Modeling of sampling designs for peary caribou survey in Bathurst Island complex Canada
MetadataShow full item record
Sustainability of wildlife resources requires effective management strategies. Unbiased estimation of wildlife populations through efficient survey methodology is therefore crucial in formulating effective wildlife management policy. I expected intensive survey for low Peary caribou populations to produce good precision and accuracy. Also, for moderate and low survey coverage to produce useful minimum counts at medium and high Peary caribou densities. Empirical Peary caribou data points and watershed delineations obtained from previous aerial survey carried out in the Island were used in creating resource selection model. The significant variables used in formulating the realistic habitat scenarios of the resource selection model included elevations, slope, and hill-shade. Specified low, medium and high densities of Peary caribou were simulated across the Island using the Resource Selection Function (RSF). Systematic transects placement of varying spacing (low, moderate and high survey coverage) were overlaid on the different population density scenarios, and distance estimation method used in determining the population estimates. The detection probabilities (more than 50%) and the coefficient of variation (precision level as low as 18%) of the survey designs revealed that they were suitable for Peary caribou survey in Bathurst Island. However, the accuracy levels of each survey design, measured by the percent difference between the simulated and the estimated Peary caribou populations for each density scenarios varied greatly between the intensive survey design and the moderate/low survey coverage. For low density Peary caribou, 20% accuracy level was observed with intensive survey coverage while 75% - 96% accuracy level was observed for moderate and low survey coverage. Low density of Peary caribou have been reported in Bathurst Island in past surveys. I would therefore recommend that moderate or low survey coverage which produced better accuracy and relatively good precision be field-tested to assess practicability.