Reuse of sediments from surface reservoirs for agricultural production in the Brazilian semiarid region: plant growth, spatio-temporal variability, economic analysis, regulatory barriers and mapping of its attributes by remote sensing

Abstract

This study explored the potential of sediment reuse from surface reservoirs located in the Brazilian semiarid region as a soil conditioner for agricultural production, to reduce the demand for high-consumption chemical fertilizers through sediment replacement. The thesis is divided into three chapters organized in the format of scientific papers. The first chapter (I) evaluates: the heterogeneity of sediments physicochemical properties at regional scale and the effect of the substrate containing sediment on the growth and physiology of maize. In general, the results showed that reservoirs sediments in semiarid regions have higher nutrient contents compared to soil. Sediments increased the leaves' chlorophyll content, photosynthesis rate and growth in relation to the control treatment containing only soil. Additionally, there weren’t significant differences between the biomass production and plants' nutrients extraction with the largest dose of the most enriched sediment when compared to the treatment with chemical fertilizer. In the chapter two (II) the sediment variability, the economic and environmental benefits and regulatory barriers for applying the practice were analyzed. We also conducted a cost-benefit analysis of reservoir sediment reuse, in face of the latest global economic crisis. The analysis of the reservoirs’ volume dynamics showed that the sediment has been exposed frequently and for long time periods, enabling its excavation. Spatio-temporal variability of the physical-chemical characteristics of the sediment may be an obstacle to extending the practice. Chapter three focused on mapping physicochemical properties of soils and reservoirs sediments from the Brazilian semiarid region by multi- and hyperspectral data. We concluded that the combination of reflectance spectroscopy with multivariate statistical techniques seems to be an effective approach for predicting electrical conductivity, organic carbon and clay content of soils and reservoir sediments in semi-arid regions.