Evaluation of caribou Rangifer tarandus groenlandicussurvey methodology in West Greenland

Abundance estimates are important to management of most harvested species of wildlife. In West Greenland, recent estimates of barren-ground caribou Rangifer tarandus groenlandicuspopulation size have been derived from aerial surveys conducted in early March of numerous short (7.5 km) transects that focused on obtaining high detection probabilities. The resultant study area coverage was low (= 1.6%), in part due to the survey design. In this article, we conducted a critical review of the current West Greenland caribou survey methodology using data from past surveys and recent GPS collar data, and present recommendations to improve the methodology. On an annual basis, movement rates of collared females were lowest in March, supporting survey timing. March distribution of collared caribou, however, differed markedly between 2009 and 2010, indicating that stratification flights prior to each survey are required to produce the most accurate and precise estimates. A viewshed analysis in GIS supported the use of a 300-m strip width, but demonstrated that the current 15-m survey flight altitude resulted in 4-5% availability bias due to the portion of the strip width hidden by topography and out of sight of observers, and a corresponding nil detection probability for caribou in these areas. A 30-m or 45-m flight height may be more appropriate to reduce the availability bias in this rugged terrain. Examination of the population composition data collected during and after abundance estimates suggested that robust calf:cow and bull:cow ratio data could be obtained with less sampling effort distributed proportionate to the population density. We suggest that systematic strip transects should be considered to increase survey coverage; this design would increase survey efficiency (ratio of helicopter time to coverage) and inherently increase precision. Distance sampling collected by group would be an improvement over the current negatively biased, transect-total method to calculate detection probabilities. Managers should ensure that sufficient resources are available to obtain robust estimates of abundance and composition of West Greenland caribou. These recommendations may be applicable to other areas in which ungulate populations exist in heterogeneous habitats with low sightability.