LIMITS AND POTENTIALITIES OF METHODOLOGICAL APPROACHES AIMING TO ESTIMATE SOIL FUNCTIONS AND ECOSYSTEM SERVICES IN VIEW OF THEIR IMPLEMENTATION IN GEOSPATIAL? DECISION SUPPORT SYSTEMS

In the last decade, geoSpatial-Decision Support Systems (S-DSS), have become interesting and accessible tools to support end-users/stakeholders' decisions in agriculture and environmental protection at different spatial and temporal scales.The diffusion of S-DSS is very promising since they aim at ameliorating both land planning and land management by providing operational tools to a large community of end-users and stakeholders.In order to provide such answers these S-DSS systems typically employ different approaches and models to estimate functions and ecosystem services (ES) provided by soils. In general terms these approaches/models - ranging from well tested to new coming - refer to 3 different categories: (i) empirical based methods; (ii) physical based methods; (iii) hybrid methods having a mix of empirical and physically based models.In this specific contribution, we examine, through both literature review and by means of new simulations over case studies, a selection of the above approaches employed to estimate key soil hydrological functions (e.g. soil filtering capacity) and soil ES (e.g. crop production) in view of their potential incorporation into S-DSS systems.In addition, a supplementary sensitivity analysis will be performed by evaluating the impact of different levels of soil spatial information (e.g. spatial resolution, soil types) over model outputs under current or climate change conditions.Our findings show that simplified empirical estimates of soil functions and ecosystem services could result in poor and static information and in scarce replicability of DSS systems in new areas. Despite detailed information even at larger spatial extent are increasingly available (e.g. earth observation data), main concerns refer to an inappropriate combination of model complexity and data availability, as for the applied spatial and temporal scale that could be not able to satisfy the specific needs of the end-user.As outcome from this analysis the paper aims to provide a framework of potentialities and limitations of using the analysed approaches in developing new S-DSS at different scales and for different applications.