Geospatial analysis of ecological vulnerability of coffee agroecosystems in Brazil

摘要

Geographical information systems and statistics were used to characterize the climatic vulnerability of coffee agroecosystems in Brazil. Average annual mean air temperature, mean air temperature of the coldest month, and moisture index were used to characterize climatic vulnerability for Coffea arabica and Coffea canephora species cultivation based on high-resolution interpolated climate surfaces from the average of the period of 1950 to 2000 and the A2 2080 climate change scenario. Soil vulnerability for coffee cultivation was derived from soil classes 1:5,000,000 scale and slope estimated from SRTM digital elevation model at 90 m spatial resolution. Coffee tree production at municipal district level from 1990 to 2006 was used to validate the obtained results of the vulnerability of coffee agroecosystems. A coffee tree index was developed using the principal components technique, based on variables related to coffee yield, coffee harvested and coffee cultivated areas. The coffee tree fraction index explained 87.0 % of coffee tree fraction and was classified in five levels inside the municipal district boundaries using natural breaks method. Based on the adopted methodology, it was possible to observe relationship between coffee tree cultivation areas and coffee climatic vulnerability in Brazil for the scenario of 1950 to 2000. Considering A2 2080 scenario of climate change, suitable areas for coffee cultivation were moved to the states of the south and southeast of Brazil.