Current distributions of ecosystem functional types in temperate South America

PARUELO, JM; EG JOBBÁGY; OE SALA

ECOSYSTEMS (NEW YORK. PRINT)

Springer

Lugar: New York; Año: 2001 vol. 4 p. 638 - 698

1432-9840

We described, classified, and mapped the functionalheterogeneity of temperate South America usingthe seasonal dynamics of the Normalized DifferenceVegetation Index (NDVI) from NOAA/AVHRR satellitesfor a 10-year period. From the seasonalcurves of NDVI, we calculated (a) the annual integral(NDVI-1), used as an estimate of the fraction ofphotosynthetic active radiation absorbed by thecanopy and hence of primary production, (b) therelative annual range of NDVI (RREL), and (c) thedate of maximum NDVI (MMAX), both of whichwere used to capture the seasonality of primaryproduction. NDVI-1 decreased gradually from thenortheastern part of the study region (southernBrazil and Uruguay) toward the southwest (Patagonia).High precipitation areas dominated by rangelandshad higher NDVI-1 and lower RREL valuesthan neighboring areas dominated by crops. Therelative annual range of NDVI was maximum forthe northern portion of the Argentine pampas (highcover of summer crops) and the subantarctic forestsin southern Chile (high cover of deciduous treespecies). More than 25% of the area showed anNDVI peak in November. Around 40% of the areapresented the maximum NDVI during summer. Thepampas showed areas with sharp differences in thetiming of the NDVI peak associated with differentagricultural systems. In the southern pampas, NDVIpeaked early (October–November); whereas in thenortheastern pampas, NDVI peaked in late summer(February). We classified temperate South Americainto 19 ecosystem functional types (EFT). Themethodology used to define EFTs has advantagesover traditional approaches for land classificationthat are based on structural features. First, the NDVItraits used have a clear biological meaning. Second,remote-sensing data are available worldwide. Third,the continuous record of satellite data allows for adynamic characterization of ecosystems and landcoverchanges.