IMPROVING THE CHOICE OF HYDROCLIMATIC VARIABLES FOR MODELING VEGETATION RESPONSES TO MOISTURE IN THE ESPINHAÇO RANGE (BRAZIL)

The capacity of ecological models to better capture and predict ecosystems variations and changes is dependent on the choice of its environmental inputs’ variables, being precipitation the mostly used hydroclimatic variable. Yet, water available to plants is a function of rainfall and atmospheric evaporative demand. We assessed which hydroclimatic variable better explains variations in vegetation productivity in a seasonally dry tropical mountain system. We modeled NDVI temporal responses of different vegetation types to Climatic Water Deficit (CWD) and Precipitation, answering the following questions: 1) Are the responses of vegetation to different hydroclimatic variables specific for each vegetation type? 2) If so, which hydroclimatic variable better explains vegetation productivity for each vegetation type? We found that seasonally dry vegetation types were more responsive to CWD, while moist vegetation productivity was poorly explained by all hydroclimatic variables. The timing of responses of vegetation to CWD or precipitation varied according to site specificity.