Continental fire density patterns in South America

DI BELLA, C.M., JOBBAGY, E.G. .; PARUELO, J.M. AND PINNOCK, S.

GLOBAL ECOLOGY AND BIOGEOGRAPHY

Año: 2006 p. 192 - 199

0960-7447

ABSTRACT Aims Quantification of the effects and interactions of natural and anthropogenic factors, including climate, canopy structure, land use and management conditions, on vegetation burning. The study of these relationships is fundamental to predict regional fire patterns and develop sound management and regulation policies for biomass burning at national and global levels. Location Southern South America, including Argentina, Brazil, Paraguay, Uru- guay, Bolivia and Chile. Methods Based on National Oceanic and Atmosphere Administration–Advance Very High Resolution Radiometer (NOAA–AVHRR) satellite images, we identified fires in southern South America with a daily frequency for two periods (1999/2000 and 2000/01) using a contextual fire detection algorithm and integrating the density of these fires at a monthly scale into a 0.5 × 0.5.äó“­êÀgrid. We combined vegetation and climate global databases and land use information from national census data to explore the relationship of these factors with fires across the region. Results The whole study region had a mean fire density of 0.10 and 0.05 fires km -2 year -1 in 1999/2000 and 2000/01, respectively, with extreme values as high as 1.37 in fires km -2 year -1 in Para State, Brazil. Water deficit estimates, derived from a climatic water balance, showed the better correlation with fire density (r = 0.28; P < 0.001; n = 4467), interacting strongly with land use. In areas with low agricultural use fire density increased with water deficit, whereas in highly agricultural areas this relation- ship was not observed. Agriculture significantly reduced fire density in prairies and savannas but increased its frequency in rain forests. Main conclusions These results suggest that agriculture prevents biomass burning in semiarid areas but enhances it in humid environments, where biomass accumulates at faster rates. Keywords Agricultural fires, fire occurrence, land cover changes, NOAA/AVHRR, remote sensing, water deficit.