Models by Sex and Age
Male, female, and young moose selected different habitat features at different scales. Male moose selected habitat at the broadest scale of the three groups, with a best-fitting spatial scale of 4750 m radius, followed by females at 2500 m, and then young at 1500 m. We did not distinguish between females with and without calves in the female, as the habitat used by females with calves was captured in modelling young.
Nine of the fourteen predictors that best explained male moose habitat use were human footprint variables: recreation, residential, cultivation, clearings, borrowpits, facilities, mines, seismic lines, and wells (Figures 3, 4). Male moose were the only group with no burn area predictors in the top model. Instead, the remaining five predictors were land cover types: water, closed mixed or deciduous forests, bogs and wetlands, closed shrublands, and open conifer stands. Of all predictors, recreation areas had the largest β estimate and wells had the smallest. Only closed shrublands, mines, and open conifer stands had negative associations with male moose habitat use, whereas all other variables had positive associations (Figures 3, 4).
Female moose were the most detected (Figure 2) and selected habitat slightly differently than males (Figures 3, 4). Area burned (including areas burnt 11-15 and 16-20 years ago) was the strongest predictor of female moose, with strong negative effect sizes. Female moose avoided nearly all burned areas but had positive habitat associations with harvest, pipelines, and many other human features. Unlike males, females had positive associations with closed shrubland and mines.
Young moose, which had the lowest detection rate (Figure 2), had the strongest effect sizes of model predictors and the largest number of predictors that differed from the other groups (Figures 3, 4). Five of the 12 variables that predicted young moose habitat were human footprint variables. Young moose were the only group to have positive associations with burned areas (including areas burnt 11-15 and 16-20 years ago). Like female moose, young moose had positive associations with closed shrublands. However, there were two predictors not in the top female moose models where young moose differed from male moose, including facilities and seismic lines. Young moose strongly avoided both features, despite their very slight positive association with adult male moose use (Figures 3, 4).
Discussion
Industrial resource extraction has altered moose relative abundance and distribution across the nearly 10,000 km2 of the Whitefish Lake First Nation territory, just as observed by elders and community members. Human footprint metrics explained variance in moose detections in all models, though often with very small effect sizes. However, the young moose model, where we expected to see strong selection for high-quality forage areas in open spaces, had the strongest negative relationship with anthropogenic landscape features (Figures 3, 4). The strong negative relationship between young moose and petroleum exploration “seismic” lines (Dabros et al., 2018) and industrial processing facilities (Fisher & Burton, 2018) supports the observations of Indigenous community members and reiterates the importance of both broad and local-scale impacts of human land use on boreal species important to Indigenous communities. Importantly, young moose strongly selected for slightly older burned areas, suggesting that early-seral forage generated by natural disturbance is important, but that human-caused disturbance (which also has ample early-seral forage vegetation (Routh & Nielsen, 2021)) is not equivalent to the young, open-canopy patches created by fire. Fire (or here, burned area) has a large and important role in boreal environments, but is rapidly shifting in size and severity across Canada, impacting wildlife species (DeMars et al., 2019; Palm et al., 2022). The roles of both fire and human land use as strong drivers in moose habitat selection are well-known (DeMars et al., 2019; Dickie et al., 2020; Ethier et al., 2024; Fisher & Burton, 2018; Fisher & Wilkinson, 2005; Johnson & Rea, 2023), but we highlight the important distinction between open areas caused by fire, versus those caused by resource extraction, for young moose seeking high-quality forage. For traditional territories highly impacted by development, such as WLFN’s territory, this has substantial consequences for food sovereignty.