The recent decoupling of agricultural production and deforestation in the southern Amazon has been made possible thanks to (1) the adoption of intensive agricultural practices, including irrigation, and (2) the diversification of economic activities, including fish farming. Whereas this new agricultural model has brought out positive results to contain deforestation, it also implied new pressures on the environment, and especially on water resources. Many small artificial water reservoirs have been built with different uses, e.g. crop irrigation, energy generation, fish farming or livestock watering. In this paper, we introduce a method to automatically map small water bodies based on time series of Landsat images. The method was tested in the municipality of Sorriso (state of Mato Grosso, Brazil). The statistical results (Overall Accuracy = 0.872; Kappa index = 0.745) validated the efficiency of the methodology although the spatial resolution of Landsat images limited the detection of very small and linear reservoirs. In Sorriso, we estimated that the cumulated area and the number of small water reservoirs increased more than tenfold (from 153 to 1707 ha) and fivefold (86 to 522), respectively, between 1985 and 2015. We discuss the numerous socio-environmental implications raised by the cumulated impacts of these proliferating small reservoirs. We conclude that integrated whole-landscape approaches are necessary to assess how anthropized hydrosystems can counteract or exacerbate the socio-environmental impacts of deforestation and intensive agriculture.

No abstract

The agricultural expansion in the Southern Brazilian Amazon has long been pointed out due to its severe impacts on tropical forests. But the last decade has been marked by a rapid agricultural transition which enabled to reduce pressure on forests through (i) the adoption of intensive agricultural practices and (ii) the diversification of activities. However, we suggest that this new agricultural model implies new pressures on environment and especially on water resources since many artificial water reservoirs have been built to ensure crop irrigation, generate energy, farm fishes, enable access to water for cattle or just for leisure. In this paper, we implemented a method to automatically map artificial water reservoirs based on time series of Landsat images. The method was tested in the county of Sorriso (State of Mato Grosso, Brazil) where we identified 521 water reservoirs by visual inspection on very high resolution images. 68 Landsat-8 images covering 4 scenes in 2015 were pre-classified and a final class (Terrestrial or Aquatic) was determined for each pixel based on a Dempster-Shafer fusion approach. Results confirmed the potential of the methodology to automatically and efficiently detect water reservoirs in the study area (overall accuracy = 0.952 and Kappa index = 0.904) although the methodology underestimates the total area in water bodies because of the spatial resolution of Landsat images. In the case of Sorriso, we mapped 19.4 km 2 of the 20.8 km 2 of water reservoirs initially delimited by visual interpretation, i.e. we underestimated the area by 5.9%.