Predicting alfa, beta and gamma diversity from physiognomic and physical indicators as a tool for ecosystem monitoring

Abstract

We searched for predictive models for alpha, beta and gamma plant diversity based in easy to measure field indicators. The study was conducted on the upper belt of the Córdoba mountains (Argentina). We established 222 permanent plots of 4 × 4 m distributed on sites with different physiognomy, topography and management. At each plot we measured physical and physiognomic indicators and recorded the presence of all vascular plants. We estimated alpha diversity as the number of species detected in a plot, beta diversity as the floristic dissimilarity between two plots, and gamma diversity as the number of species detected in a landscape. Through linear regression we found predictive models for alpha and pair-wise beta diversity. Then we analysed if predicted average alpha and beta diversity were good estimators of gamma diversity. We recorded a total of 288 species (5–74 species per plot). Alpha diversity was highest in sites on shallow soils with high structural richness (i.e. high number of cover categories), half covered by lawns, at sunny slopes and rough landscapes (r2 = 0.66). For beta diversity, the difference between plots in structural richness and in cover of thick tussocks grasses and lawns were the best predictors (r2 = 0.45). For different sets of simulated landscapes, gamma diversity was well explained by predicted average alpha and beta diversity, plus the sampling effort (r2 = 0.92). We concluded that using easy to measure field indicators it is possible to estimate plant diversity at different levels with a good accuracy.