Can pellet counts be used to accurately describe winter habitat selection by moose Alces alces?

Pellet counts have been used to measure habitat selection of a variety of ungulate species often under the assumption that this method gives an unbiased sample of the true distribution of the species among habitats. The validity of this method has been questioned and comparisons with other methods have sometimes showed divergent results. We tested the validity of pellet group distribution as a tool for habitat selection studies by comparing the distribution of moose pellet groups in four different forest age categories (forest age <30, 31–60, 61–90, >90 years) and mire with GPS positions from collared moose (Alces alces). Sample plots (n = 531) were cleaned from pellet during the fall 2007 and the number of new pellet groups were counted in spring 2008, thus resulting in a defined period of accumulation. In addition, pellet groups were counted in paired, uncleaned, control plots. GPS data from 15 collared moose monitored during the same period were used for comparison with habitat composition and distribution of pellet groups. Both the distribution of pellet groups and GPS positions differed significantly from the habitat composition within the study area. Young forest stands (<30 years) were significantly more used than both forests >30 years and mire. The selection by moose, as calculated by Manly’s alpha, showed identical ranking among habitat classes for cleaned sample plots and GPS data whereas uncleaned plots showed a shifted rank order for two of the habitat classes. We conclude that pellet group counts can be used to accurately predict habitat use for moose during winter.