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

CINGOLANI A. M., VAIERETTI V. M., GURVICH D. E., GIORGIS M. A. & CABIDO M.

BIOLOGICAL CONSERVATION

ELSEVIER SCI LTD

Año: 2010 vol. 143 p. 2570 - 2577

0006-3207

We searched for predictive models for alpha, beta and gamma plant diversity based in easy to measurefield indicators. The study was conducted on the upper belt of the Córdoba mountains (Argentina). Weestablished 222 permanent plots of 4  4 m distributed on sites with different physiognomy, topographyand management. At each plot we measured physical and physiognomic indicators and recorded thepresence 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 speciesdetected in a landscape. Through linear regression we found predictive models for alpha and pairwisebeta diversity. Then we analysed if predicted average alpha and beta diversity were good estimatorsof gamma diversity. We recorded a total of 288 species (5–74 species per plot). Alpha diversity was highestin sites on shallow soils with high structural richness (i.e. high number of cover categories), half coveredby lawns, at sunny slopes and rough landscapes (r2 = 0.66). For beta diversity, the differencebetween 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 predictedaverage alpha and beta diversity, plus the sampling effort (r2 = 0.92). We concluded that using easyto measure field indicators it is possible to estimate plant diversity at different levels with a goodaccuracy.