. Tomassen, cm) with a lower ratio 305 in the Fertilized plots than in the Control ones (Fig.4). and the Fertilized mesocosms and no 311 stimulation of Molinia growth was observed. This is in agreement 312 with the results of Tomassen et al. (2003) who found an effect of 313 N addition on M. caerulea biomass only after 3years of N input. In 314 addition, the different forms of N dissolved in peat water were 315 not affected by the N addition treatment (Table 1). However, N 316 addition induced an increase in the concentration and stock of N 317 in the living parts of Sphagnum (the first two layers of the peat 318 mesocosms: 0-0.5 cm and 0.5-2.5 cm; Table 2, Fig. 2). Thus, the 319 lack of a short-term effect of N fertilization on M. caerulea growth 320 can be explained by the high capacity of Sphagnum mosses to 321 retain N deposition (van Breemen, 1995). Indeed, Sphagnum 322 species can capture the atmospheric N supply, limiting its 323 availability for the surrounding vascular plants (van Breemen, Sphagnum mesocosms than in Sphagnum + Molinia 301 mesocosms (Table S2 and Fig. 3). The variability of peat 302 density between vegetation did not affect the stoichiometry of 303 the peat. The C/N ratio was only affected by the N treatment 304 in the top two layers, vol.324, 1995.