistent land and Cambridge, UK and Insight In- with conservation outcomes. — PLoS ONE vestment, London, UK. e81831. http://dx.doi.org/10.1371/journal. Van Ballenberghe, V. 2006. Predator control, pone.0081831 politics, and wildlife conservation in Alaska. Schneider, R.R., Hauer, G., Adarnowicz, — Alces 42: 1–11. W.L. & Boutin, S. 2010. Triage for conserv- Walker, S., Brower, A.L., Stephens, R.T.T. ing populations of threatened species: the & Lee, W.G. 2009. Why bartering bio- case of woodland caribou in Alberta. — Bio- diversity fails. — Conservation Letters 2: logical Conservation 143: 1603-1611. http:// 149-157. http://dx.doi.org/10.1111/j.1755- dx.doi.org/10.1016/j.biocon.2010.04.002 263X.2009.00061.x Scofield, R.P., Cullen, R. & Wang, M.2011. Webb, N.F., Allen, J.R. & Merrill, E.H. 2011. Are predator fences the answer to New Zea- Demography of a harvested population of land’s terrestrial faunal biodiversity crises? — wolves (Canis lupus) in west-central Alberta, New Zealand Journal of Ecology 35: 312-317. Canada. — Canadian Journal of Zoology 89: Serrouya, R.D. 2013. An adaptive approach 744-752. http://dx.doi.org/10.1139/z11- to endangered species recovery based on a 043 management experiment: reducing moose Wittmer, H.U., Serrouya, R., Elbroch, L.M. to reduce apparent competition with wood- & Marshall, A.J. 2013. Conservation strate- land caribou. Ph.D. thesis, University of Al- gies for species affected by apparent compe- berta, Edmonton. tition. — Conservation Biology 27: 254-250. Serrouya, R., Wittmann, M.J., McLellan, http://dx.doi.org/10.1111/cobi.12005 B.N., Wittmer, H.U. & Boutin, S. 2015. Using predator-prey theory to predict outcomes of broadscale experiments to reduce apparent competition. — The American Naturalist 185: 665-679. http://dx.doi. org/10.1086/680510 Smith, K.G. & Pittaway, L. 2011. Little Smoky woodland caribou calf survival enhancement project. — Rangifer 19: 97-102.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 120 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 121 This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 122 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 15th North American Caribou Workshop Whitehorse, Canada 12-16 May, 2014

Habitat restoration as a key conservation lever for woodland caribou: A review of restoration programs and key learnings from Alberta

Paula Bentham1 & Brian Coupal2 1 Golder Associates Ltd., 16820 107 Avenue, Edmonton, Alberta, Canada T5P 4C3 1 (Corresponding author: [email protected]). 2 Golder Associates Ltd., 102, 2535 - 3rd Avenue S.E., Calgary, Alberta, Canada T2A 7W5.

Abstract: The Recovery Strategy for the Woodland Caribou (Rangifer tarandus caribou), Boreal Population in Canada (EC, 2012), identifies coordinated actions to reclaim woodland caribou habitat as a key step to meeting current and future caribou population objectives. Actions include restoring industrial landscape features such as roads, seismic lines, pipelines, cut-lines, and cleared areas in an effort to reduce landscape fragmentation and the changes in caribou population dynamics associated with changing predator-prey dynamics in highly fragmented landscapes. Reliance on habitat restoration as a recovery action within the federal recovery strategy is high, considering all Alberta populations have less than 65% undisturbed habitat, which is identified in the recovery strategy as a threshold providing a 60% chance that a local population will be self-sustaining. Alberta’s Provincial Woodland Caribou Policy also identifies habitat restoration as a critical component of long-term caribou habitat management. We review and discuss the history of caribou habitat restoration programs in Alberta and present outcomes and highlights of a caribou habitat restoration workshop attended by over 80 representatives from oil and gas, forestry, provincial and federal regulators, academia and consulting who have worked on restoration programs. Restoration initiatives in Alberta began in 2001 and have generally focused on construction methods, revegetation treatments, access control programs, and limiting plant species favourable to alternate prey. Specific treatments include tree planting initiatives, coarse woody debris management along linear features, and efforts for multi-company and multi-stakeholder coordinated habitat restoration on caribou range. Lessons learned from these programs have been incorporated into large scale habitat restoration projects near Grande Prairie, Cold Lake, and Fort McMurray. A key outcome of our review is the opportunity to provide a unified approach for restoration program planning, best practices, key performance indicators, and monitoring considerations for future programs within Canada.

Key words: Alberta; federal recovery strategy; habitat restoration; woodland caribou.

Rangifer, 35, Special Issue No. 23, 2015: 123-148 DOI 10.7557/2.35.2.3646

Introduction In 2012, the federal Recovery Strategy for the reclaim woodland caribou habitat as a key step Woodland Caribou (Rangifer tarandus caribou), to meeting caribou population and distribution Boreal Population in Canada was publicly re- objectives (EC, 2012). Actions include restor- leased and it described coordinated actions to ing industrial landscape features such as roads,

Figure 1. Natural regeneration of a typical conventional seismic line in the boreal forest. Photo courtesy of Brian Coupal.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 124 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 stakeholder groups and individual companies avoidance of habitat adjacent to human dis- in recent years, including restoration projects turbance (Dyer, 1999; Neufeld, 2006; Oberg, near Grande Prairie, Cold Lake, and Fort 2001). In addition, seismic lines were reported McMurray. The objectives of these initiatives to have very slow reforestation rates (Revel et have been to restore habitat on historical an- al., 1984; Osko and MacFarlane, 2000), with thropogenic footprint in an attempt to create slow tree regeneration attributed to root dam- intact habitat areas for caribou and/or to slow age from the original disturbance, compaction down predation rate as a result of the footprint. of the soil in tire ruts, insufficient light reach- Given a lack of formal guidelines on habitat ing the forest floor, introduction of competitive restoration objectives or techniques, as well seed mixes (i.e., plant seed mixes), drainage of as a lack of reporting on program learning’s, sites, and repeated disturbances (e.g., all terrain Golder Associates Ltd. (Golder) organized a vehicles) on seismic lines (MacFarlane, 1999 Restoration Workshop in Edmonton, Alberta, and 2003; Sherrington, 2003). Rehabilitation in June 2013. More than 80 participants from of existing anthropogenic disturbances within industry, government, academia, and consult- caribou range was expected to reduce the degra- ing attended the one day workshop to discuss dation of functional habitat over the long-term, caribou restoration efforts in Northern Alberta. with caribou no longer exhibiting avoidance of The intent of the workshop was to provide an the disturbance feature (e.g., Oberg, 2001). The opportunity to improve common understand- CRRP piloted techniques with the objectives of ing from on-the-ground restoration programs promoting revegetation of these features, while in terms of key performance indicators, suc- discouraging access for predator, primary prey, cesses, best practices and outcomes and to link and human use. the results of these programs back to provincial The CRRP was a multi-stakeholder group guidance on habitat restoration considerations. initiated and steered by the provincial govern- The workshop balanced learning and discus- ment agency Alberta Sustainable Resource De- sion with knowledge sharing presentations by velopment (ASRD), and the Boreal Caribou government and industry. Breakout groups Committee (BCC) (Dzus, 2001). Although the focused on a series of key questions regarding CRRP was not extended beyond 2007, the pro- restoration efforts. Here we outline how lessons ject did incorporate silviculture methods based learned from past restoration initiatives educate on knowledge of forestry treatments, focusing the objectives and techniques for implementa- on access control treatments and enhancing the tion of habitat restoration for current and fu- vegetation recovery rate of historical seismic ture restoration projects. lines, pipelines, and lease roads. Based on the outcome of treatments and learnings on linear Habitat restoration initiatives restoration, the CRRP prepared a Guidance A Caribou Range Restoration Project (CRRP) Document (CRRP, 2007a) which included was first established within Alberta in 2001 (Sz- recommended practices for implementing a korupa, 2002) in an effort to address growing habitat restoration program, from the planning concerns with the relationship between indus- through to the treatment stages. A monitoring trial development and declining local caribou protocol document for revegetation (unpub- populations. At that time, research from James lished) (CRRP, 2007b) was also prepared. Key (James, 1999) suggested wolves were gaining a learnings during the CRRP included recogni- predation advantage using linear features cre- tion that restoring linear development features ated by industry, and that indirect habitat loss is not equivalent to replanting a typical mono- for boreal caribou was occurring through the

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 126 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 Cody, 2013; Golder, 2015b), and Fort McMur- has been recommended as an area where direct ray (COSIA, 2014; OSLI, 2012a), Alberta. management activities, including access control to reduce repeated disturbance, and silviculture Existing knowledge preparations to address site deficiencies, should Conventional seismic lines, which are gener- be applied to set a line on a natural successional ally 6 to 8 m wide, have been reported to have trajectory (MacFarlane, 2003). very slow reforestation rates (Revel et al., 1984; Positive results for establishing native vegeta- Osko and MacFarlane, 2000; Lee and Boutin, tion on seismic lines and pipeline rights-of-way 2006). Tree regeneration along seismic lines is (ROWs) have been recorded using techniques influenced by the characteristics of the adjacent such as planting tree and shrub seedlings, and forests (e.g., site productivity, tree and shrub creating microsites by methods such as mound- species and heights) (Bayne et al., 2011), meth- ing that are conducive to seedling growth and od of clearing from the original disturbance, natural vegetation encroachment (DES, 2004; compaction of the soil from human use, insuf- CRRP, 2007b; Golder, 2010; 2011; 2012a; ficient light reaching the forest floor, mainte- OSLI, 2012a; Macadam and Bedford, 1998; nance of apical dominance from surrounding MacIsaac et al., 2004; Roy et al., 1999). Meas- stands, introduction of competitive species ures such as the use of coarse-woody debris such as graminoid dominated seed mixes, natu- (slash rollback) can address site condition issues rally poor drainage of sites and repeated distur- including competition from non-target or un- bances (e.g., all-terrain vehicles, animal brows- desired plant species, erosion, frost, and heat or ing, repeated exploration) (Revel et al., 1984; moisture deficiencies, as well as to create micro- MacFarlane, 1999; 2003; Sherrington, 2003; sites for germination (CRRP, 2007b; Pyper and Lee and Boutin, 2006). The slow pace of re- Vinge, 2012; Vinge and Pyper, 2012). covery of plant communities on seismic lines Transplanting native vegetation has been at-

Figure 3. Alder shrub seedling planting on a pipeline ROW after 1 growing season. Photo courtesy of Enbridge Pipelines (Athabasca).

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 127 tempted along seismic lines and pipelines but is visual screening vary considerably between challenging to implement on a large scale due negligible and high effectiveness in controlling to the inconsistent availability of vegetation human access within caribou ranges (CLMA suitable for transplant, the potential for degra- and FPAC, 2007). Effectiveness of access con- dation of neighboring vegetation communities trol measures are dependent on suitable place- if transplants are sourced from adjacent stands, ment (e.g., placed to prevent detouring around approval requirements to move vegetation, and access control point), enforcement, and public less than ideal storage conditions for plant ma- education of the intent of the access control, terials due to weather. Other treatments such as which facilitates respect of the control meas- seeding and seedling planting have been shown ures (AXYS, 1995). Excavator mounding is a to be more successful and predictable in com- well-researched and popular site preparation parison (Golder, 2012b). technique in the silviculture industry (Macad- Both natural revegetation and seedling am and Bedford, 1998; Roy et al., 1999; Ma- planting initiatives on both seismic lines and cIsaac et al., 2004). Mounding has been found pipelines have benefited from minimal dis- to discourage human access such as off road turbance construction during frozen ground vehicular use and also creates microsites that conditions that reduce or avoid grubbing and improve vegetation establishment (CLMA and grading and minimize disturbance to the duff FPAC, 2007). Physical access control measures layer (e.g., DES, 2004; TERA, 2011; 2012; provide short-term solutions to manage access Enbridge, 2010; TCPL, 2014). and allow for natural regeneration (Golder, The ability of linear developments to regen- 2009). It has been suggested that once linear erate to native species is affected considerably features have regenerated to a pole sapling or by human use. Oberg (2001) identified that young forest structural stage, they no longer recovery of conventional seismic lines within facilitate vehicular access (Sherrington, 2003). the foothills to functioning caribou habitat oc- A number of the techniques used to block curs within 20 years following disturbance in human access use of regenerating industrially west-central Alberta. Within a boreal caribou disturbed features also contribute to initiatives area, seismic lines that were allowed to regen- to block line–of–sight. Short-term manage- erate naturally achieved an average height of 2 ment for access and line-of-sight blocking is m, across all boreal vegetation types, within 20 understood to lead to long-term access control to 25 years, if the line had not undergone a re- by providing the necessary conditions for the peated disturbance (e.g., re-cleared to ground disturbance to regenerate to natural vegetation level for winter access or exploration use). The conditions (CLMA and FPAC, 2007). Expe- average age of trees on the revegetated seismic diting growth of visual barriers along linear lines was only 10 years, suggesting sites that are features can be achieved by concentrating rec- continually disturbed or re-cleared by seismic lamation efforts on productive upland habitats, exploration or vehicular access take longer to since tree and shrub (e.g., alder which is less regenerate. Restoration efforts are also negated palatable for prey species) species grow more when human use destroys or damages seedlings quickly on these sites compared to lowland after planting (Enbridge, 2010; Golder, 2011; sites. On deciduous and mixedwood upland 2012a). sites, encouraging deciduous tree species and Subjective expert ratings suggest that ef- shrub growth is important to quickly establish fectiveness of access control measures such as visual and physical barriers in the short-term. gates, berms, mounding, slash rollback, and Tree-felling has recently been applied through

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 128 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 the Cenovus Energy Linear Deactivation examined vegetation trajectories associated (LiDEA) project in northeastern Alberta and with the natural recovery of linear features over early results suggest it is effective in providing time. The values provided in CEMA (2012) an immediate access control through remote are based on practitioner opinion as well as es- camera monitoring (Cenovus, 2014). Although timates based on ecosite and tree species growth regeneration of conifer species is the endpoint potential. The revegetation matrix can be used for caribou habitat use and minimizes habitat to simulate how vegetation height may change creation for other prey species, conifer species over time (CEMA, 2012). growth rates are slower than the growth rates While there has been some effort to assess of deciduous species. Faster growth rates pro- wildlife use of regenerating seismic lines (e.g., vide for access control and line-of-sight barriers Bayne et al., 2011) and reclaimed areas (e.g., more quickly (DES, 2004). Recent field trials Hawkes, 2011), few researchers have docu- suggest that planting shrubs along with conifer mented the relationship between natural habi- tree species may allow trees to grow healthier, tat recovery and wildlife responses to recovery faster and with less competition for nutrients with respect to caribou. A pilot study to meas- and water from fast-growing grasses than when ure the effects of revegetating linear distur- planted without shrubs (OSLI, 2012a). Plant- bances on wildlife use and mobility collected ing shrubs may also provide important habitat data for a group of predators (i.e., cougar, wolf, benefits for wildlife, compared to only plant- coyote, lynx, grizzly and black bears) and prey ing tree seedlings, by providing hiding cover (i.e., moose, deer and caribou) (Golder, 2009). (Bayne et al., 2011). Results indicated that revegetated seismic lines The OSLI program (now COSIA) includes with a minimum of 1.5 m of consistent vegeta- on-going studies to determine what the most tion regrowth were preferred by both predator efficient vegetation introduction techniques are and prey species (including caribou) compared for peatland areas, such as planting frozen seed- to open, low (< 1.5 m vegetation) vegetation lings in the winter instead of summer planting, control lines. The line-of-sight measured on and whether to use seed or seedlings, depend- the revegetating lines was typically less than 50 ent on site conditions and other variables. The m. In general, control lines were used primar- OSLI/COSIA program also involves voluntary ily for travel by both predators and prey spe- restoration of legacy footprint within caribou cies. Human use was primarily limited to the critical habitat in an effort to restore large, late control lines. Golder (Golder, 2009) suggested seral stage patches of caribou habitat to increase that moose and deer may have been attracted habitat intactness and discourage corridor use to the revegetated lines for forage availability (OSLI, 2012a). and perceived cover protection. The preference The Government of Alberta has not provid- for regenerating seismic lines by wolves may be ed a manual for reclamation that can be uti- explained as a response to increased prey use lized for developing silvicultural prescriptions of these lines. More recently, pre-treatment for large scale habitat restoration programs. (Dickie, 2015) and post-treatment wolf move- However, a revegetation matrix was developed ment data is being gathered through the Uni- by Alberta Environment and Parks and pub- versity of Alberta to look at the effectiveness of lished within the Cumulative Effects Manage- line-blocking within the Cold Lake region of ment Association (CEMA) document ‘Stony Alberta. Wolves selected conventional seismic, Mountain 800 Linear Footprint Management pipelines, railway, roads, trails, and transmis- Plan’ (CEMA, 2012). The revegetation matrix sion lines, but did not select low-impact seis-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 130 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 velopment feature in a bog or fen is least likely ration priority areas mapping, available upon to regenerate naturally; and a narrow (<3m) request, to help direct where restoration efforts line has improved regeneration over a wider should be focused (D. Hervieux pers. comm., line (van Rensen et al., 2015). Natural regen- 2013). eration to 3 m vegetation height is inversely re- Learnings from existing restoration programs lated to terrain wetness, line width, proximity were presented and included an awareness that to roads as a proxy for human use of lines, and not all linear disturbances are equal and that lowland ecosites (fens, bogs) (van Rensen et al., restoration on linear disturbances differs from 2015). Areas adjacent to major rivers illustrate silviculture prescriptions applied to cutblocks, high probability of regeneration. Overall, ter- given the higher variability in site conditions. rain wetness and the presence of fens has the As a result, the toolbox for restoration treat- stongest negative effect on natural regenera- ments needs to consider a number of variables, tion (van Rensen et al., 2015). Lack of time se- in particular the lack of a seed bed and mineral quence recording for regenerating seismic lines layer for plant growth and compaction. and other linear developments reduces the abil- It was discussed, based on previous initia- ity to estimate natural rates and types of veg- tives, that prior to applying treatments on the etation recovery, however predictive models do ground, linear feature (and polygon) invento- exist (e.g., van Rensen et al., 2015). ries of the existing footprint are the first steps in designing a restoration program. Collecting in- Workshop results ventories help ensure an efficient allocation of Although the federal Recovery Strategy (EC, resources committed to habitat restoration. For 2012) for boreal caribou describes the require- example, a pilot habitat restoration program in ment for habitat restoration, it is not clear what west-central Alberta approximately four town- defines successful habitat restoration. During ships in size and another pilot northwest of the workshop participants discussed a proposed Cold Lake, approximately eight townships in definition of habitat restoration: Restored (dec- size, reported that approximately 85% of lin- ades) - disturbed caribou range is returned to ear features observed were already on a natural functional habitat that can support self-sustain- recovery trajectory and revegetation treatments ing caribou population without ongoing inter- were not recommended (CRRP, 2007c; Golder, vention (e.g., predator control). Participants 2010) (Fig. 1). Inventories are gathered using identified that habitat restoration needs to con- remote sensing to spatially map linear distur- sider spatial and temporal scales, trajectories, as bances and the level of natural regrowth (e.g., well as predator/prey dynamics. van Rensen et al., 2015). In addition to the During the restoration workshop, a number amount of natural regrowth, field truthing of of the presentations discussed key elements of candidate treatment sites is completed to docu- program planning, including authorization to ment detailed ground conditions. Data is col- implement restoration measures. Government lected on classifying the type(s) of disturbance of Alberta representatives acknowledged that (roads are considered severe disturbance where- an approval process needs to be developed that as a cutline is often minimal disturbance), level provides a consistent approach to authorize im- of human (e.g., all-terrain vehicle) and wildlife plementation of restoration treatments on his- (game trails) use, width and orientation of a torical seismic lines, and that development of line (impacts light penetration and moisture the process is under discussion. As well, Alberta level), compaction level (impacted from con- Environment and Parks presented draft resto- struction practices), soil mineral layer (nutri-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 131 ents) and microsite availability, adjacent ecosite ibou recovery management measures indicate phase / forest attributes (very wet to very dry, that habitat restoration is linked to improving upland/transitional/lowland), coarse woody caribou population projections, the feasibility debris level/availability/fuel loading considera- (cost, large scale application, rate of restoration tions from a fire management perspective, and as compared to rate of ongoing development historical seeding practices which often results pressure) and predicted outcomes of restoration in high levels of competing vegetation to coni- activities remain highly uncertain (ALT, 2009; fer seedlings (Vinge, 2013; CRRP, 2007c). Wilson et al., 2010). This uncertainty includes During the remote sensing and ground truth- the time lag required to recover disturbed areas ing of site conditions, treatment sites and pre- to effective habitat to support self-sustaining scriptions are finalized and often located into caribou populations. priority areas for restoration and to areas where Based on monitoring of revegetation of ex- human access control treatments will prevent isting disturbances, it is expected that vegeta- repeated use. This ensures that the ‘right lines tion recovery of disturbed areas will take dec- for restoration’ are selected. For large scale res- ades, with or without intervention. To address toration programs, future development plans in the time lag associated with natural revegeta- the area (e.g., forestry harvest plans, lease areas, tion of linear features (Fig. 1), industries and development footprints, pay depth to bitumen, governments have built a toolbox of habitat etc.) (ALT, 2009), provincial priority areas, as restoration treatment best practices, focused on well as a focused plan to create large, contigu- establishing vegetation similar to adjacent for- ous intact habitat areas should be considered. est communities, creating line-of-sight breaks, Restoration program development considers and discouraging human, predator and prima- not only a planning scale, but a tactical scale ry prey usage of linear features. The treatment with efficiency of operational implementation best practices, including their objectives and considerations. For example, the OSLI/CO- recommended specifications, are summarized SIA program used a modelling approach called in Table 1. Inclusion of a reference in Table Landscape Ecological Assessment and Planning 1 was based on if the results of implemented (LEAP) to enhance efficiency in bringing land- treatments were successful, for the objective scape data sources together to assess and de- outlined (e.g., if a treatment met the objec- velop restoration scenarios, strategic to tactical tive of establishing vegetation along a segment implementation plans, and monitoring plans of linear feature where vegetation did not ex- (OSLI, 2012a). ist prior to the treatment). Specifications and considerations for each treatment are also pro- Restoration toolbox vided, based on positive evidence of success. The objectives of past and current habitat res- The treatments designed to promote revege- toration programs for caribou have been to tation of linear features are intended to address restore habitat on existing anthropogenic foot- micro-site deficiencies, and are well recognized print to create large contiguous habitat patches silvicultural practices modified for linear fea- that can support self-sustaining caribou popu- ture application. When implemented properly, lations with historical predator-prey encounter these practices will meet their objective of es- rates. This objective implies that habitat resto- tablishing vegetation. Additional monitoring ration must address revegetation, predator and on site preparation treatments such as mound- primary prey access, predator efficiency, and ing and spreading woody debris are currently forage for primary prey species. Although the being researched in NE Alberta to determine federal recovery strategy and analyses to set car-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 132 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 their efficacy in achieving the goal of discourag- ested areas should be consulted prior to de- ing predator and primary prey usage of linear termining specifications and design of a tree features. Although long-term results are not and shrub seedling planting program. For ex- yet available, preliminary results indicate these ample, the Government of Alberta guidelines methods are achieving this objective (Cenovus, for forest reclamation in the oil sands region 2014). (AENV, 2010) specify ranges of seedling plant- Implementing practices to reduce a new ing densities that vary by the site type and spe- project’s impacts at the construction phase will cies planted. These guidelines are not specific reduce the need for, and the amount of, habi- to caribou habitat restoration, and may need tat restoration required following construction. to be modified with consideration to measure- Construction practices which enhance the abil- able objectives for caribou habitat restoration. ity of a site to restore naturally will reduce the The Science and Community Environmental level of effort and cost of site preparation (e.g., Knowledge branch of the Government of Brit- mounding) and tree/shrub planting over the ish Columbia has recently commissioned the entire project. For example, three practices that creation of a Boreal Caribou Habitat Restora- can be implemented during or immediately fol- tion Operational Toolkit for British Columbia lowing the construction phase of a project are that contains reclamation recommendations minimizing line width (e.g., low impact seis- specific to caribou ranges, focusing on linear mic <3m width; Dickie, 2015), minimal dis- feature restoration (Golder, 2015a). Considera- turbance vegetation removal (e.g., DES, 2004), tions for determining species, planting density and controlling off road vehicle access (Revel et and locations of planting should include site al., 1984). type (dry, moist/poor, moist/rich, wet rich), Since the ability of cleared areas to quickly surrounding vegetation community, distur- regenerate to native species following construc- bance level (high with no LFH layer, low with tion is affected considerably by human use, LFH layer intact), coarse woody debris level, applying human access control measures, with and site preparation (Vinge, 2013). effectiveness monitoring, along linear features A critical component of a successful habitat should occur immediately following construc- restoration program is protection of the treat- tion. Woody debris treatments, excavator ment locations from disturbance. Sites that mounding, berms, tree-felling and steel gates have been developed using methods that pro- are treatment types that are effective immedi- mote speedy natural revegetation or planted ately and can be considered; but require moni- to enhance revegetation, line-of-sight break toring. The type of control can be determined locations, or access control treatments should by the amount of expected human use at the be clearly marked in the field and protected location, width of the linear feature, ecosite with physical barriers if necessary. For example, phase, and topography. For example, a seismic seedlings planted on an upland graded site can line seldom used by humans, crossing a newly be damaged or destroyed from human use of constructed pipeline ROW, may be treated with the ROW unless they are protected by a suf- excavator mounding and planted seedlings (e.g, ficient layer of coarse woody material. Fig. 4), while a pipeline crossing of a winter ac- Monitoring cess road, well-used by humans, may be treated with a greater density of excavator mounding, Monitoring of construction practices, the suc- planted seedlings, along with a steel gate. cess of treatments to establish vegetation, lines Reclamation criteria and guidelines for for- undergoing natural revegetation trajectories and the effectiveness of access control methods

Figure 6. Lease road after treatment with mounding, tree-felling, tree-bending, and tree transplanting. Photo courtesy of MEG Energy. is necessary for any habitat restoration program. within restoration priority areas. During the Monitoring programs should be linked to res- workshop participants discussed monitoring toration objectives and measureable targets for programs and the overall consensus was that the program to determine success or oppor- there is a need for consistent design in what’s tunities for adaptive management measures being measured, that there should be near term

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 134 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 variables measured to determine if a site is on (Pembina Institute, 2012). For any new project trajectory (with consideration of revised recla- planned or project expansion within a caribou mation certificate criteria); successional trajec- range in northeast Alberta, under the SARA, tories or milestones should be determined and the new project footprint could be deemed de- monitored against; and there is a disconnect struction of critical habitat for the species. As between the end goal of caribou population such, planning for approved future develop- lambda and the desire to consider habitat re- ment projects within caribou ranges will need stored as early as possible. Adaptive manage- to consider the entire mitigation hierarchy: ment on restoration programs will need to be avoidance of caribou range; minimizing im- implemented by adjusting and/or supplement- pacts through project planning, utilizing the ing restoration measures, where warranted, to least footprint necessary, overlapping land uses achieve the objectives of the habitat restoration (e.g., coordinated access planning, integrated initiatives. Monitoring programs will need to land use management planning); planning out consider a number of response metrics includ- a comprehensive habitat restoration plan; and ing the wildlife response to restoration (multi- include off-sets to address residual project ef- species including caribou population trends, fects due to the time lag and uncertainty around wolf movement and behavior, and primary habitat restoration success to caribou recovery. prey population response) and the site level Habitat restoration has been highlighted response both short-term and long-term with within the federal recovery strategy, as well successional trajectories or milestones devel- as within the Alberta Caribou Policy (GOA, oped (Cody, 2013). Given the relatively short 2011) as a critical component of long-term time period since large scale habitat restoration caribou habitat management. Given the cur- programs have begun to be implemented, field rent range condition for caribou in Canada, re- results are currently in the early stages of recent National Energy Board and Federal Joint porting regarding the success of caribou habitat Review Panel conditions for pipeline ROW restoration methods meeting their objectives. occurring within caribou ranges have included Monitoring outcomes will inform adaptive preparing, implementing and monitoring Car- management, allowing for modification of un- ibou Habitat Restoration Plans (e.g., NGTL, successful measures to continuously improve, 2014a; 2014b). These Caribou Habitat Resto- and are an important means of addressing un- ration Plans provide details on the objectives certainty. of restoration plans, the criteria used to identify potential habitat restoration sites, the pro- Discussion cess to identify restoration actions to be used At the national scale, Alberta’s woodland cari- at different types of sites, quantifiable targets bou are among the least viable in Canada (EC, and performance measures that will be used to 2011). Under the Species At Risk Act, in 2012 evaluate the effectiveness of restoration meas- the federal government released its recovery ures to offset impacts to habitat, as well as a strategy for woodland caribou, with a clearly follow-up monitoring program (NEB, 2013). outlined habitat threshold to meet critical hab- Long-term vegetation removal and the time-lag itat levels (EC, 2012). In four caribou ranges associated with vegetation re-establishment to in northeastern Alberta underlain by oil sands suitable caribou habitat are considered residual deposits, on average only 24% of caribou habi- effects and are to be addressed with habitat off- tat remains undisturbed, far below the recov- set measures for caribou (NEB, 2013). ery plan target of 65% undisturbed habitat Although habitat restoration activities have

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 135 moved from pilot projects beginning in 2001 monitoring programs (Golder, 2013). to large scale project implementation since the The driver to implement large-scale habi- release of the recovery strategy, some caution- tat restoration programs has been to lower ary details need to be considered. First, there is the anthropogenic footprint within caribou currently no direct link to indicate that imple- ranges, and to address how caribou, wolves mented restoration treatments are having a pos- and primary prey species utilize habitat with- itive effect on caribou populations. Although in restored areas. Although we have identified modelling scenarios of management options the planning and physical measures that can for caribou indicate that restoration of habitat be implemented for a restoration program to should have benefits in the long-term by con- begin restoring caribou habitat following contributing to the restoration of large contiguous struction or along historical linear features, it is habitat patches that are preferred by caribou unreasonable to directly associate local caribou (e.g., ALT, 2009), additional management population trends with these programs due to measures must be applied by governments to the time lag to grow vegetation; as well as the address the proximate cause of caribou de- other factors contributing to these population clines. Specifically, governments must look to trends, specifically the effects of apparent com- implement immediate population manage- petitioninduced mortality on secondary prey ment of predators with effective habitat con- such as boreal caribou (DeCesare et al., 2010; servation measures (Hervieux et al., 2014) and Hervieux et al., 2014), and the current rate of primary prey (CAPP, 2012). It has been noted development. Monitoring and adaptive man- that industry actions and planning around agement of the restoration toolkit measures, minimizing and eliminating project footprints and the wildlife response to these measures, will be of no value if caribou populations are will be a critical element of industry led habitat not stabilized through aggressive wildlife (i.e., restoration programs. predator and alternate prey) management and long-term habitat conservation. It is recognized Acknowledgments that the full benefits of habitat recovery will not This paper includes a summary of restoration be realized for decades because there is a 30 to programs and studies which have been con- 50 year lag time following reclamation before ducted in Alberta since 2001. We would like re-establishing vegetation becomes old enough to acknowledge the efforts of those collabora- to be considered low quality for other prey, and tors involved in the CRRP for initiating habitat suitably old to be used by caribou (ALT, 2009). restoration treatments on linear disturbances as At a minimum, predator management through well as to those industry collaborators and gov- wildlife control will need to be continued for ernment partners who have been involved in this entire lag period (ALT, 2009). Intuitively, larger scale projects in northern Alberta. Tim extirpation risk of local herds will be reduced Vinge, Alberta Environment and Sustainable if habitat restoration begins as soon as possible Resource Development, provided guidance on (CAPP, 2012). Lastly, there is not a clear un- the restoration toolbox available to industry. derstanding of the desired objectives provided The Historic and Current Habitat Restoration by regulators regarding landscape level habitat Initiatives Summary, was originally prepared restoration programs. With no official frame- for, and is shared by, the Ministry of Forests, work, legislation or best practices within the Lands and Natural Resource Operations, Fish provincial jurisdiction, it is difficult to imple- Wildlife and Habitat Management Branch, of ment consistent caribou habitat restoration and the Government of British Columbia.

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 136 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 References try/Hazard%20Assess%20%20Abate%20 AXYS (AXYS Environmental Consulting Guidance%20Doc%20FINAL%20 Ltd.). 1995. A Compendium of Physi- with%20all%20Links%20April%202012. cal Access Control Measures for Roads and pdf. Accessed: September 15 2014. Other Rights-of-Way. March 1995. A com- BC MFR (BC Ministry of Forests and Range). ponent of the Access Management Initiative 1998. Provincial Seedling Stock Type Selec- in Northeastern BC. 27 pp. + Appendices. tion and Ordering Guidelines. BC Ministry AENV (Alberta Environment). 2010. Guide- of Forests, Forest Practices Branch. Available lines for Reclamation to Forest Vegetation in at: http://www.for.gov.bc.ca/hfp/publica- the Athabasca Oil Sands Region, 2nd Edition. tions/00078/. Accessed: February 27 2014. Prepared by the Terrestrial Subgroup of the BC MFR. 2014. Microsite Planting Presenta- Reclamation Working Group of the Cumu- tions. Available at: http://www.for.gov.bc.ca/ lative Environmental Management Asso- hfp/training/00009/microsite.htm. Ac- ciation, Fort McMurray, Alberta. December cessed: February 27 2014. 2009. Updated June 2011. BC MOF (BC Ministry of Forests). 2000. ALT (Athabasca Landscape Team). 2009. Extension Note: Opportunities for Improve- Athabasca Caribou Landscape Management ments to Reforestation Success. Prepared by Options Report. 99 pp. Marek J. Krasowski, Research Branch, Brit- Archuleta, J.G. & Baxter, E.S. 2008. Subsoil- ish Columbia Ministry of Forests, and Ron- ing Promotes Native Plant Establishment on ald J.F. Elder, R.J.F. Elder Forestry Consult- Compacted Forest Sites. — Native Plants ing. January 2000. 15 pp. 9: 117-122. http://dx.doi.org/10.2979/ CAPP (Canadian Association of Petroleum NPJ.2008.9.2.117 Producers). 2012. Athabasca Caribou Project Bayne, E., Lankau, H. & Tigner, J. 2011. Industry-Led Management Options. Submit- Ecologically-based criteria to assess the impact ted to: Canadian Association of Petroleum and recovery of seismic lines: the importance of Producers Species Management Committee. width, regeneration, and seismic line density. Prepared by: Golder Associates Ltd. 22 pp. Environmental Studies Research Funds Re- Cenovus (Cenovus Energy). 2013. Woodland port No. 192. Edmonton, Alberta. 102 pp. Caribou Mitigation and Monitoring Plan: Website: http://www.esrfunds.org/pdf/192. Cenovus FCCL Ltd. Christina Lake Ther- pdf. Accessed July 2012. mal Project (Approval 48522-01-00). Feb- BCFS (British Columbia Forest Service). ruary 22 2013. 1998. Silviculture Note 16 - Forest Floor Cenovus . 2014. Presentation given to mem- Planting: a Discussion of Issues As They Re- bers of Regional Industry Caribou Collabo- late to Various Site-Limiting Factors. Forest ration. December 10 2014. Site Management Section, Forest Practices CRRP (Caribou Range Restoration Project). Branch. March 25, 1998. Victoria, BC. 11 2007a. Guidelines for Planning and Imple- pp. mentation. Unpublished document created BC MFLNRO (BC Ministry of Forests, for the West Central Alberta Petroleum Pro- Lands, and Natural Resource Operations). ducers Group, Canadian Association of Pe- 2012. A guide to fuel hazard assessment and troleum Producers and Environment Cana- abatement in British Columbia. Wildfire da. September 19 2007. Management Branch. Available at: http:// CRRP. 2007b. Caribou Range Restoration bcwildfire.ca/Industry_Stakeholders/Indus- Project: Permanent Sample Plot Manual for

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 137 the Caribou Range Restoration Project in port Year 2 (2003). Prepared for: Canadian Alberta, March 2, 2007. Draft Unpublished Natural Resources Ltd. 18 pp. + Appendices. Document. Dickie, Melanie. 2015. The Use of Anthropo- CRRP. 2007c. Little Smoky Caribou Habi- genic Linear Features by Wolves in North- tat Restoration Project: Implementation of eastern Alberta. — A thesis submitted in Field Work-Plan. Unpublished document partial fulfillment of the requirements for prepared by Brian Coupal and John Ward. the degree of Master of Science in Ecology Prepared for Suncor Energy, ConocoPhillips Department of Biological Sciences, Univer- Canada, Alberta Newsprint Company and sity of Alberta. Canadian Forest Products. Dyer, S.J. 1999. Movement and Distribution CLMA (Caribou Landscape Management of Woodland Caribou (Rangifer tarandus Association) and FPAC (Forest Products caribou) in Response to Industrial Devel- Association of Canada). 2007. Audit of opment in Northeastern Alberta. — M.Sc. Operating Practices and Mitigation Measures Thesis. University of Alberta. 106 pp. Employed within Woodland Caribou Ranges. Dzus, E. 2001. Status of the Woodland Caribou Prepared for: CLMA and FPAC. Prepared (Rangifer tarandus caribou) in Alberta. Al- by: P. Bentham, Golder Associates, Edmon- berta Environment, Fisheries and Manage- ton, AB. 135pp + appendices. ment Division, and Alberta Conservation Cody, Michael. 2013. Restoration Efforts Association, Wildlife Status Report No. 30, in Northern Alberta: Knowledge Sharing. Edmonton, Alberta. 47 pp. Cenovus Linear Deactivation (LiDEA) Pro- Enbridge (Enbridge Pipelines (Athabasca) jects Update. June 6 2013 Golder Associates Inc.). 2010. Waupisoo Pipeline Project: Restoration Update Workshop. 2010 Green Area Post-Construction Recla- COSIA (Canadian Oilsands Innovation Al- mation Assessment. Submitted to Alberta liance). 2014 internet site. Caribou Habitat Sustainable Resource Development. Restoration. Available online at: http://www. EC (Environment Canada). 2011. Scientific cosia.ca/initiatives/land/caribou-habitat-res- review for the identification of critical habitat toration. Accessed on January 13 2014. for woodland caribou (Rangifer tarandus cari- CEMA (Cumulative Environmental Manage- bou), Boreal population, in Canada. August ment Association). 2012. Stony Mountain 2008. Ottawa: Environment Canada. 72 pp 800 Linear Footprint Management Plan. plus 80 pp Appendices Process Guidelines Revised Final Draft. Pre- EC. 2012. Recovery Strategy for the Woodland pared by Compass Resource Management. Caribou (Rangifer tarandus caribou), Boreal DeCesare, N., Hebblewhite, M., Robinson, population, in Canada. Species at Risk Act H. & Musiani, M. 2010. Endangered, ap- Recovery Strategy Series. Environment Can- parently: the role of apparent competition ada, Ottawa. xi + 138pp. in endangered species conservation. — FRI (Foothills Research Institute). 2014. Anim. Conserv. 13: 353-362. http://dx.doi. Habitat Restoration within Caribou Ranges: org/10.1111/j.1469-1795.2009.00328.x Interim Progress Report September 2014. DES (Diversified Environmental Services). Prepared for the Science and Community 2004. Canadian Natural Resources Ltd. Lady- Environmental Knowledge (SCEK) Fund. fern Pipeline d-87-H/94-H-1, British Colum- Prepared by Laura Finnegan and Doug Mac- bia to NE Section 20- Township 94-Range 12 Nearney. W6M, Alberta: Revegetation Monitoring Re- Golder (Golder Associates Ltd). 2005. Fire-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 138 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 bag Project 2004 Seismic Line Reclamation Golder. 2015a. Boreal Caribou Restoration Study Annual Report. Submitted to: Suncor Operational Toolkit for British Columbia. Firebag Inc. Submitted to: BC Science and Community Golder. 2009. Caribou Habitat Restoration Pilot Environmental Knowledge (SCEK) Fund’s Study. Prepared for Conoco Phillips Canada, Research and Effectiveness Monitoring Suncor Energy and Canadian Association of Board (REMB). Petroleum Producers. 41pp. Golder. 2015b. Habitat Enhancement Pro- Golder. 2010. Canadian Natural Resources gram. Prepared for: Canadian Natural Re- Limited: Primrose and Wolf Lake: Wildlife sources Primrose and Wolf Lake. Submitted Habitat Enhancement Program Develop- to Paul Kip, Canadian Natural Resources ment and Implementation. Submitted to: Ltd. Wolf Lake Office. May 2015. 28 pp + Canadian Natural Resources Ltd, Calgary, Appendices. AB. GOA (Government of Alberta). 2011. A Golder. 2011. Waupisoo Pipeline 2011 Seed- Woodland Caribou Policy for Alberta. June ling Monitoring Plot Summary. Memo 2011. ISBN: 978-0-7785-9477-2 (Printed submitted by Brian Coupal, Golder Associ- Version); 978-0-7785-9478-9 (Online Ver- ates, to Jeannette Gasser, Enbridge Pipelines sion). (Athabasca) Inc. December 16, 2011. 6 pp + Government of Alberta. 2013. Integrated Appendices. Standards and Guidelines – Enhanced Ap- Golder. 2012a. Interconnect Pipeline 2011 proval Process. Version 2. 85 pp. Available at: Seedling Monitoring Plot Summary. Memo http://esrd.alberta.ca/forms-maps-services/ submitted by Brian Coupal, Golder Asso- enhanced-approval-process/eap-manuals- ciates, to Paul Kip, Canadian Natural Re- guides/documents/EAP-IntegratedStand- sources Ltd. February 23, 2012. 6 pp + Ap- ardsGuide-Dec01-2013.pdf. pendices. Hawkes, V. 2011. Early successional wild- Golder. 2012b. Boreal Caribou Habitat Res- life monitoring on reclamation plots in the toration. Prepared for: Ministry of Forests, Athabasca Oil Sands Region. 9 pp. Lands and Natural Resource Operations. Hervieux, D. 2013. Personal communication Fish Wildlife and Habitat Management with Paula Bentham, Golder Associates. Branch, Government of British Columbia. Hervieux, D., Hebblewhite, M., DeCesare, Prince George, BC. 24 pp. N.J., Russell, M., Smith, K., Robertson, Golder. 2012c. Waupisoo Pipeline Project: S. & Boutin, S. 2013. Widespread declines Seedling Planting and Monitoring Final in woodland caribou (Rangifer tarandus cari- Report. Submitted by Golder Associates to bou) continue in Alberta. — Canadian Jour- Alberta Environment and Sustainable Re- nal of Zoology 91: 872-882. http://dx.doi. source Development, on behalf of Enbridge org/10.1139/cjz-2013-0123 Pipelines (Athabasca) Inc. December 27, Hervieux, D., Hebblewhite, M., Stepnisky, 2012. 16 pp + Appendices. D., Bacon M. & Boutin, S. 2014. Manag- Golder. 2013. Woodland Caribou Restoration ing wolves (Canis lupus) to recover threat- Workshop. Submitted to the Canadian As- ened woodland caribou (Rangifer tarandus sociation of Petroleum Producer’s Species caribou) in Alberta. — Can. J. Zool. 92: Management Committee. Report Number: 1029-1037. http://dx.doi.org/10.1139/cjz- 13-1334-0040-2000-2004. March 2014. 8 2014-0142 pp + Appendices. James, A.R.C. 1999. Effects of Industrial De-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License Rangifer, 35, Spec. Iss. No. 23, 2015 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com 139 velopment on the Predator-Prey Relationship the Enbridge Northern Gateway Project, Vol- Between Wolves and Caribou in Northeastern ume 2, Appendix 1. The Panel’s conditions. Alberta. Ph.D. Thesis Submitted to the Uni- Cat No. NE23-176/2013E-PDF versity of Alberta. Edmonton, AB. NEIPC (Northeast Invasive Plant Commit- Keim, J.L., DeWitt, P.D., Shopik, T., Fitzpat- tee). 2010. Peace – Liard Re-Vegetation rick, J. & Lele, S.R. 2014. Understanding Manual. Approved by the NEIPC Member- and mitigating the effects of linear features ship on April 28, 2010. Available at: http:// and snow condition on caribou predator- prrd.bc.ca/services/environmental/weed_ prey overlap in the Alberta Oil Sands. 15th control/documents/NEIPC_Reveg_manu- North American Caribou Workshop, White- al_PeaceLiard_April2010.pdf. Accessed No- horse, Yukon, May 14 2014. Oral presenta- vember 2014. tion. Neufeld, L.M. 2006. Spatial dynamics of wolves Lee, P. & Boutin, S. 2006. Persistence and de- and woodland caribou in an industrial forest velopmental transition of wide seismic lines landscape in west-central Alberta. M.Sc. The- in the western Boreal Plains of Canada. — sis, Wildlife Ecology and Management, De- Journal of Environmental Management 78: partment of Renewable Resources, Univer- 240-250. http://dx.doi.org/10.1016/j.jenv- sity of Alberta, Edmonton, Alberta. 155 pp. man.2005.03.016 NGTL (NOVA Gas Transmission Ltd). Macadam, A. & Bedford, L. 1998. Mound- 2014a. Final Caribou Habitat Restoration ing in the Sub-boreal Spruce Zone of West- Plan for the NOVA Gas Transmission Ltd. Central British Columbia: 8-year Results. — Northwest Mainline Expansion (Revised The Forestry Chronicle 74: 421-427. http:// from Original Filing January 2014). Pre- dx.doi.org/10.5558/tfc74421-3 pared for: TransCanada Ltd. Prepared by: MacIsaac, D.A., Hillman, G.R. & Hurdle, TERA Environmental Consultants. Submit- P.A. 2004. Alternative Silvicultural Systems ted to the National Energy Board in May for Harvesting and Regenerating Spruce-Dom- 2014. 63 pp + Appendices. inated Boreal Mineral Wetlands. Canadian NGTL. 2014b. Final Caribou Habitat Resto- Forest Service, Northern Forestry Centre, ration Plan for the Leismer to Kettle River Information Report Nor-X-399. Crossover Project. Prepared for: TransCana- MacFarlane, A. 1999. Revegetation of wellsites da Ltd. Prepared by: Golder Associates Ltd. and seismic lines in the boreal forest. B.Sc. Submitted to the National Energy Board in Honour’s Thesis, Department of Biological January 2014. 41 pp + Appendices. Sciences, University of Alberta, Edmonton, Nexen (Nexen Energy). 2013. Algar Caribou Alberta, Canada. Habitat Restoration Program, 2012/13 Phase MacFarlane, A. K. 2003. Vegetation Response to 2 & 3 Areas. Draft report prepared for Cana- Seismic Lines: Edge Effects and On-line Succes- da’s Oil Sands Innovation Alliance (COSIA). sion. Master of Science Thesis, Department Calgary, AB. 109 pp. of Biological Sciences, University of Alberta, Oberg, P.R. 2001. Responses of Mountain Cari- Edmonton, AB. 131 pp. bou to Linear Features in a West-central Al- MEG (MEG Energy). 2014. MEG Energy berta Landscape. M.Sc. Thesis, University of Christina Lake Regional Project: Linear De- Alberta, Edmonton, AB. 126 pp. activation Trial. March 2014. OSLI (Oil Sands Leadership Initiative). NEB (National Energy Board). 2013. Consid- 2012a. Algar Caribou Habitat Restoration erations: Report of the Joint Review Panel for program, Field Operations: Phase 1 Area. Re-

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 140 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 port. Calgary, AB. 76 pp. Photography. Master of Science Thesis, De- OSLI. 2012b. Monitoring Program Algar partment of Geography, University of Cal- – Area 1. Developed with The Silvacom gary, Calgary, AB. Group. Developed for: ConocoPhillips, Szkorupa, T. 2002. Caribou Range Recovery in Nexen, Shell, Statoil, Suncor Energy and Alberta: 2001/02 Pilot Year. Alberta Sustain- Total. Website: http://www.osli.ca/projects/ able Resource Development, Fish and Wild- land. Accessed: July 2012. life Division, Alberta Species At Risk Report Osko, T. J. & Glasgow, M. 2010. Remov- No. 48. Edmonton, AB. 8 pp. ing the Well-Site Footprint: Recommended TCPL (TransCanada Pipelines Ltd.). 2014. Practices for Construction and Reclamation First Year Post-Construction Monitoring Pro- of Well-sites on Upland Forests in Boreal gram Report for the TransCanada Pipelines Alberta. Department of Biological Sciences, Limited Cheecham/Kettle River Area Expan- University of Alberta. sion Pipeline Project. Submitted to the Na- Osko, T. J. & MacFarlane, A. K. 2000. Natu- tional Energy Board. January 2014. ral reforestation rates on seismic lines and TERA (TERA Environmental Consultants). wellsites in comparison to natural burns or 2011. Post-Construction Monitoring Report logged sites. Alberta-Pacific Forest Indus- for the Trans Mountain Pipeline L.P. TMX tries, Boyle, Alberta. – Anchor Loop Project. Prepared for Trans Pembina Institute. 2012. Federal recovery Mountain Pipeline L.P. December 2011. 78 strategy confirms protecting habitat is key pp. + Appendices. to protecting caribou. Available online at: TERA. 2012. Supplemental Update for the http://www.pembina.org/blog/651. Ac- Caribou Habitat Restoration Plan for the cessed February 3, 2014. NOVA Gas Transmission Ltd. Horn River Pyper, M. & Vinge, T. 2012. A visual guide to Mainline Project - Part IIA. Submitted to the handling woody materials for forested land rec- National Energy Board in October 2012. 8 lamation. Oil Sands Research and Informa- pp + Appendices. tion Network, University of Alberta, School USDA (United States Department of Agri- of Energy and the Environment, Edmonton, culture). 2009. Multipurpose Subsoiling Alberta Report No. TR-31. 10 pp. Excavator Attachments. Prepared by Bob Revel, R. D., Dougherty, T. D. & Downing, Monk, Forester. U.S. Department of Agri- D. J 1984. Forest Growth and Revegetation culture, Forest Service National Technology Along Seismic Lines. University of Calgary and Development Program. Revised June Press, Calgary, AB. 2009. 21 pp. Roy, V., Bernier, P.Y., Plamondon, A.P. & van Rensen, C.,Nielsen, S., White, B., Vinge, Ruel, J.C. 1999. Effect of Drainage and Mi- T. & Lieffers, V. 2015. Natural Regenera- crotopography in Forested Wetlands on the tion of Forest Vegetation on Legacy Seismic Microenvironments and Growth of Planted Lines in Boreal Habitats in Alberta’s Oil Black Spruce Seedlings. — Canadian Journal Sands Region. — Biological Conservation of Forest Research 29: 563-574. http://dx.doi. 184: 127-135. http://dx.doi.org/10.1016/j. org/10.1139/x99-024 biocon.2015.01.020 Sherrington, P. M. 2003. Measuring Boreal Vinge, T. 2013. Getting Online: Restoration Forest Fragmentation Change in Response to Effects in Northern Alberta Provincial Per- Seismic Line, Wellsite and Road Revegetation spective on Restoration. 28 pp. Presented at with Scanned False-Colour Infrared Aerial the Golder Associates Woodland Caribou

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 142 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015

et al. 1999

et al., Bedford, 1998 Roy MacIsaac 2004 2010 Golder, OSLI 2012a, 2012b Nexen, 2013 2007b CRRP, Archuleta & 2008 Baxter, USDA, 2009 BC MFR, 2014 BCFS, 1998 2000 BC MOF, BC MFR, 1998 References Macadam & - Winter (fro zen-ground conditions) Ideal Timing for Ideal Timing Treatment • - - - Sufficient frost is re quired to access sites in the winter when crossing lowland ar This varies from eas: winter to Research regarding machines that can operate in lowlands during non-frozen conditions is under Alberta way in NE Considerations to take into account • • - - - For the purposes of creating microsites for planted seed lings, mounding is a well-re searched site preparation technique in the silviculture It is commonly used industry. low-lying areas to in wetter, create elevated, welldrained microsites for seedlings Mounding treed fen and bog areas can enhance a site to promote natural revegetation drier over time, as higher, spots are created that seed can eventually settle into and germinate Mounding has been used as an access control measure on decommissioned roads, seismic lines, and pipelines to discourage off-road It is effective vehicle activity. immediately following imple mentation Ripping is a standard forestry site preparation method that has been modified in this case for tighter workspaces Positive Experiences with this Technique • • • • - -

For access control purposes, mounds The should be created using an excavator. holes left behind by the mounds should generally be approximately 0.75 m deep, The excavated material is po if feasible. sitioned right beside the hole, creating mounds. Ripping should focus on upland sites where excessive moisture is not a concern. created by ripping should be posi Troughs tioned to reduce erosion potential. density of mounding for access Target control and/or microsite creation purposes can vary from 600 to 2,000 mounds/hectare (ha), depending on the size of hole and mound. When completing in synergy with seedling planting, seedlings are generally planted near the hinge of mound: Slightly higher up from the hinge for lowland and transitional sites At or slightly lower than the hinge for upland sites Specifications • • • • • • • - Create microsites in areas where it is deemed to be effec tive for increasing survival and growth rates of planted seed and seedlings, and natural regrowth of woody species Access control Objective(s) • • - preparation: Mound ing and/or ripping using an excavator (Figures 4 and 6) Treatment Mechanical site Table 1. Caribou habitat restoration treatment best practices. treatment 1. Caribou habitat restoration Table

2011 BC MFR, 1998 Cenovus, 2013 2007b CRRP, DES, 2004 Golder 2005, 2010, 2012b, 2011, 2012c MEG, 2014 OSLI 2012a, 2012b Nexen, 2013 NEIPC, 2010 References 2010, AENV, - - Seedlings can be planted on frozen sites in the winter (OSLI, 2012a; MEG, 2014; Cenovus, 2013) Non-frozen stock are gen erally planted as summer stock in consideration of the Least Risk Timing Windows (BC) or Restrict ed Activity Period (AB) for caribou Treatment Ideal Timing for Ideal Timing • • - - Use of frozen seed lings need to consider preparation of nurs ery stock, storage, planting temperature, and use of snow packing following planting to avoid winter freeze/thaw seedling mortality into account Considerations to take • - - - -

Seedling planting is consid ered a long-term restoration treatment due to the length of time it takes to establish effective hiding cover and access deterrents Seedlings should ideally be sourced at least six months prior to planned planting dates Seedlings and/or seed for growing seedlings may not be available for every spe cies prescribed and therefore seed may need to be collect ed and grown in the nursery Seedling planting during win ter is generally restricted to lowland and transitional sites with organic soil that have been treated with mechanical site preparation immediately prior to planting Seedling planting density for reclamation purposes has generally been based on adjacent site type and quickly providing hiding cover; it can range from 2,000 to 2,500 stems/hectare Technique Positive Experiences with this • • • • • - 2007b) 2007b) Tree/shrub species are determined based Tree/shrub on site conditions, the adjacent forest stand and restoration objectives (e.g., low palatability for ungulates). Coniferous tree species (Spruce sp., Pine sp.) are recom mended to meet caribou habitat needs. Considerations for the use of shrubs: Alder is generally planted because it forms an effective access control and line of sight break in a relatively quick period of time Alder has a similar palatability rating for ungulates as conifer species (CRRP Willow is avoided due to the high palatability rating for ungulates (CRRP Shrub and tree seedlings are often planted depending on site conditions and together, anticipated natural revegetation of both species Specifications • • • • • access control erosion control reduce line-of-sight restore habitat Objective(s) • • • • planting (Figures 3 and 4) Treatment seedling Tree/shrub Table 1 continued. Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 144 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 2010 2011 Alberta, 2013 2012b 2012 2012 2012 References 2007b CRRP, Enbridge, 2010 Osko & Glasgow, 2010, Golder, Government of OSLI 2012a, BC MFLNRO, Pyper & Vinge, & Pyper, Vinge - Winter (fro zen-ground conditions) Ideal Timing for Ideal Timing Treatment • - - - -

Potential for fuel loading is a concern. The BC MFLNRO specifies acceptable levels of woody material while con sidering fire man agement objectives. Consultation with the local fire regulations recommended prior to treatment Storage and use of woody materials may be compromised if bark beetle is a con cern in the area and would be discussed with the local forest officer Storage of woody material for extend ed periods without increasing fire hazard can be challenging and should be discussed with district fire managers as part of the planning process when using woody materials Considerations to take into account • • • - - - The length of a treated segment is dependent on sufficient quantities of woody material available. Longer segments are a more effec tive treatment at controlling ATV human access since riders will be less inclined to attempt to travel through the woody material or traverse around it in adjacent forest stands if the woody material continues for an extended distance material can also Woody conserve soil moisture, moderate soil temperatures, provide nutrients after it decomposes, prevent soil erosion, provide a source of seed for natural revegetation, provide microsites for seed germination and protection for introduced tree seed lings, and protect seedlings from wildlife trampling and browsing Spreading of woody material is effective as an access control immediately following implementation material can be Woody brought to a site from anoth er location that has identical tree species Positive Experiences with this Technique • • • • /ha of woody material 3 /ha). An exemption may /ha). 3 Spread woody material evenly across the entire corridor or polygon feature Ensure woody material is consistently dense ATV enough on the ground to discourage and wildlife use Assessment The Guide to Fuel Hazard Abatement in British Columbia (2012) and recommends woody loads do not exceed 99 tonnes/ha (~175 m be allowed for larger volumes from the local fire centre under Section 25 or 26 of the Wildfire Regulation. and Pyper recommend applying Vinge between 60 to 100 m to reclaimed sites mimic the natural range of variability for woody material in the forest Implement at sites left for natural recovery when woody material is available as well sites that are planted with seedlings Specifications • • • • • control of human access during snow free periods erosion control protect planted seedlings from extreme weather, wildlife trampling, and damage from ATVs provide site nutrients when the wood decomposes provide microsites for natural seed ingress Objective(s) • • • • • material (Figure 2) Treatment Spreading of woody Table 1 continued. Table

Bend (hinge) mature trees partially across the line with an excavator while treating features for mounding purposes or spread ing woody material Fell mature trees across the line on upland and transitional sites (e.g., white spruce, pine, aspen, and black spruce) An excavator is preferred for felling trees if site conditions are by pushing them over, suitable for excavator access can be felled with a chain saw if Trees site access is suitable to address safety concerns are to be felled perpendicular the Trees are not to be felled parallel the Trees line. line to reduce a fire hazard locations to occur approximately Treatment every 20 m on lowland and upland sites At each treatment location, 2 or more trees to be felled, from opposite sides of the line, to create an access control and line of sight break locations should occur where Treatment sufficient sized timber is present. locations should be as frequent Treatment as possible to discourage wildlife use, understanding that locations will be variable depending on forest stand adjacent to line More trees to be felled near access points and intersections to restrict access Additional trees can predator movement. be felled along identified lines where the adjacent trees are of suitable height (de pends on width of line, need to cover across entire corridor) Specifications • • • • • • • • • • access control reduce line-of-sight reduce shade effect Objective(s) • • • Bending (Figures 5 and 6) Treatment Tree Tree-felling/ Table 1 continued. Table

This journal is published under the terms of the Creative Commons Attribution 3.0 Unported License 146 Editor in Chief: Birgitta Åhman, Technical Editor Eva Wiklund and Graphic Design: H-G Olofsson, www.rangiferjournal.com Rangifer, 35, Spec. Iss. No. 23, 2015 References 2007a CRRP, - Winter (fro zen-ground conditions) Ideal Timing for Ideal Timing Treatment • - -

Fences are logisti cally challenging to establish in areas without pick-up access. Used infrequently in the past and unknown efficacy Installing fences during summer may be difficult to imple ment due to access availability Considerations to take into account • • • - Fences could also be es tablished using poles and geotextile or similar style decomposable matting Gates can be installed on fences if desired to allow some operational access Positive Experiences with this Technique • • Fences can be installed at intersections with linear corridors and/or along a corridor where predator/human access control and line-of-sight breaks are required Where natural topography or bends in the corridor do not break line-of-sight, fences can be placed to limit access and sight lines panels should be pre-constructed Wooden off-site, fastening the panels together at treatment site to create a fence Specifications • • • access control Objective(s) • Treatment Installing fences Table 1 continued. Table