Karine Pigeon (fRI Research Caribou Program and Grizzly Bear Program), Megan Hornseth (Borealis Ecology), Doug MacNearney (fRI Research Caribou Program) and Laura Finnegan (fRI Research Caribou Program and Grizzly Bear Program)
March 31, 2016
Executive Summary
Anthropogenic disturbance is extensive within caribou range in west-central and north-west Alberta, and habitat restoration is necessary for caribou persistence. Given the extent of the disturbance footprint, it is crucial to understand how disturbance is impacting caribou and how restoration efforts will be most effective in helping stabilize caribou populations. We used GPS data from caribou and wolves, field data on human and wildlife use of seismic lines and pipelines, vegetation heights extracted from LiDAR, non-invasive fecal DNA collections, and a suite of GIS variables associated with landscape, habitat, and anthropogenic features to 1) determine how caribou, their predators, and humans respond to seismic lines at different stages of regeneration, 2) evaluate whether the zone of influence of linear features changes in relation to the level of regeneration, 3) assess how human activity on linear features is affected by landscape attributes and regeneration, 4) assess the size and health of caribou populations, and 5) produce spatially explicit models to aid in the prioritization of areas for restoration.
In year one of this project, analysis of field data failed to reveal associations between wildlife use of linear features and regeneration (n = 95 plots). However, using GPS data, we found that central mountain caribou selected areas further from well sites in the drilling phase during winter, and that Chinchaga boreal caribou selected all types of anthropogenic features less than expected from chance. Chinchaga caribou also selected relatively high elevation, open mature conifer stands on shallow slopes and plains during spring and summer, and low elevation, open mature conifer stands and open habitats on shallow slopes and plains during fall and winter. Overall, regeneration height of seismic lines in the Chinchaga range was low (70% <1m). However, despite this extensive footprint, caribou selected areas further from seismic lines with low vegetation height during spring, summer, and late winter. In comparison, wolves selected flat areas at relatively high elevation in mixed forest and non-forested habitat near small streams, and away from high densities of linear features at the landscape scale but close to areas with high densities of anthropogenic features at the local scale. Wolves also selected for areas near seismic lines with low vegetation heights when in close proximity to seismic lines during the snow-free period but the influence of regeneration height on seismic lines diminished when wolves were located farther from seismic lines. Finally, both species also selected areas near seismic lines with high CTI values (high wetness).
Activity patterns of humans in relation to seismic lines (n = 1250) and pipelines (n = 435) were not consistent across caribou ranges, and models using GIS and field variables neither predicted human use of pipelines, nor human use of seismic lines in north-west Alberta. In west-central Alberta, high levels of human use occurred on seismic lines with low vegetation heights, in areas with low densities of well sites and high ungulate counts, and on seismic lines that were close to paved roads. The probability of human use was zero when vegetation heights reached 2.7m. The models for west-central Alberta correctly predicted 62-82% of human use on seismic lines. Using non-invasive fecal sampling, we measured stress (cortisol and corticosterone) and reproductive hormones (progesterone and testosterone) in west-central caribou herds. The final year of fecal collection is still underway but preliminary data revealed no differences in hormone levels among herds. Progesterone levels indicated that 51 out of 60 female caribou tested in west-central Alberta were pregnant at the time of sampling.
Analyses failed to reveal clear regeneration thresholds to identify when caribou and wolf habitat selection patterns were no longer affected by the presence of seismic lines. We were therefore unable to define a breakpoint at which seismic lines and their surrounding habitat should be considered functional caribou habitat again. Final results from fecal surveys (spring 2017) could help assess caribou functional habitat further. Nevertheless, our research findings reveal a suite of variables that can be used to prioritize seismic lines for restoration, and these variables can also be used to direct mitigation of human impact within caribou ranges. Based on our findings, we overlaid spatially explicit maps of the relative probability of habitat selection for caribou and wolf in the Chinchaga range. These resulting maps identify areas surrounding regenerating seismic lines that have the highest probability of overlap between caribou and wolves. From these probabilities, we then created a map of seismic lines ranked with respect to priority for restoration. This approach classified 1539km (2.9%) of seismic lines in the Chinchaga range as high or very high priority for restoration.
Overall, our detailed analyses of animal and human response to regenerating seismic lines and well site status contributes new knowledge towards understanding the effect of anthropogenic disturbances on animal behavior, and towards understanding areas of overlap between caribou and their predators. The covariates identified here, along with seismic lines identified as high priority for restoration, can be used by land planners and industrial partners to identify areas where restoration of previously disturbed areas will have the greatest benefit to caribou in the Chinchaga range. Ultimately, these results may be used to expedite restoration of caribou habitat to reach the disturbance targets outlined in the federal recovery strategies.
Forest Ecology and Management Paper
#15-ERPC-01