The internal EJP SOIL project SERENA contributed to the evaluation of soil multifunctionality aiming at providing assessment tools for land planning and soil policies at different scales. By co-working with relevant stakeholders, the project provided co-developed indicators and associated cookbooks to assess and map them, to report both on soil degradation, soil-based ecosystem services and their bundles, under actual conditions and for climate and land-use changes, at the regional, national, and European scales. The map is the result of applying the SOC loss cookbook developed in SERENA/EJP-Soil. It is based on Tuscany Region SOCdatabase (Gardin et al., 2021). Not freely available (Lamma Consortium). More information and requests to: info@lamma.toscana.it. Three versions of the map were produced. In versions v1 and v2, the dataset corresponds to a map of SOC concentrations for the Tuscany Region produced using a geostatistical approach. Ordinary kriging (Chilès & Delfiner, 2012) coupled with local geostatistics (LGS) was used to map SOC concentration (%) in the topsoil (0-30 cm) at a regional scale in the Tuscany Region (Italy). The map has a 50 m spatial resolution. In version v3, a different approach to applying SOC loss cookbook was used. Particularly, Random Forest regression algorithm with numerous spatial covariates was applied for mapping SOC content (%) in the topsoil (0-30 cm) at regional scale in Tuscany Region (Italy). The map has a 100 m spatial resolution.

The data are derived from the calculation of indicators based on a standard methodology established as part of the EJP Soil SERENA programme. Please keep in mind that: It is the result of a modelling exercise and does not necessarily reflect reality. Despite the efforts made to provide reliable data, the results may contain inconsistencies, depending in particular on the raw data available and level of accuracy of the techniques chosen and their prior knowledge . It is necessary to consider how the results have been obtained in order to decide on their relevance in relation to the intended purpose of reuse. These results are interesting from a scientific point of view, but their use for environmental management and policy issues should be done keeping the previous aspects in mind and complementing when necessary the provided results with the best available data. ==> Finally, it is the responsibility of the users of this information to decide whether it is appropriate to use these data and whether the data meet their needs. The authors of this resource can in no way be held responsible for the results obtained from the use of this data.

In versions v1 and v2 local geostatistics (LGS) approach (Chilès & Delfiner, 2012) was used as an alternative to a global model of spatial dependence. According the LGS approach, the geostatistical parameters involved in variogram-based modelling were locally (within a 20-km grid mesh) optimized ensuring a better adequacy between the geostatistical model and the data.