The Ya Ha Tinda Elk project is now amongst the longest running elk research project in the world.  Initiated in 2000, the Ya Ha Tinda elk project is the result of a collaboration between University of Alberta, University of Montana, Parks Canada, and Alberta Environment and Parks, Fish and Wildlife Division.  While early studies in the late 1970’s and early 1980’s lead by Dr. Luigi Morgantini laid the foundation for our latter studies (Morgantini and Hudson 1988), there was a ~ 20-year gap in active research on Alberta’s most important elk population.  Initiated at first because of questions regarding the changing migratory dynamics of the migratory Ya Ha Tinda elk population, the project has since evolved into North America’s longest running wild, free-ranging elk research projects focused on fundamental and applied research.

The Ya Ha Tinda is one of Alberta’s most pristine montane rough fescue winter ranges and elk which provides the habitat foundation for one of Canada’s most iconic and largest populations.  The Ya Ha Tinda elk population is also a transboundary system, with annual elk migratory cycles that have spanned two provinces (elk have migrated into Yoho National Park, British Columbia), two land management regimes including Banff National Park, Provincial Forest Land Use Zones, and Provincial multiple use zones.  Ya Ha Tinda is also managed as a premier bull elk harvest area that provides much sought-after hunting opportunities to residents and guided hunters alike.  Meaning mountain prairie in the Stoney Sioux language, Ya Ha Tinda has long been important to First Nation communities for hunting and traditional land use practices.  And the region is also home to recovered populations of grizzly bears, wolves, and other large mammal predator-prey species, including for the first time in over a century – Plains bison. In this transboundary setting, our long-term research has contributed directly to enhancing interagency cooperation and management of this important elk population.

Our long-term studies have documented dramatic changes in migratory behavior and population dynamics arising from this complex landscape of gradients in carnivore densities, habitat productivity, and differences in land management practices. Over the last 30-40 years migrant to resident ratios have substantially decreased from 12:1 (1977-1987, (Morgantini and Hudson 1988)), 3:1 in the early 2000’s (Hebblewhite et al. 2006), to more recently a ratio closer to 1:1 (Berg 2019).  Early studies in the 2000’s demonstrated that migratory elk moving west into BNP experienced much higher forage quality which translated to higher calf 8-month old weights and higher pregnancy rates (Hebblewhite et al. 2008).  Yet western migrants also experienced reduced predation risk from wolves, but higher risk from grizzly bear predation (Hebblewhite and Merrill 2007, MacAulay 2019). Resident elk remaining year-round near Ya Ha Tinda, in contrast, experienced lower forage quality, but, compensated for this by reducing predation risk by seeking out fine-scale predation risk refugia surrounding human development at the Ya Ha Tinda (Hebblewhite and Merrill 2009, Robinson et al. 2010). Commensurate with these shifts in the migratory dynamics of this population have been correspondingly significant population shifts and changes (Hebblewhite et al. 2006, Killeen et al. 2015, Berg 2019). 

In the last decade, a new migratory strategy has emerged with female elk now undertaking an eastward migration into Provincial multiple use lands in and adjacent to the 2001 Dogrib fire (Killeen et al. 2015). Long-term research revealed individual elk are making density-dependent switches in migratory behavior, evidently to seek out these new beneficial areas (Eggeman et al. 2016). In 2013 – 2017, we lead a neonate calf survival research component to understand spatial variation in calf survival (Berg 2019).  Calf survival and cow:calf ratios have indicated that calf survival of elk migrating east on to industrial forest experienced higher calf survival. In January 2019, Jodi Berg defended her PhD thesis on elk calf survival, and her thesis details are provided in our publications section below. This new migratory behavior has seemingly stabilized the Ya Ha Tinda elk population, which has fluctuated between 400 – 600 elk now for almost a decade. Our long-term predator-prey research shows, however, that this stabilization is not likely a result of wolf or grizzly bear predation stabilizing the population at low density (Hebblewhite et al. 2018). Instead, our research suggests that migratory behavior itself may be providing escape from low densities (Hebblewhite et al. 2018).

We analyzed 18 years of demographic and migratory data from the Ya Ha Tinda elk population in and adjacent to Banff National Park (BNP), Alberta, Canada. The winter range is the Parks Canada-owned Ya Ha Tinda Ranch where ~100 - 180 horses are boarded during winter for use in the National Parks. The Ya Ha Tinda elk population has been studied since 1972 (Morgantini et al. 1988), with Global Positioning System (GPS) radiocollared individuals starting in 2001. From 1972 the population size grew from 800 in the early 70’s, reached a peak of 2200 elk in the mid 1990’s, and then has declined and stabilized around 500 elk since 2010. The winter range of the Ya Ha Tinda elk population is a rough fescue (Fesctuca campestris) grassland which covers 20 km2. In winter, Chinooks (warm air and high winds) clear snow from the grassland acting as a refuge for ungulates. The summer range consists of three distinct areas used by 3 different migratory tactics, 1) residents who remain year-round on the rough fescue grassland, 2) eastern migrants which migrate to the eastern foothills that experience logging and oil and gas extraction, and 3) western migrants- migrate into BNP living in alpine or subalpine ecotypes (Fig. 1). Each summer range has different levels of forage quality and predation risk by human and non-human predators (Hebblewhite and Merrill 2009). This population is hunted in the provincial portions of their range and treaty First Nations harvest can occur year-round for males and females. For more details, see (Hebblewhite et al. 2006).

Aerial count data [YHT_MinimumCountWinterElkSurveys.csv]— Elk count data were collected annually by Parks Canada or Alberta Environment and Parks Biologists on the winter range in January-March, based on helicopter (most commonly a Bell 206 Long Ranger) availability and ideal survey conditions (full snow cover). Aerial surveys were not conducted in 2015 and 2018. Because these surveys were not repeated within a year, nor were sightability parameters estimated, we assume that the counts from the aerial surveys are an index of the true population size but represent a minimum population count. To align our estimates of the aerial surveys with our population count data and ratio data which is also collected in February and Mark, we estimated our total population size and cow: calf ratios 2 months prior to the model anniversary on 1 June.

Calf: cow ratio data [YHT_CalfCowRatioData.csv]— We conducted ground-based surveys during the months of February- March. During these surveys we collected sex and age ratio data of observed elk groups. We estimated the cow: calf ratio by summing the total number of observed calves divided by the total number of observed cows (Hurley et al. 2011). The calf to female variance was calculated following the methods of Cochran (1977). The error rate associated with determining yearling and adult females is very high and thus the denominator included both adult and yearling females.

Survival data [YHT_AdultFemaleSurvivalDataRaw.csv; YHT_AdultFemaleSurvivalRate.csv]— We captured and radio-collared adult female elk in late winter (February-March) from 2001-2018. In the early years of the study (2001-2012), female captures were conducted using baited corral traps and aerial net-gunning. After 2013, all female elk were captured using ground-darting from horseback. Female elk were fitted with a Global Positioning System (GPS) and/or a Very High Frequency (VHF) collar to provide location and mortality data. All elk were captured and handled using techniques approved by the University of Montana IACUC (AUP-MH004-16). Trained personnel determined cause of death as soon as mortality of a collared elk was detected (on average within 5.2 days (SE= 8.0), from Hebblewhite and Merrill (2011)) using consistent criteria (Alt and Eckert 2017). We used a modified Kaplan Meier- non-parametric survival model to estimate adult survival. Monthly survival was estimated using a Bayesian, generalized-linear mixed-effect model with a random effect for migratory tactic: year and a fixed effect for migratory tactic. These monthly survival estimates were multiplied to obtain a yearly survival estimate for each tactic (YHT_AdultFemaleSurvivalRate.csv). The raw data provided has an indicator of the state of the animal indicating if the observation resulted in the elk being alive was alive (1) or dead (2- unknown/natural/other,3-wolf, 4 grizzly bear, 5 cougar, 6 human caused).

Calf survival data [YHT_CowCalfObsData.csv]: Calf survival was estimated using cow-calf resight of marked female elk. Individual female elk were observed multiple times (x̄=3) between 1 July and 1 April with the presence or absence of a calf being recorded for each occasion. Monthly survival rates were estimated using a modified Cormack-Jolly-Seber model where all observed individuals were assumed to have a calf at the first occasion. If a calf was seen with a marked female it was recorded as alive and if it was not observed with a marked female it was recorded as unobserved.