https://orcid.org/0000-0002-4797-6284
Abstract
Ecosystem-based management (EBM) is a landscape-level management and planning process that is common across North America. A primary tenet of EBM is that the area and intensity of anthropogenic disturbance should mimic the historical natural disturbance of the focal ecosystem. Biodiversity should persist, at least at a coarse scale, where anthropogenic disturbance, such as forest harvesting, matches natural disturbance. However, EBM is failing some species, particularly those that are dependent on old forest. Across many areas of Canada, woodland caribou (Rangifer tarandus caribou) are declining because of the direct and indirect effects of habitat loss and fragmentation. This is even though forest management often follows the principles of EBM. We conducted a qualitative comparison of the responses of woodland caribou to wildfire and forest harvesting, considering a broad range of responses, including habitat selection and distribution, forage, movement patterns, and population dynamics. We found that while wildfire and forest harvesting both influence caribou, the negative effects are generally greater following forest harvesting. For example, wildfire and forest harvesting result in the loss of habitat, but caribou are more likely to shift, abandon or contract their range in response to harvest. The literature also suggested a stronger negative population response of caribou to forest harvest when compared to wildfire. This difference could be the result of greater residual forest structure associated with wildfire as well as the extensive resource roads that are necessary for forestry operations. Although there is sound theoretical support for EBM, the practice, as implemented, may not be effective for maintaining the habitat and ultimately populations of woodland caribou.
Introduction
Ecosystem-based management (EBM) is a land management strategy focused on maintaining biological diversity, sustainability, and resilience (Grumbine 1994). A primary tenet of EBM is that anthropogenic disturbances emulate natural disturbances, with the goal of maintaining natural ecological processes (Bergeron et al. 2004; Christensen et al. 1996; DeLong 2007; Grumbine 1994; Kuuluvainen and Grenfell 2012; Mori et al. 2013; Seip 1998). There is much evidence to suggest that for many ecosystems, human activities can result in disturbance patterns and ecological outcomes that differ from natural disturbance (Bergeron et al. 2004; Bergeron and Fenton 2012; Best et al. 2024; McClelland et al. 2023; Nitschke 2005; O’Hara 2016; Stockdale et al. 2016). To reduce these differences, forest management has adopted many of the principles of EBM (Bergeron et al. 1999, 2001; Burton et al. 2006; Harvey et al. 2002; Moussaoui et al. 2016). Theory and empirical evidence suggests that biodiversity and ecosystem function are more likely to be maintained when forest management emulates the size, severity, and frequency of wildfire (Angelstam 1998; Bergeron et al. 2004; Bichet et al. 2016; Franklin 1993; Gauthier et al. 1996).
Wildfires are extremely variable in extent, severity, and ecological effect (Bergeron et al. 2002). Across the forests of North America, wildfires occur when fuel, topography, environmental conditions, and an ignition event align. Wildfires vary in size from <1 ha to >600 000 ha in extreme cases, happen at frequencies of between 0 and 500 years, impact a range of stand types and ages, and can range in severity from surface fires that remove understory vegetation, to crown fires that kill or remove all vegetation, including mature trees (Bergeron et al. 2002; McRae et al. 2001). Wildfires result in a complex mosaic of forest types that differ in age, relative to the dominant tree species, and the understory plant community (Bergeron et al. 2002; DeLong and Tanner 1996; McRae et al. 2001). In comparison, the maximum size of a harvest block is controlled by legislation (3.5–150 ha), harvest occurs every 40–100 years, harvest blocks are clustered at the landscape scale, productive stands are targeted, and most of the mature trees and understory cover are removed (Andison 2024; Bergeron et al. 2002; McRae et al. 2001; Mladenoff et al. 1993). In comparison to wildfires, harvesting usually results in more homogeneity, with large areas of regeneration interspersed with fragments of mature forest (McRae et al. 2001; Bergeron et al. 2002; Lorente et al. 2012). Harvesting also creates roads that are not associated with natural disturbances (McRae et al. 2001). As a consequence, wildfire and forest harvesting likely have differential effects on the habitat of wildlife (Best et al. 2024; Kalies et al. 2010; but see Huggard et al. 2015), which is particularly relevant when using EBM to manage forests for the conservation of biodiversity. However, there is a paucity of research comparing the effects of EBM-based forest harvesting to wildfires relative to the habitat, distribution, or abundance of wildlife (McRae et al. 2001; Koivula and Vanha-Majamaa 2020).
Caribou and reindeer (Rangifer tarandus) are members of the deer family (Cervidae) with a circumpolar distribution. Populations of caribou in Canada are legally recognized as distinct evolutionary units (i.e. designatable units, DUs) for conservation assessment and recovery. Currently, many of the caribou DUs found across boreal and montane ecosystems, collectively known as woodland caribou (R. t. caribou), are assessed as Threatened or Endangered (COSEWIC 2002, 2011, 2014). The leading cause of decline for many populations of woodland caribou is human-caused forest loss or change leading to unsustainable predation by wolves, bears, and cougars (COSEWIC 2014; Johnson et al. 2022). Anthropogenic disturbance, primarily resource extraction by the forestry and oil and gas sectors, is the principal cause for the loss and fragmentation of habitat for caribou (Festa-Bianchet et al. 2011; Vors and Boyce 2009). The loss of old forest, typically used by caribou, results in early-seral habitat favorable to apparent competitors of caribou—moose (Alces americanus), deer (Odocoileus spp.), and elk (Cervus canadensis). Resource extraction also results in linear features, such as roads, that facilitate predator movement (DeCesare et al. 2010; Mumma et al. 2018; Serrouya et al. 2017). An increase in the density and distribution of apparent competitors and predators leads to unsustainable predation of caribou (Hervieux et al. 2013; Serrouya et al. 2021; Sorensen et al. 2008).
Caribou are a species of cultural significance to many Indigenous peoples across Canada, a flagship for the conservation of boreal and montane landscapes (Hummel and Ray 2008), and an umbrella species for biodiversity (Bichet et al. 2016; Drever et al. 2019; Labadie et al. 2024; but see Micheletti et al. 2023). As such, many Indigenous peoples, and federal and provincial governments, are focused on the recovery of small or declining populations of caribou. Conservation initiatives can be intensive and controversial and have resulted in constraints on land-use activities associated with industrial development (Hebblewhite 2017; Johnson et al. 2022).
Across Canada’s boreal and montane forests, wildfire is the dominant natural stand-replacing disturbances, while forest harvesting is the dominant anthropogenic disturbance when considering disturbance footprint alone (Kuuluvainen and Gauthier 2018; Natural Resources Canada 2022). These two natural disturbance agents can act independently, cumulatively, or interactively, and their effects will likely be exacerbated by climate change (Bergeron and Dansereau 1993; DeMars et al. 2023; Labadie et al. 2023; Natural Resources Canada 2022; Wang et al. 2015). Both wildfire and forest harvesting have the potential to affect caribou habitat, movement, and population dynamics (Labadie et al. 2023). Caribou require home ranges with large tracts of intact mature forest that allows them to space out from other ungulates and their predators (Bergerud et al. 1990; Seip, 1992). In addition, caribou have specialized diets, particularly during winter, composed of terrestrial and arboreal lichens (Antifeau, 1987; Thomas et al. 1996b), which are generally most abundant within mature forests consisting of pine (Pinus spp.) and spruce (Picea spp.). Like many species, woodland caribou have evolved to exist within the ecological context of natural disturbance regimes that govern ecosystem dynamics. Thus, EBM-based forest harvesting that mimics the stand structure and understory succession of natural disturbance could allow for the persistence of woodland caribou and other species that are dependent on these landscapes (Bergeron et al. 2004).
Past research suggested that woodland caribou had a stronger negative response to anthropogenic disturbances even where EBM-based forestry practices had been adopted (Johnson et al. 2020; Konkolics et al. 2021). Recently, some studies have assessed the potential effects of silviculture (McKay and Finnegan 2023; Nadeau-Fortin et al. 2016; Vitt et al. 2019), as well as compared differences in caribou habitat following wildfire and forest harvest across large geographic areas (Best et al. 2024). Despite this, there has been limited research directly evaluating the efficacy of EBM or comparing the effects of forestry and wildfire for the habitat, movements, distribution, or abundance of woodland caribou. In lieu of direct study, EBM guidelines for caribou are dependent on the extensive caribou-related literature, but there has been no comprehensive review of research describing caribou response to fire or the differences in response to wildfire versus forest harvesting.
We conducted a comprehensive review of the literature to assess the differential effects of forestry and wildfire on the habitat, population, and distributional dynamics of woodland caribou across boreal and montane ecosystems. When considering anthropogenic disturbances, we focused on forestry as that was the most common cause of habitat change for caribou across much of the species’ range and was the focus of EBM. We did not conduct a quantitative metanalysis as the literature represented a broad range of ecosystem conditions, disturbance types, and outcomes for caribou. Although we did not directly evaluate the effectiveness of EBM for conserving caribou and their habitat, our findings can reveal potential limitations of that management approach and inform empirical studies that can directly address assumptions about our ability to use forestry to mimic wildfire.
Methods
We searched Google Scholar, a commonly used and publicly available academic database, with terms that were relevant to the disturbance of habitat of woodland caribou (e.g. wildfire, forestry, forest harvest, anthropogenic disturbance). We performed secondary reviews of the literature cited within papers or reports that were identified through the Google Scholar search. We reviewed and referenced studies focused on woodland caribou from across Canada, while acknowledging that the responses of caribou to wildfire and forestry were likely ecosystem dependent (Konkolics et al. 2021; Nagy-Reis et al. 2020). We recognize that responses of caribou to disturbance will vary according to ecosystem type and that such variation will influence the strategy and potential efficacy of EBM. However, given the wide range of ecosystems that are used by caribou (i.e. low-elevation fens and bogs to high-elevation, low-productivity forest) we were unable to partition the literature by type. Furthermore, some studies did not report specific ecosystem or forest type or use a common ecosystem classification, beyond broad ecogeographic categories typical of the ecotype or DU of the study population. Nonetheless, we attempted to provide a coarse-scale description of the ecological context or geographical location of the studies that we summarized. We considered a broad range of potential responses of caribou to fire and forestry that aligned with typical measures of the spatial or population ecology of the species: habitat selection and distribution, forage, movement patterns, and population dynamics.
Results
Habitat selection and distribution
Wildfire
Generally, caribou do not alter their home ranges in response to wildfires (Dalerum et al. 2007; Faille et al. 2010), but instead avoid the burned portions of their ranges (Joly et al. 2003; Schaefer and Pruitt 1991). However, there is variation in the response among regions and populations. For example, caribou in wildfire-dominant landscapes like northern Saskatchewan shifted their home range to include more burned habitats in comparison to their home ranges pre-burn. This suggests that recently burned areas provide some value as habitat for caribou (Silva et al. 2020). There is also some variation in the fine-scale responses of caribou to wildfires. In black spruce–dominant forests in east-central Alaska, caribou avoided the core areas of burns but not areas within 500 m of the burn edge (Joly et al. 2003). Caribou in the southern Yukon (spruce-dominant forests) and northern Saskatchewan (jack pine–dominant forests) selected remnant forest patches within burned areas (Russell 2018; Skatter et al. 2017).
Caribou generally select for older forest and avoid recent burns. That pattern is relatively consistent across regions and ecosystems (Dalerum et al. 2007; Johnson et al. 2015; Joly et al. 2003; Konkolics et al. 2021; Lafontaine et al. 2019; Palm et al. 2022; Robinson et al. 2012; Shepherd et al. 2007; Silva et al. 2020; Skatter et al. 2017). For example, in Alaska, northern Canada, and Saskatchewan, caribou avoided severely burned areas for up to 30–50 years (Joly et al. 2003; Palm et al. 2022; Russell 2018; Skatter et al. 2017). However, there is some variation among regions; in Jasper National Park, Alberta, caribou selected forests that were >75 years post-fire (Shepherd et al. 2007), while a few studies reported that caribou selected for recently burned areas. In deciduous and mixed-wood upland forests in north-eastern British Columbia, Mumma et al. (2018) found that caribou selected burns <40 years old. Lafontaine et al. (2019) found that caribou selected for burned areas <20 years old in black spruce, balsam fir, and jack pine forests in Québec without previous exposure to wildfire. In mixed-wood forests in west-central Alberta, Peters (2010) also found that caribou selected for burns <20 years old.
There is seasonal variability in the effect of fire on habitat selection. Research from eastern Alaska, the Yukon, Northwest Territories, and northern Alberta showed that caribou avoided burned areas more strongly during winter than summer (Palm et al. 2022). In eastern Alberta, avoidance of burned habitat by individual caribou was greatest during late winter and least during early winter and summer (Konkolics et al. 2021). Conversely, caribou in northern Saskatchewan avoided recent burns (<5 years old) during all seasons except during the spring calving season (Silva et al. 2020), while other research reported that caribou selected burns equally among seasons (Mumma et al. 2018). The general avoidance of recently burned areas may be the result of a reduction in the availability or quality of forage (Dunford et al. 2006; Fisher and Wilkinson, 2005; Joly et al. 2010; Schaefer and Pruitt 1991; Shepherd et al. 2006). Across eastern Alaska and northern Canada, avoidance of burns was consistent with a lack of available forage (Palm 2021; Russell 2018). However, the interaction between forage availability and caribou response post-fire varies across the distribution of caribou (DeMars et al. 2019). For example, in many regions, caribou avoided burned habitat between 6 and 60 years post-fire even when terrestrial lichen was available (Dalerum et al. 2007; Fisher and Wilkinson 2005; Joly et al. 2003; Klein 1982; Lafontaine et al. 2019; Schaefer and Pruitt 1991; Thomas et al. 1996b). Moose, deer, and bears may take advantage of the regrowth of shrubby vascular plants that occurs in burned areas (Bergerud 1974; Robinson et al. 2012). Thus, avoidance of burned areas by caribou may be a response to increased predation risk associated with the regeneration of habitat for their apparent competitors (Bergerud 1974; Courtois et al. 2007; Robinson et al. 2012).
Selection of young burns may represent a trade-off between successional changes in forage availability and exposure to predators (Lafontaine et al. 2019; Mumma et al. 2018). For example, prior to the regrowth of early-seral vegetation, burned areas may have winter forage (i.e. terrestrial lichens) in remnant patches and reduced predation risk (DeMars et al. 2019; Neufeld et al. 2021; Rettie and Messier 2000). Furthermore, the reduction in horizontal cover may increase predator detection (Skatter et al. 2017). As vascular plants re-establish, caribou may take advantage of that high-quality forage during spring and summer, but that would represent a trade-off with increased predation risk as apparent competitors such as moose and deer would also be attracted to those post-burn habitats, as well as shared predators like bears and wolves (Denryter et al. 2022; McGreer et al. 2015).
Forestry
Caribou shift their home ranges to avoid forest harvesting or abandon heavily harvested areas within their home range entirely (Beauchesne et al. 2013; Courtois et al. 2007; Donovan et al. 2017; Faille et al. 2010; Honsberger 2011; MacNearney et al. 2016; Mahoney and Virgl 2003; Slater 2013; Smith et al. 2000). In multiple regions of western Canada, broad-scale forest harvesting has resulted in the extirpation of caribou from low-elevation valley bottoms (Poole et al. 2000; Smith et al. 2000; Williams et al. 2021). Some populations demonstrate an initial home range expansion in response to forest harvesting. This may be a compensatory response for lost habitat or displacement from human disturbance (Beauchesne et al. 2013; Chubbs et al. 1993; Courtois et al. 2007; Cumming and Beange 1993).
In general, caribou select mature patches of forest and avoid areas that are recently harvested (Bowman et al. 2010; Chubbs et al. 1993; Courbin et al. 2009; Courtois et al. 2007; DeCesare et al. 2012; Gagné et al. 2016; Rettie and Messier 2000). The avoidance period ranges from 12 years in Ontario mixed-spruce, jack pine forests (Cumming and Beange 1993) to 20 years in Québec balsam-fir, white-birch forests and mixed-wood forests in west-central Alberta (Leblond et al. 2016; Peters 2010) to upwards of 40 years in mixed-wood forests in Québec, Alberta, and British Columbia (Gagné et al. 2016; Mumma et al. 2018; Rudolph et al. 2019). While some caribou may be more tolerant of harvest blocks if they are adjacent to mature forests, as seen in white birch–balsam fir and black spruce forests in Québec (Hins et al. 2009), there was evidence that the area of avoidance extended beyond the harvest block boundary, with caribou avoiding areas >1 km from recent harvests in coniferous forests in Newfoundland, Ontario, and Alberta (Schaefer and Mahoney 2007; Smith et al. 2000; Vors et al. 2007).
The strength of avoidance of harvest blocks varies by region, season, sex, and reproductive status. Caribou in Québec avoided harvested areas more strongly in summer (Vanlandeghem et al. 2021), but caribou in western Canada avoided harvested areas more strongly in winter (Palm 2021). In Québec, caribou selected for forests >20 years post-harvest during all seasons except in spring, when those harvested areas were avoided (Leblond et al. 2016). In Newfoundland, female caribou avoided harvest blocks, while male caribou showed no response (Schaefer and Mahoney 2007). In Ontario, female caribou with calves avoided stands <20 years post-harvest (Walker et al. 2021), while in Québec, female caribou with calves selected for harvest blocks <20 years post-harvest (Leblond et al. 2016). Contradictory results may be an indicator of variation in the trade-offs between forage and predation risk (Walker et al. 2021; Leblond et al. 2016), or confounding factors such as roads associated with forest harvesting (Walker et al. 2021).
Caribou may change their use of habitats following exposure or displacement from forest harvesting. In Newfoundland, females displaced by harvesting selected mature black spruce forests, while females that were not displaced continued to use all habitat types in proportion to their availability (Chubbs et al. 1993). In Québec, caribou in harvested landscapes selected closed canopy mature conifer forest more than other forest types during calving and summer, while caribou in unharvested landscapes showed equal selection of forest types (Moreau et al. 2012).
Partial or small-block harvest has been proposed as a strategy to reduce the impact of forestry on caribou habitat. For example, retaining patches of old forest can provide winter forage that increases the quality of caribou habitat, ultimately resulting in greater use of harvested regions (Serrouya et al. 2006). In Alberta, caribou were more likely to use retention patches that were >20% of the harvested area in clear cuts 15–18 years post-harvest (Franklin et al. 2019). While the habitat conditions in partially harvested stands are more similar to mature forest, the associated road networks may result in increased predation risk for caribou (Nadeau-Fortin et al. 2016).
In addition to the direct habitat changes occurring within harvest blocks, roads associated with forest harvesting and other anthropogenic disturbances (e.g. wellsites, pipelines, and seismic lines) in the surrounding area may reduce the value of remnant habitat (Beauchesne et al. 2014; Johnson et al. 2015; MacNearney et al. 2016). These anthropogenic disturbances are consistently avoided by caribou, and their effects are cumulative across landscapes (Beauchesne et al. 2014; DeCesare et al. 2012; Johnson et al. 2015). Models of habitat selection that consider the surrounding habitat matrix, so-called “Functional response models” (Holbrook et al. 2019), have demonstrated that the response of caribou to forest harvest varies according to the availability of mature forest within the broader home range (Losier et al. 2015; Moreau et al. 2012). The response of caribou to forest harvest likely varies with the density of roads and other disturbances across the broader landscape, as reported for other boreal species (McKay and Finnegan 2022; Muhly et al. 2019). However, to our knowledge, no studies have specifically assessed caribou response to forest harvest (or fire) as a function of the density and status (i.e. active vs. inactive vs. restored) of roads in the surrounding habitat matrix.
Forage
Wildfire
Terrestrial (Cladina spp., Cladonia spp., Stereocaulon spp., Cetraria spp.) and arboreal (Alectoria sarmentosa, Bryoria spp., Usnea spp.) lichens are the dominant food source for caribou during winter. However, the importance of terrestrial versus arboreal lichen varies geographically; e.g. arboreal lichen is more important in the diet of caribou that occupy montane ecosystems (Bergerud et al. 2007; Denryter et al. 2017; Johnson et al. 2001; Klein 1982; Schaefer and Pruitt 1991; Thomas et al. 1996b; Thompson et al. 2015). Forage lichens are more resilient to low-intensity wildfires compared to stand-replacing wildfires (Miller et al. 2018; Shepherd et al. 2006). Large wildfires typical of the boreal forest initially reduce the availability of both terrestrial and arboreal lichen (Ray et al. 2015; Russell and Johnson 2019).
Arboreal lichens may take up to 60 years to regenerate post-fire and are most abundant in stands >100 years old (Berryman and McCune 2006; Foster 1985; Horstkotte et al. 2011; Shepherd et al. 2006). The successional trajectory of terrestrial lichen varies with ecosystem type that may differ according to soil moisture and nutrients, topography, and the presence of permafrost (Arseneault et al. 1997; Bergeron and Fenton 2012; Brulisauer et al. 1996; Coxson and Marsh 2001; Foster 1985; Hart and Chen 2008; Kershaw 1977; Russell and Johnson 2019). In boreal and foothills ecosystems, pioneer species like bryophytes and vascular plants typically dominate burned habitat for the first 21–40 years until terrestrial lichens recover (Dunford et al. 2006; Russell and Johnson 2019; Thomas et al. 1996b). At more northern latitudes or elevations, terrestrial lichens may take as long as 75 years to recover post-fire (Shepherd et al. 2006; Thomas et al. 1996a). In peatlands disturbed by fire in northern Alberta, terrestrial lichen was less abundant for the first 20 years but was comparable to mature forest within 30–40 years after fire (Dunford et al. 2006). In the Canadian shield of northern Québec, bryophytes and vascular plants dominated burned habitat up to 30 years following wildfire, after which lichens dominated (Arseneault et al. 1997). Lichen growth after wildfire is not always linear, as increasing canopy closure from tree regrowth may allow feather mosses and vascular plants to outcompete terrestrial lichens (e.g. 70–300 years post-disturbance). In pine–lichen woodlands in the central interior of British Columbia, bryophytes and Stereocaulon spp. lichens dominated stands up to 20 years following wildfire, Cladonia spp. lichens dominated from 20 to 150 years after fire, and feathermosses overgrew the terrestrial lichen in stands 100 to 150 years after fire (Brulisauer et al. 1996; Coxson and Marsh 2001; Sulyma and Coxson 2001). In the pine and pine/spruce stands in west-central British Columbia, the understory did not shift from terrestrial lichens to feather mosses until 200–300 years following wildfire (Cichowski and Banner 1993). Where xeric acidic soils limit growth of vascular plants and bryophytes, terrestrial lichens may be the dominant ground cover in stands as young as 10 years after burning and can persist beyond 300 years (Brulisauer et al. 1996; Kershaw 1977).
Bryophytes and vascular plants can supplement caribou diet (Denryter et al. 2017; Rettie et al. 1997; Thomas et al. 1996b); however, the temporal availability of post-fire vegetation is limited, and colonizing species can be of low nutritional value to caribou (Denryter et al. 2017; Joly et al. 2003). While growth of terrestrial lichens slows in older stands, there is some uncertainty whether burning forests to encourage an increase in terrestrial lichens would have an overall benefit to caribou populations (Apps et al. 2001; Klein 1982; Shepherd et al. 2006; Szkorupa and Schmiegelow 2003; Thomas et al. 1996b).
The reduced forest canopy resulting from wildfires can influence snow accumulation and ablation rates (Johnson et al. 2001; Maxwell et al. 2019; Thomas et al. 1996b). Increased snow depth and snow hardness can limit access to terrestrial lichens and can shift the diet of caribou from terrestrial lichens to arboreal lichens (Johnson et al. 2001, 2004; Kinley et al. 2007; Thomas et al. 1996b). In montane areas with a consistent deep snowpack, caribou feed almost exclusively on arboreal lichens during winter (Terry et al. 2000; Webber et al. 2022). With climate change, the compounding impacts of wildfires and changes in precipitation and temperature (e.g. reduced snow packs, rain-on-snow events; Putkonen and Roe 2003) may increase the importance of arboreal lichens as winter forage for caribou.
Forestry
Forest harvesting reduces the abundance of both terrestrial and arboreal lichens. Terrestrial lichens are removed or damaged through a variety of mechanisms, including mechanical damage from harvesting machinery, changes in ecological conditions following canopy removal, and silvicultural activities such as replanting and brushing (Sulyma and Coxson 2001; Waterhouse et al. 2011). Across the conifer forests of interior British Columbia, harvesting directly reduces the availability of arboreal lichens by removing trees and snags and indirectly reduces lichens by increasing wind exposure on the residual trees (Stevenson et al. 2001). The most significant long-term effect of forest harvesting results from the change in the composition and successional stage of the harvested stand. In boreal mixed-wood forests in Northern Ontario, vascular plants (including forbs, graminoids, and shrubs) dominate the understory 7–25 years post-harvest (Hart and Chen 2008). In lodgepole pine–dominant forests in British Columbia, terrestrial lichens generally recolonize 20–40 years post-harvest (Cichowski et al. 2022; Harris 1992; Waterhouse et al. 2011), but in west-central Alberta terrestrial lichens have been recorded as recently as 6 years post-harvest (T. McKay, unpublished data). In many regions, including conifer forests in Oregon, USA, Labrador, Canada, and northern Sweden, arboreal lichens do not typically re-establish on new trees until the regenerating stands are 60–100 years old (Berryman and McCune 2006; Foster 1985; Horstkotte et al. 2011).
The recolonization of terrestrial lichens is dependent on silvicultural methods, pre-harvest condition, and geographic location. Harvesting and silvicultural practices (timing, site preparation, density of stems removed, retention patch size) can vary considerably and differ greatly in their influence on the regrowth of lichens and vascular plants (Bartemucci et al. 2022; Coxson et al. 2003; Kranrod 1997; Nadeau-Fortin et al. 2016; Stevenson et al. 2001; Vitt et al. 2019; Waterhouse et al. 2007, 2011). The post-harvest application of herbicides, intended to accelerate growth of conifers by inhibiting competitors like shrubs, grasses, and deciduous trees (McCormack 1994), can also decrease terrestrial lichens (Newmaster et al. 1999). In general, minimizing ground disturbance during harvesting improves retention of terrestrial lichens. That can be achieved by harvesting during winter (Coxson and Marsh 2001; Kranrod 1997) or avoiding silvicultural or processing practices associated with heavy machinery, such as scarification or stump-side delimbing (Kranrod 1997). Harvesting on a deep snow pack or using low-impact methods that minimize physical damage can maintain the pre-harvest understory community (Hart and Chen 2006, 2008; Nguyen-Xuan et al. 2000). However, the removal or reduction of canopy cover and associated changes in light, temperature, and humidity inevitably result in changes to the growth or composition of the vegetation community (Hart and Chen 2006, 2008; Nguyen-Xuan et al. 2000). Lichen colonization is generally greatest near forest edges, thus, harvest blocks with smaller openings have increased lichen colonization and reduced recovery times for caribou forage (Bartemucci et al. 2022).
In comparison to clear-cut, partial-cut harvesting (group selection or single tree selection) may result in greater retention, growth, or decreased recovery times of arboreal and terrestrial lichens (Nadeau-Fortin et al. 2016; Vitt et al. 2019). Group or single tree selection maintains arboreal lichen in the residual trees immediately post-harvest (Coxson et al. 2003). Removing <30% of the total basal area may maintain or increase the productivity of arboreal lichens, likely due to the increased ventilation beneficial to some species (Coxson and Stevenson 2007; Nadeau-Fortin et al. 2016; Waterhouse et al. 2007). Arboreal lichens in partial-cut stands are at increased risk of wind exposure (Stevenson et al. 2001), but the blow-down of trees in partially cut stands may increase the availability of arboreal lichens for caribou several years following harvest (Terry et al. 2000). Partial thinning can maintain or increase the growth rate of terrestrial lichen (Vitt et al. 2019; Waterhouse et al. 2011), although access to forage may be limited by snow accumulation associated with decreased canopy cover (Seip and Jones 2008). Partial harvesting may have similar effects to low-intensity wildfires (Bergeron et al. 2002), although to our knowledge, there are no published studies comparing partial cutting to wildfire intensity.
Movement patterns
Wildfire
Wildfire can change the structure of the forest, limiting or reducing the efficiency of caribou movement. In multiple regions across Canada, recently burned forest (5–10 years post-fire) have greater densities of downed trees and understory vegetation compared to unburned habitats, which can restrict caribou movement (Metsaranta et al. 2003; Schaefer and Pruitt 1991; Shepherd et al. 2007). The loss of canopy cover and change in forest structure resulting from wildfire also influences the accumulation and melting rates of snow when compared to unburned stands (Kirchhoff and Schoen 1987; Maxwell et al. 2019; Skidmore 1994; Winkler 2011). The resulting deep or thinly crusted snow can increase the energetic costs of caribou movements (Avgar et al. 2013; Fancy and White 1985; Stuart-Smith et al. 1997; Telfer and Kelsall 1979).
Forestry
Forest harvesting within caribou ranges can disrupt the movement of caribou (Bloomfield 1979; Finnegan et al. 2021). For example, caribou move less when occupying ranges with extensive forest harvesting (Smith et al. 2000). Such responses can vary between sexes, with females moving up to two to three times further from harvesting than males (Chubbs et al. 1993). A high density of replanted or naturally regenerating trees may impede caribou movement (Stevenson et al. 2001; Wilson et al. 2023). The road networks associated with forest harvesting may act as barriers to caribou movement (Apps and McLellan 2006; Dyer et al. 2002; Stevenson et al. 2001) and facilitate increased predation (Dickie et al. 2017; James and Stuart-Smith 2000; Mumma et al. 2018). Furthermore, access roads provide increased hunting opportunity and can result in the demand for continued access into these regions (Brinkman et al. 2009).
As with wildfire, harvesting reduces canopy cover (Telfer 1978), and the associated changes in snow depth and hardness in clear-cut and partially harvested stands can influence the movement of caribou (Jones 2007; Kirchhoff and Schoen 1987; Seip and Jones 2008), as well as other sympatric ungulates and their shared predators (see “Movement—Wildfire”). However, caribou have the lowest foot loading and are the most efficient at moving across snow, possibly providing an ecological advantage relative to their predators and apparent competitors (Droghini and Boutin 2018; Telfer and Kelsall 1979, 1984).
Population dynamics—recruitment and survival
Wildfire
The population dynamics of caribou are thought to be relatively insensitive to the occurrence of wildfire in the absence of anthropogenic disturbance (Environment Canada 2011; Gonet 2020; Johnson et al. 2020; Konkolics et al. 2021; Neufeld et al. 2021; Palm 2021; Sorensen et al. 2008; Stewart et al. 2020). While wildfire may have a small negative effect on recruitment in some regions for up to 20 years post-burn (Johnson et al. 2020), recruitment was unaffected in other caribou populations with up to 76% of their home ranges burned (Dalerum et al. 2007). Konkolics et al. (2021) reported that use of burned habitat did not influence the survival of adult female caribou in eastern Alberta. Overall, wildfire accounts for only 5% of the variation in recruitment rates of boreal caribou across Canada (Environment Canada 2011; Palm 2021; Sorensen et al. 2008).
The apparent competitors of caribou (moose, deer, and elk) generally respond positively to the early-seral habitat created by wildfires (Boyce et al. 2003; Joly et al. 2016; Lord and Kielland 2015; Maier et al. 2005; but see DeMars et al. 2019). Subsequent increases in deer, moose, and elk within caribou ranges are linked to increases in the distribution and densities of the predators they share with caribou (Hebblewhite et al. 2009; Kittle et al. 2017; Serrouya et al. 2017). Furthermore, bears forage in the early-seral plant communities associated with wildfires and are known predators of caribou calves (Frenette et al. 2020; Leblond et al. 2016). Despite the established relationships between early-seral habitat and the increase in predation risk associated with caribou–predator overlap (DeMars and Boutin 2018; Mumma et al. 2018), there is no strong evidence linking wildfire or remnant burned forest to survival of caribou. Current research indicates that the survival of female caribou is generally not influenced either by the presence of burned habitat or by caribou use of burned areas (Apps et al. 2013; Dalerum et al. 2007; Johnson et al. 2020; Konkolics et al. 2021; Stewart et al. 2020), although such effects may occur at coarser spatial and temporal scales than have been evaluated to date.
Forestry
There is much evidence linking the area of forest harvested to declines in the distribution and the abundance of caribou (Environment Canada 2011; Fryxell et al. 2020; Johnson et al. 2015; Lochhead et al. 2022; Serrouya et al. 2021; Vors et al. 2007). Variation in the survival of caribou is related to the extent of early-seral stands in home ranges (Wittmer et al. 2007), and caribou mortality rates increase with increasing densities of harvest blocks (Grant et al. 2019; Losier et al. 2015; Vanlandeghem et al. 2021). As the effects of harvesting are cumulative and temporally dynamic (e.g. emergence of early-seral vegetation), there may be a time-lag between forest harvesting and a decrease in the survival or recruitment of caribou (Vors et al. 2007). In the short term (e.g. 5 years post-harvest), caribou populations may remain stable despite increased rates of forest harvesting (Mahoney and Virgl 2003).
Forest harvesting can provide ecological conditions that increase early-seral forage for moose, deer, and elk, with concurrent increases in the predators of caribou (Gagné et al. 2016; James et al. 2004; Kuzyk et al. 2004). Even in the same stand types, harvested stands have different ecological legacies than wildfires (Best et al. 2024). In some cases, partial harvest may better mimic natural disturbances and the habitat conditions of caribou, but also their apparent competitors. For example, moose may select partial cut stands because they provide a juxtaposed mix of thermal cover and open areas with increased browse (Eastman 1977).
Disturbances that remove mature forest result in greater predation of caribou (Courtois et al. 2007; Wittmer et al. 2007; Apps et al. 2013; Environment Canada 2011; Fryxell et al. 2020; Vanlandeghem et al. 2021). The magnitude of the effect varies across demographic groups, as female caribou and calves are more likely to be depredated when occupying landscapes with recent forest harvest (Courtois et al. 2007). The spatial configuration of forest harvesting may also influence rates of predation (Vanlandeghem et al. 2021). For example, group selection harvesting that leaves multiple small openings can create habitat that favors the apparent competitors of caribou (moose and mule deer) and their shared predators (bears and wolves), increasing predation risk for caribou (J. Bradshaw, unpublished data). In some cases, caribou mortality is not associated with higher proportions of younger forests or edge habitat (Apps et al. 2013), but this may be related to caribou avoidance of harvested areas at coarse spatial scales (DeCesare et al. 2012).
Forestry operations create resource roads, and there is abundant evidence relating road occurrence and road density to caribou mortality (Apps et al. 2013; James and Stuart-Smith 2000; Vanlandeghem et al. 2021). Predators, specifically wolves, use roads and other linear features to travel faster and further into caribou habitats (Dickie et al. 2017; James and Stuart-Smith 2000; Mumma et al. 2018), resulting in increased caribou–wolf encounters and increased predation risk for caribou (Mumma et al. 2017; Whittington et al. 2011). Lochhead et al. (2022) found that roads buffered by 50 m was the best predictor of the decline of 12 subpopulations of mountain caribou in southern British Columbia.
The negative effects of forest harvesting and roads, as well as other anthropogenic disturbances, are cumulative and interactive (Beauchesne et al. 2014; Johnson et al. 2015). For example, wolves are more likely to use roads and seismic lines when the density of harvest blocks and seismic lines, respectively, are less in the surrounding area (Muhly et al. 2019; Pigeon et al. 2020). Similar responses have been reported for moose (Finnegan et al. 2023; Mumma et al. 2019). It is probable that the interactions between the polygonal and linear disturbances created by forestry operations may be responsible for the differential demographic responses of caribou to forestry relative to wildfire.
Synthesis and discussion
Effective EBM requires that the disturbance resulting from forestry emulates the ecosystem changes that occur during and following wildfire. Wildfires and forest harvesting both influence the habitat use, forage availability, movement, and population dynamics of caribou across a range of ecosystem types. However, the scientific literature revealed that the negative outcomes from disturbance were typically greater following forest harvest and associated road construction and silvicultural practices (Fig. 1). Following from that general conclusion, forest practices should be modified to better mimic wildfire or the other dominant natural disturbance types where maintenance of caribou habitat is the primary objective across managed landscapes.
Summary infographic comparing the effects of wildfire and forest harvesting on woodland caribou habitat selection and distribution, demographics, movement, and forage availability.
Substantial evidence from across Canada indicates that post-harvest ecosystems do not match post-fire ecosystems (Bergeron and Fenton 2012; Hobson and Schieck 1999; Krawchuk and Cumming 2011; McRae et al. 2001; Mladenoff et al. 1993; Nitschke 2005; Zimmerling et al. 2017), although the extent of that mismatch may vary by location and forest type (e.g. Huggard et al. 2015), and overall cumulative disturbance. Forest characteristics such as patch size and shape, stand age, and successional pathways differ between harvest blocks and burned areas (McRae et al. 2001), with trickle-down impacts on boreal and montane species like caribou. Resource roads associated with forestry are an unnatural part of any EBM harvest prescription and represent a major difference between fire and forest harvesting, which has largely been ignored in studies assessing caribou response and demography in relation to forest harvesting (but see Lochhead et al. 2022). These linear features can increase the movement, hunting efficiency, and distribution of the predators of caribou (Blagdon and Johnson 2021; Dickie et al. 2017). While the impact of linear features could be managed through deactivation and restoration, these efforts are slow or not always effective (e.g. Bentham and Coupal 2015). The definition of restoration success also varies (Ruiz-Jaen and Aide 2005), including a reduction in wildlife use (Keim et al. 2019; Tattersall et al. 2020; Dickie et al. 2021; Lacerte et al. 2022), wolf movement (Neufeld 2006; Dickie et al. 2022), or a change in vegetation height or composition (Lacerte et al. 2021). Although studies are generally monitoring short-term success, and over time further insights may emerge, so far, no deactivation and restoration treatments of linear features have achieved both functional (predator movement) and structural (vegetation composition) restoration (Ray 2014).
EBM prioritizes types and rates of human disturbance that maintain ecological patterns and processes, with the intended outcome being complete, naturally functioning, and resilient ecosystems (Grumbine 1994). Where EBM mimics natural disturbance operationally and ecologically, resource extraction is expected to maintain a wide range of social, economic, and environmental values (Gauthier et al. 2009). Thus, forest harvest and silvicultural practices that emulate the natural disturbance regime of the focal ecosystem could maintain natural levels of biodiversity and allow old forest–dependent species, such as caribou, to persist (DeLong 2007). For example, in western Canada, harvest prescriptions are designed to emulate the dominant disturbance regime specific to the ecosystem occupied by caribou (Carlson and Kurz 2007; Government of Alberta 2017; Seip 1998). Across low-elevation, pine-dominated ecosystems, a more frequent and widespread fire regime dictates large, aggregated harvest blocks, with less old forest. In contrast, the prescribed harvest regime for wetter, high-elevation ecosystems is greater retention of old forest (i.e. less “natural disturbance”) and small harvest blocks (e.g. group selection harvesting) that are more representative of localized fire or wind events (Seip 1998). Similarly, managed forests in Alberta are moving toward large areas of aggregated harvest blocks, as this pattern of human disturbance more closely resembles fire regimes across those ecosystems (Carlson and Kurz 2007; Government of Alberta 2017).
More broadly, EBM requires that the area of forest disturbed, frequency of disturbance, and successional pathways match the natural disturbance dynamic for the region. Forest loss and change from human disturbance cannot be additive to natural disturbance. That is not the case in many ecosystems that support woodland caribou, especially where economic priorities demand greater volumes of timber than might be expected through historical natural disturbance dynamics. Moving forward, regulatory and land-use decisions (e.g. allocating Annual Allowable Cut) may not be able to reconcile the disconnect between ecological process and socioeconomic demands (DeLong 2007; Nagy-Reis et al. 2020).
The divide between ecological and economic realities will likely broaden with climate change (Pau et al. 2023; St-Laurent et al. 2022). Across much of North America, there is less control over large and frequent fires as well as increasing tree mortality from forest “pests” (Allen et al. 2010; Kasischke et al. 2010; Kuuluvainen and Gauthier 2018; Ratajczak et al. 2018). At the same time, there is an increase in demand for timber products from Canada’s forests, likely resulting in pressure to maintain “unnatural” harvest levels (Natural Resources Canada 2022). Combined, the loss of trees from natural disturbance and the sustained pressure to maintain forest harvest to meet economic demands may result in changes in forest composition (e.g. younger age distribution), which are beyond the adaptive capacity of many wildlife species. Old forest–dependent species such as caribou continue to experience dramatic losses in habitat and rapid population declines across their range, suggesting that the current paradigm of EBM (e.g. Seip 1998) is ineffective, at least in practice (Johnson et al. 2015; Nagy-Reis et al. 2020).
Our review of the literature suggests that for woodland caribou, there are significant differences between the outcomes of natural disturbance and forest harvesting, although the confounding impacts of resource roads and other anthropogenic disturbances are likely influencing some of these differences. Perhaps of most concern is the fact that human disturbance has a consistently stronger influence on caribou demography than wildfire (Dalerum et al. 2007; Environment Canada 2011; Johnson et al. 2020; Palm 2021; Stewart et al. 2020). For example, 60% of the variation in recruitment for boreal caribou was attributed to habitat change caused by anthropogenic disturbance (Environment Canada 2011).
There is sound theoretical support for EBM (DeLong 2007; O’Higgins et al. 2020), but the implementation of EBM is constrained by social, economic, operational, and technical limitations. As highlighted by McRae et al. (2001) and two decades later by Kuuluvainen et al. (2021), there remains little empirical research assessing the ecological impacts of EBM versus wildfire and how those impacts might vary among ecosystems. For example, although alternate harvesting techniques have been proposed to mitigate the impacts of forestry on caribou (Stevenson et al. 2001; Waterhouse et al. 2011, 2015), research to date has been focused in one study system, or did not compare different harvesting techniques to wildfires in the same system (Franklin et al. 2019; Serrouya et al. 2006; Vitt et al. 2019), a critical piece of the EBM puzzle.
Caribou have been proposed as an umbrella species (Bichet et al. 2016; Drever et al. 2019; Labadie et al. 2024), providing opportunities for EBM that benefits other boreal and montane species. However, our literature review revealed that there has been little quantitative evaluation of the effectiveness of EBM for caribou and, by association, other species. Despite decades of published research on caribou, the nature of scientific research and funding means that most studies are limited in time or space (but see Best et al. 2024). Also, our literature review was focused on the direct impacts of wildfire and forestry for caribou, but understanding how wildfire and forestry affect forage, movement, habitat use, and demography of the apparent competitors (moose, deer, elk) and shared predators (bears, wolves, cougars) of caribou is also crucial (DeMars et al. 2023).
Conclusions and future work
We provided a detailed comparative review of the observed response of caribou and caribou habitat to wildfire and forestry practices. Many of those responses are direct and relatively obvious and could be mitigated by choosing ecologically appropriate harvest patterns, reducing the area and rate of harvest, and applying post-harvest silviculture that best mimics community succession that follows wildfire. For example, retention of old forest for boreal caribou or small-block harvest for mountain caribou could maintain winter habitat. Appropriate management prescriptions that may be used to address the indirect effects of forestry, including successional pathways of the post-harvest plant communities and associated changes in predator–prey dynamics, are less clear. This may be especially challenging where caribou have multiple apparent competitors (e.g. moose and deer) that are also of value to the public and to Indigenous peoples, and in the case of moose, are themselves declining in some areas (Parlee et al. 2012; Priadka et al. 2022; Timmermann and Rodgers 2017; Titus et al. 2009). In some cases, caribou populations are extremely small and close to extirpation (Johnson et al. 2015), and although an EBM approach may be a sound long-term strategy, it may be insufficient to address population declines in the short term. Instead, more aggressive actions may be required, e.g. reducing forest harvest within caribou ranges (Government of Alberta 2017). There may also be an argument for management strategies that deviate from what we expect from natural disturbance. Silvicultural approaches such as site preparation, mechanical or chemical stand tending, and herbicide application that aggressively limit forage for moose and deer may not coincide with the “natural” outcomes of wildfire but may be essential for arresting the decline of small populations of caribou (McKay and Finnegan 2023).
Despite extensive scientific literature reporting different outcomes of forestry and wildfire for caribou, our review revealed a lack of quantitative research focused on the mechanisms for those differences. Future research should progress from simply describing differences in outcomes to exploring why forestry has, in general, a larger negative effect on the habitat, movement, and population ecology of caribou. For example, existing and large datasets of GPS-collared animals and the establishment of long-term, standardized monitoring programs could be used to evaluate the differences or similarities in the ecology of caribou across broad geographic areas that have disturbance of different types, but of similar age and ecosystem conditions. Existing national wildfire and forest harvest data (e.g. https://nfi.nfis.org/) could be used to determine the influence of stand type, disturbance size, shape, and regeneration on caribou as well as their apparent competitors and predators. Furthermore, there may be opportunities to conduct retrospective experiments that assess habitat conditions, including forage, within burned areas as well as areas with forest harvest and different types of silvicultural treatment after controlling for confounding factors such as time since disturbance, ecosystem type post-disturbance treatment, and the relative use of those areas by monitored caribou (sensu Best et al. 2024).
In addition to few studies addressing the causal factors that might explain the different outcomes from forestry and wildfire, we found no studies that explicitly assessed the efficacy of EBM for caribou. This is despite direct application of that approach to conservation and management of woodland caribou in some jurisdictions, such as British Columbia (Seip 1998). EBM has been criticized as amorphous, conflicted, and insufficient as a stand-alone strategy that will meet the needs of individual species that need intensive recovery actions (Lackey 1998; Simberloff 1998). Thus, it is not surprising that there have been no definitive tests of this strategy for caribou, a low-density species with a broad distribution that overlaps with powerful and dominant socioeconomic interests (Collard et al. 2020; Hebblewhite 2017).
Despite the limitations, EBM has the potential to address the cumulative impacts of disturbance (Beauchesne et al. 2014; Johnson et al. 2015) that results from natural and anthropogenic causes. There is much to be gained, relative to the knowledge limitations that we identified, in linking EBM to active adaptive management experiments (Serrouya et al. 2019; Wilman and Wilman 2017). When paired, these approaches could provide both learning and solutions to the complex and confounding interactions of forest change with linear features and the longer-term effects of climate change (DeMars et al. 2023). Those studies and resulting practices could address basic knowledge gaps, such as the additive or interactive effects of roads and forestry or wildfire disturbance, as well as outcomes of restoration that may differ according to disturbance type.
In addition to the immediate technical challenges of managing today’s forests are the unknown dynamics of future disturbance within a context of landscapes that have been dramatically altered by a century of forest harvest, fire suppression, and other human-caused disturbance (Allen et al. 2010; Kasischke et al. 2010; Kuuluvainen and Gauthier 2018). The combined impacts of climate change and historical forest management are likely to change the size, frequency, and severity of wildfires and insect outbreaks into the future (Bleiker et al. 2019; Hanes et al. 2018; Pau et al. 2023; Wang et al. 2015). While the outcome of these changes is likely to occur over decades and is inherently difficult to anticipate, evidence suggests that forest changes resulting in less caribou habitat and greater density or distribution of their apparent competitors and shared predators will lead to worsening conditions for this iconic Canadian species (DeMars et al. 2023; Johnson et al. 2015). Given that uncertainty, a sound starting point would be forest practices and planning that create conditions that are representative of the habitats that allow caribou to persist today. Those conditions are more aligned with wildfire and natural disturbance than extensive forest harvest in combination with other human-caused forest change (Fryxell et al. 2020). Although current knowledge is incomplete and uncertainty about the future is accelerating, the principles of EBM and adaptive management are likely the best starting point in planning for caribou across changing landscapes.
Acknowledgements
We respectfully acknowledge that the research summarized in this paper took place across the traditional territories and ancestral homelands of the Indigenous peoples of Canada. This project was funded by the Alberta Regional Caribou Knowledge Partnership (ARKCP) which receives support through the Forest Resource Improvement Association of Alberta (FRIAA).
Author contributions
Suzanne Stevenson (Writing—original draft, Writing—review & editing), Laura Finnegan (Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing—original draft, Writing—review & editing), Chris Johnson (Supervision, Writing—review & editing), and Tracy L. McKay (Writing—review & editing).
Conflict of interest
None declared.
Funding
This work was supported by the Alberta Regional Caribou Knowledge Partnership (ARCKP). Funders had no role in the production of this document.
Data availability
There are no new data associated with this article.
