Résumé : The two main sources of Ca in forest ecosystems are mineral weathering release and atmospheric inputs. We use the87Sr/86Sr isotopic ratio (Sr is used as a proxy for Ca) to determine the Ca contribution from atmospheric deposition input intwo forest ecosystems (beech stands) growing on soils formed from parent materials with contrasting total Ca contents andisotopic ratios: Pleistocene loess in Central Belgium (acid leached soil) and Lower Devonian shales in the Ardenne Massifof High Belgium (ochreous brown earth). The 87Sr/86Sr ratio and the Ca and Sr contents were measured in bulkprecipitation, in vegetation (beechwood growth rings and leaves) and in main soil horizons (total, acid-extractable and labilepools). The relative contributions of atmospheric input and soil mineral weathering to vegetation were calculated usingmixing equations.Calculations based on Sr isotope ratios of bulk precipitation (end-member 1; 87Sr/86Sr close to ocean water: 0.7090),0.1 M HCl-extractable soil fraction (end-member 2) and beech wood (mixing compartment) indicate that about 43%(Central Belgium) and 39% (Ardennes) of Sr uptake originates from atmospheric inputs. For Ca, the contributions ofatmospheric input are 75% and 78%, respectively. These estimations are, however, very sensitive to the choice of theappropriate 87Sr/86Sr and Sr/Ca ratios for the weathering end-member. The isotopic composition of the mineral sourceis estimated from the soil mineralogical composition and the relative weatherability of the Sr-bearing minerals. Due tosoil processes (mineral weathering, element recycling) and geomorphological events (addition of allochtonous mineralsto the autochtonous pool inherited from the local bedrock), the distribution of minerals in both soil profiles isheterogeneous and varies from horizon to horizon. Which horizons are relevant and which soil pool (total soil, acidextractablefraction or labile pool) should be selected for isotopic measurement of the weathering end-member, istherefore open to uncertainty. The choice of the appropriate 87Sr/86Sr and Sr/Ca ratios for the weathering end-memberis discussed.Our results emphasize the importance of the atmospheric contribution of Ca for tree mineral nutrition in the studiedforest stands, suggesting that these ecosystems on acid soils are sensitive to chemical changes in the atmosphericenvironment, e.g. acid depositions associated with decreasing input of atmospheric cations. Our results support thereforethe conclusion of Hedin et al. [Hedin, L.O., Granat, L., Likens, G.E., Bulshand, T.A., Galloway, J.N., Butler, T., Rodhe,H., 1994. Steep declines in atmospheric base cations in regions of Europe and North America. Nature, 367: 351–354],suggesting that the atmospheric base cation trends in Europe and North America are ecologically relevant on the scale ofdecades for poorly buffered ecosystems. This is a serious possibility of depletion of the soil available cation pool of theHigh Belgium beech forest, which grows on a bedrock with extremely low levels of total calcium.