Water balance partitioning for ecosystem service assessment. A case study in the Amazon

Rainforests ensure fundamental water-related ecosystem services that are currently threatened by land-use change, in particular deforestation. Quantitative assessments of water-related ecosystem services have traditionally focused on the benefits linked to direct water availability for humans. Under this perspective, forests have been considered as water consumers, due to high interception and transpiration rates that reduce water flows available to downstream human activities. In contrast, their role as water suppliers through transpiration from tree canopies has often been neglected. Integrating this second perspective into assessments of water-related ecosystem services from forests and other land covers is key to providing a comprehensive support to decision-making processes on land-use change. In addition, specific indicators are necessary to incorporate the contribution of the different water balance components into ecosystem service assessment. In this paper, we investigate the use of water balance partitioning for the assessment of water-related ecosystem services. We use three water balance models (two existing ones and one developed in this study) and compare, via model simulations, water balance partitioning and water-related ecosystem services of a forested and a deforested area of the Brazilian State of Rondônia. Then, we propose a set of indicators that, based on the outputs of the models, provide a multidimensional characterization of different land-use types. Regardless of the model used, the values of these indicators for the selected case study consistently point out the key role of vegetation in regulating the water cycle. Forests act as effective indirect water suppliers for human activities and are key players in moisture recycling, thanks to high evapotranspiration rates (about 68% of total precipitation). Deforested areas, instead, act as direct suppliers of water flows for human activities due to higher drainage rates (about 50% of total precipitation). The proposed methodology helps highlight the importance of comprehensive water-related ecosystem services assessments. Moreover, the indicators quantitatively support the impact assessment of land-use change on the different processes involved in the water cycle and on human activities relying, directly or indirectly, on these processes