My master’s project was developed through a unique collaboration between the University of Montana, Round River Conservation Studies and the Taku River Tlingit First Nation (TRTFN). Completed in the spring of 2010, my study supported the management initiatives of the TRTFN by developing woodland caribou habitat models that identified important seasonal habitats.

The project objective was to use an innovative combination of habitat modeling approaches to determine the effect of cumulative human developments on the Atlin herd. To support this effort, we collaborated with the University of Calgary to develop an updated landcover classification for the range of the Atlin herd based on satellite imagery. The landcover classification improved on previous forest landcover models, and the new product was used in the development of seasonal caribou habitat models.

I developed resource selection habitat (RSF) models using GPS collared caribou data and habitat suitability index models with the traditional ecological knowledge (TEK) of the TRTFN. The results of these models are available in my thesis and results of the RSF models were published in the Journal of Biological Conservation (Polfus et al. 2011). The TEK habitat modeling paper was published in the Journal of Wildlife Management (Polfus et al. 2014).

This approach has generated substantial interest at national and international conferences and has the potential to promote new methods for caribou management across northern Canada.

Other Projects

As part of my collaboration with Round River Conservation Studies, I also worked on a number of other projects related to the Atlin herd of northern mountain woodland caribou. Full explanations of these projects can be found in the Atlin Woodland Caribou Final Report 2011.

Cumulative Effects Toolkit

There is a growing need to understand how potential future development might affect habitat selection of the Atlin herd. A GIS-based toolkit that uses information from the RSF habitat models was developed by Rick Tingey. This toolkit assesses the influence of new human infrastructure on the statistical habitat models and predicted the future reduction in habitat quality. This tool will allow managers, such as the TRTFN, to make informed decisions about the effects of proposed projects by examining the effects of future development scenarios before development occurs. The interface with ArcGIS is intended to allow easy updating of human infrastructure layers, including potential or proposed projects to determine how these projects will alter the underlying habitat quality. This provides a dynamic evaluation of proposed projects on potential caribou habitat through simple metrics that measure the loss of habitat quality. The toolkit report can be downloaded here.

Caribou Pregnancy Testing through Fecal Pellet Collection

Understanding female ungulate pregnancy status is an important component for studying long-term population demography. Knowledge of pregnancy rates and reproductive intervals provides researchers and managers with a useful tool for detecting changes in population growth rates and calf survival. Analysis of progesterone and estrone conjugates from collected fecal samples during late pregnancy stages (March and April) has been used as a non-invasive method of detecting pregnancy status.

In March 2009 and 2010 we collected fecal pellets from woodland caribou in the Atlin herd. To collect fecal pellets we located small groups of female caribou from a helicopter. We collected 10-15 pellets per individual pile along individual caribou paths. To rule out collection of male samples and multiple samples from the same individuals we sent samples for genetic testing to Wildlife Genetics International.  Another set of samples were shipped frozen to the Toronto Zoo Reproductive Physiology Lab. Analysis of pregnancy hormones was conducted on samples known to be individual female caribou following genetic sampling.

Our results indicate that between 86 and 98% of female caribou were pregnant in the Atlin area. This result confirms findings from other studies on woodland caribou that suggest that between 88-100% of adult female caribou become pregnant each year.

Predator Diet Stable Isotope Analysis

Information coming soon.

Lichen Sampling

To test the RSF caribou habitat models that predict avoidance of human development we measured lichen biomass and height across the study area. Other authors have also used lichen biomass as an indirect measure for caribou and reindeer avoidance of human infrastructure (Nellemann and Cameron 1996, Nellemann et al. 2000, Nellemann et al. 2001, Vistnes et al. 2004, Dahle et al. 2008). These studies found that lichen height and biomass decreased with increasing distance to roads, resorts and power lines in Norway. We predicted that there would be a negative correlation between the height or biomass of lichen and distance to roads across the study area because caribou would overgraze lichen in areas far from human disturbance and grazing pressure would be lowest near roads in areas within the winter ZOI. We also predicted that presence of scat would increase in areas outside the winter ZOI.

In 2009 we measured lichen biomass at sites across the study area to build a biomass specific regression equation to convert percent lichen cover and lichen height to biomass (full methods can be found in the report). In 2010, we measured lichen height and cover rather than biomass at each site. We found no relationship between lichen height (average for Cladina and Cladonia spp.), lichen cover or the predicted biomass values and distance to roads. There was also no difference in the average height of Cladina and Cladonia spp. inside or outside of the ZOI buffer. However, we did detect the presence of caribou scat more often outside the ZOI buffer (30 of 53 sites) than inside the ZOI buffer (16 of 25 sites). The probability of detecting caribou scat also increased with increasing lichen cover.

Our study does not seem to support the results of previous research which indicate that lichen height/biomass decreases with increasing distance from human disturbance and is indicative of caribou grazing pressure. Interestingly, we did find a difference between the presence of caribou scat within and outside the ZOI buffer. In our study area, caribou scat may be more indicative of caribou presence than lichen height/biomass. Thus, scat transects may be an efficient way of monitoring the Atlin caribou herd in the future.