Gradient analysis of soil-plant interactions from the alpine-nival ecotone to the snowline on slopes of the Central Great Caucasus (Kazbegi Region, Georgia)

Alpine ecosystems are especially sensitive to climatic changes which affect the relationships among glaciers, snow, vegetation and soils. Our aim was to examine how the variation in the abiotic environment affected soil properties and plant species distribution at regional and local scales. We sampled soil and vegetation along two transects set on the opposite-facing slopes (North versus South), from the alpine-nival ecotone to the snowline (Central Great Caucasus, Kazbegi, Georgia). We measured also soil temperature and controlled for the slope inclination. Multivariate ordination methods were used to link abiotic factors, soil properties and plant species distribution along the gradients. We found that ordination models were better resolved when soil properties were used as environmental variables instead of abiotic ones such as elevation, inclination and slope aspect. Soil pH and plant available potassium were the best predictors of plant species distribution in these habitats. We conclude that the models that account for the role of soils as a mediator between the abiotic environment and vegetation can more accurately describe plant species distribution at local and regional scales: a potentially important amendment with implications for the monitoring of the effects of climate change on vegetation at least in high mountain systems.