The global demand for wood is growing and the sustainability of forests has become an increasingly important issue. The first metre of soil represents the reserve of nutrients cations (K, Ca, Mg) for trees. The depletion of these elements in the soil gradually leads to a decrease in soil fertility, which can have a strong impact on the development and health of forests. The aim of this study was to better understand the source and dynamics of nutrients in forested soils from a base poor environment by combining mineralogic, chemical, isotopic and numerical tools through an experimental approach. We designed batch experiments to follow the leaching of soils from two experimental plots (under spruce and under beech from the Strengbach catchment, Vosges Mountain, NE France). The studied elements show highly contrasting behaviours. Mg release results from a mixing between cationic exchange, mineral dissolution (mainly smectite) and organic matter mineralization. The dynamics of K are underestimated by modelling, which may be attributed either to an unknown source of K or to an overestimation of the secondary phase precipitation. The Sr isotopic approach, used as a proxy to identify the sources of Ca, shows that the exchangeable complex supplies nutrients first and rapidly to the solution, followed by the clay compartment and at the very end, the bulk soils. Na is mainly controlled by albite dissolution. The release of Ca is significantly lower than that of other cations as Mg or K. Our results suggest that the exchangeable Ca, estimated through classical extractions, is only partly available for plants, due to strong chemical binding with organic matter and/or the occurrence of Ca-chemical bridges between clay and organic matter. Sustainable management should thus consider that exchangeable Ca, conventionally determined, does not necessarily provide an accurate picture of the real availability of Ca and therefore of the fertility of soils, with a risk of overestimation of nutrient reservoir for trees. Calcium tree nutrition at these sites has to rely on the long term on atmospheric deposition and an efficient recycling through biological cycling. This is also true for th other nutrient cations, but to a lesser extent since soil intrinsic fluxes (i.e., weathering, mineralisation of OM…) are not null and may thus supply the exchangeable pools and participate substantially to tree nutrition. Harvesting and biomass exportation could rapidly impact the biological cycle and threaten the sustainability of these ecosystems, similar to changing tree species that may influence the nature of the litter, its turnover and the input of nutrients to the soil.