Avian rarity was investigated in ten high-altitude cloud forests in the Andes of Ecuador. Data on species compositions and abundances were obtained by a fully standardized method (standardization for area, altitude, habitat, effort and seasonality). The rare species were isolated from rank-abundance plots on the basis of the quartile definition of rarity. A positive correlation between mean abundances of species and number of sites occupied suggests that high-altitude bird species classified rare by abundance generally can also be classified rare by range. However, it is necessary to be cautious using this result in ranking conservation priorities since the generality is not obeyed by all species. Within the two abundance classes (contains one and two individuals, respectively) represented among the rare species, the one-individual class had significantly more species than the two-individual class. The quantitative rarity of taxa and ecological groupings produced similar results for all sites, while pairwise similarity of rare species between sites was very low. Together with the difficulty of identifying species that are truly rare by abundance, these results imply that sites selected for conservation preferably should be based upon a qualitative evaluation of lists of species referred to vulnerability categories such as endemic, restricted-range, CITES or IUCN threatened/near-threatened species. However, it is necessary also to incorporate other aspects of biodiversity to cover a full range of biotic diversity.
