Ecosistemas 17 (3): 64-78. Septiembre 2008.

 

La caracterización de la heterogeneidad espacial de los ecosistemas: el uso de atributos funcionales derivados de datos espectrales.

 

Néstor Fernández, Gervasio Piñeiro

 

 

 

Most remote sensing approaches to ecosystem classification continue to rely on structural attributes like vegetation physiognomy and landform, failing to incorporate key functional properties of energy and matter exchange. Here we show an application of functional classifications for ecosystem characterization using the Normalized Difference Vegetation Index (NDVI), albedo and surface temperature as estimated from remote sensing data. We illustrate the development of ecosystem functional classifications in Cabo de Gata and Doñana, two Mediterranean regions of high conservation concern under different environmental conditions. Both regions displayed a complex spatial pattern in functional class distributions only partly related to vegetation structure and landform. We identified 10 ecosystem functional classes in Cabo de Gata, with only two closely associated to specific vegetation units. In addition, seven different landform units present in the region included a variety of ecosystem classes. Out of eight different functional classes recognized in Doñana, only two were closely associated to particular landform units within the highest protected area, whereas a combination of landform and land use intensity explained the distribution of other four functional classes. We propose that ecosystem functional classifications based on remote sensing data are of key importance for ecosystem characterization and management, therefore complementing more traditional approaches solely based on structural attributes.