What drives elevational patterns of diversity? A test of geometric constraints, climate and species pool effects for pteridophytes on an elevational gradient in Costa Rica

Aim We studied pteridophyte species richness between 100 m and 3400 m along a Neotropical elevational gradient and tested competing hypotheses for patterns of species richness. Location Elevational transects were situated at Volcan Barva in the Braulio Carrillo National Park and La Selva Biological Station (100-2800 m) and Cerro de la Muerte (2700-3400 m), both on the Atlantic slope of Costa Rica, Central America. Method We analysed species richness on 156 plots of 20 x 20 m and measured temperature and humidity at four elevations (40, 650, 1800 and 2800 m). Species richness patterns were regressed against climatic variables (temperature, humidity, precipitation and actual evapotranspiration), regional species pool, area and predicted species number of a geometric null model (the mid-domain effect, MDE). Results The species richness of the 484 recorded species showed a hump-shaped pattern with elevation with a richness peak at mid-elevations (c. 1700 m). The MDE was the single most powerful explanatory variable in linear regression models, but species richness was also associated strongly with climatic variables, especially humidity and temperature. Area and species pool were associated less strongly with observed richness patterns. Main conclusions Geometric models and climatic models exclusive of geometric constraints explained comparable amounts of the elevational variation in species richness. Discrimination between these two factor complexes is not possible based on model fits. While overall fits of geometric models were high, large- and small-ranged species were explained by geometric models to different extents. Species with narrow elevational ranges clustered at both ends of the gradient to a greater extent than predicted by the MDE null models used here. While geometric models explained much of the pattern in species richness, we cannot rule out the role of climatic factors (or vice versa) because the predicted peak in richness from geometric models, the empirical peak in richness and the overlap in favourable environmental conditions all coincide at middle elevations. Mid-elevations offer highest humidity and moderate temperatures, whereas at high elevations richness is reduced due to low temperatures, and at low elevations by reduced water availability due to high temperatures.