Female Moose Prioritize Forage Over Mortality Risk in Harvested Landscapes

Abstract

Since 2010, several moose (Alces alces) populations have declined across North America. These declines are believed to be broadly related to climate and landscape change. At the western reaches of moose continental range, in the interior of British Columbia, Canada, wildlife managers have reported widespread declines of moose populations. Disturbances to forests from a mountain pine beetle (Dendroctonum ponderosae) outbreak and associated salvage logging infrastructure in British Columbia are suspected as a mechanism manifested in moose behavior and habitat selection. We examined seasonal differences in moose habitat selection in response to landscape change from mountain pine beetle salvage logging infrastructure: dense road networks and large intensive forest harvest cutblocks. We used 157,447 global positioning system locations from 83 adult female moose from 2012 to 2016 on the Bonaparte Plateau at the southern edge of the Interior Plateau of central British Columbia to test whether increased forage availability, landscape features associated with increased mortality risk, or the cumulative effects of salvage logging best explain female moose distribution using resource selection functions in an informationtheoretic framework. We tested these hypotheses across biological seasons, defined using a cluster analysis framework. The cumulative effects of forage availability and risk best predicted resource selection of female moose in all seasons; however, the covariates included in the cumulative models varied between seasons. The top forage availability model better explained moose habitat use than the top risk model in all seasons, except for the calving and fall seasons where the top risk model (distance to road) better predicted moose space use. Selection of habitat that provides forage in winter, spring, and summer suggests that moose seasonally trade predation risk for the benefits of foraging in early seral vegetation communities in highly disturbed landscapes. Our results identified the need for intensive landscape-scale management to stem moose population declines. Additional research is needed on predator densities, space use, and calf survival in relation to salvage logging infrastructure.

Mammalian predator and prey responses to recreation and land use across multiple scales provide limited support for the human shield hypothesis

“…Other anthropogenic disturbances seemed to have stronger effects on species than did recreation. Moose site use increased in more intensely logged habitat, suggesting that foraging opportunities in open habitat could outweigh the risks associated with increased visibility and predation in those areas (Boucher et al, 2009 ; Francis et al, 2021 ; Massé & Côté, 2009 ). For carnivores, responses to other forms of human disturbance varied among species, as was demonstrated in one of our single landscapes (Kananaskis, Heim et al, 2019 ).…”

Whose line is it anyway? Moose (Alces alces) response to linear features

“…Differences in seasonal habitat preferences of male and female moose are driven by differences in resource needs, reproductive status, and exposure to risk. Females with calves trade off access to forage against security and survival of their calves through the year, which drives more variable habitat selection patterns in females when compared with males (Francis et al, 2021; Mumma et al, 2019). It is also possible that narrow seismic lines with shallower/more compacted snow are more attractive to the larger males during winter (Droghini & Boutin, 2017; Telfer & Kelsall, 1984), driving the more consistent response we observed during that season for males.…”

“…Localized population differences have been attributed to the interacting effects of habitat and climate change, which have changed moose distribution and density and also changed moose mortality risk from hunters, predators, and pathogens (Murray et al, 2006; Parlee et al, 2012; Timmermann & Rodgers, 2017). While moose generally benefit from early seral vegetation (e.g., abundant forbs and shrubs) produced after disturbance (i.e., forage; Rempel et al, 1997; Serrouya et al, 2021), their response to early seral habitat created by wildfires and forest harvesting is hierarchical, varies at broad scales relative to the surrounding habitat matrix, and at fine scales relative to the characteristics and age of disturbances (Francis et al, 2021; Maier et al, 2005; McKay & Finnegan, 2022; Mumma et al, 2021). Moose response to linear features is even more variable: linear features provide forage (Finnegan, MacNearney, et al, 2018; MacDonald et al, 2020; St‐Pierre et al, 2022), specifically they provide forage adjacent to mature forest with thermal shelter (van Beest et al, 2012), greater snow interception during winter (Droghini & Boutin, 2017), and refuge from predators (Kunkel & Pletscher, 2000).…”

“…Moose response to linear features is even more variable: linear features provide forage (Finnegan, MacNearney, et al, 2018; MacDonald et al, 2020; St‐Pierre et al, 2022), specifically they provide forage adjacent to mature forest with thermal shelter (van Beest et al, 2012), greater snow interception during winter (Droghini & Boutin, 2017), and refuge from predators (Kunkel & Pletscher, 2000). However, they may also act as ecological traps by increasing moose mortality risk via encounters with vehicles, human hunters, and predators (Eldegard et al, 2012; Francis et al, 2021; LeBlanc et al, 2011). Although moose response to roads has been the focus of previous research (e.g., Bartzke et al, 2015; Eldegard et al, 2012; Yost & Wright, 2001), there has been limited research on their response to pipelines and seismic lines.…”

Using camera traps and N‐mixture models to estimate population abundance: Model selection really matters

“…Wetlands and riparian areas are important habitats for moose, as they select wetlands or areas near water throughout the year, and several forage species can be found in those habitats (Francis et al., 2020; Mumma et al., 2021; Schrempp et al., 2019). We used the Freshwater Atlas (FA) to measure the proportion of each grid cell that contained wetlands and riparian habitat.…”