From Field Data to Ecosystem Services Maps: Using Regressions for the Case of Deforested Areas Within the Amazon
Abstract
Quantifying and mapping ecosystem services (ES) is seen as one way to improve decision making and land management to better integrate environmental issues. This study aimed to characterize ES supply in deforestation context where an improvement of scientific knowledge should help develop more efficient environmental management. For three case studies in the Brazilian Amazon impacted by deforestation, seven indicators of potential ES supply were mapped at a spatial resolution of 30 × 30 m: biodiversity index (indicator of food web support); richness of pollinators (pollination); index of soil chemical quality (support to production); water available for plants (water regulation); soil carbon stocks (support to production and climate regulation); rate of water infiltration into the soil (soil erosion control); and vegetation carbon stocks (climate regulation). To map these indicators, in situ measurements of ES for 135 sampling points and remote sensing data were linked using regression methods. These methods were used to predict ES values and identify environmental factors that influence ES supply. The resulting maps help in understanding the influence of environmental factors on ES spatial distribution within the sites. The analyses illustrate the influence of land-use changes on ES supply and the role of context effects due to the heterogeneity of the biophysical environment, the temporality of deforestation and/or their diversified sociopolitical contexts. From a methodological viewpoint, the study highlights the importance of choices inherent in all cartographic practices and that need to be considered, especially in the context of rendering ES maps operational.