Cloud frequency correlates to plant species composition in the high Andes of Ecuador

Two data sets representing atmospheric Moisture are available for the high Andes of Ecuador, (i) cloud frequency obtained from weather satellites, and (ii) interpolated rainfall estimates obtained from global climate observations. We analyzed their correlation to vascular plant species composition at 18 Ecuadorian superparamo study sites. Of particular interest was whether cloud frequency Could be used as a proxy for precipitation. Cloud frequency had distinct seasonal and spatial variation among the sites. The spatial gradient Of Cloudiness was strongest during June-August and weakest around the equinoctials. Interpolated rainfall estimates also shoved seasonal and spatial variation among the sites, but there was no correlation between them and Cloudiness. Cloud frequency during June-August vas the only significant variable in the canonical correspondence analysis (CCA) of the species composition, whereas rainfall estimates, geology, presence of laciers, and size and altitudinal range of the superparamos were not significant. When the spatial components were filtered out from the species composition data by employing partial CCA, cloud frequency during December-February became the only significant variable. Our results suggest that cloud frequency data may be a useful tool in mountain ecology research, serving as an indicator of habitat humidity when exact precipitation data are lacking, (C) 2007 Published by Elsevier GmbH on behalf of Gesellschaft fur Okologie.