Litter- and ecosystem-driven decomposition under elevated CO2 and enhanced N deposition in a Sphagnum peatland

Peatlands represent massive global C pools and sinks. Carbon accumulation depends on the ratio between net primary production and decomposition, both of which can change under projected increases of atmospheric CO2 and N deposition. The decomposition of litter is influenced by 1) the quality of the litter, and 2) the microenvironmental conditions in which the litter decomposes. This study aims at experimentally testing the effects of these two drivers in the context of global change. We studied the in situ litter decomposition from three common peatland species (Eriophorum vaginatum, Polytrichum strictum and Sphagnum fallax) collected after one year of litter production under pre-treatment conditions (elevated CO2: 560†ppm or enhanced N: 3†g†m-2†y-1 NH4NO3) and decomposed the following year under treatment conditions (same as pre-treatment). By considering the cross-effects between pre-treatments and treatments, we distinguished between the effects on mass loss of 1) the pre-treatment-induced litter quality and 2) the treatment conditions under which the litters were decomposing. The combination between CO2 pre-treatment and CO2 treatment reduced Polytrichum decomposition by -24% and this can be explained by litter quality-driven decomposition changes brought by the pre-treatment. CO2 pre-treatment reduced Eriophorum litter quality, although this was not sufficient to predict decomposition. The N addition pre-treatment reduced the decomposition of Eriophorum, due to enhanced lignin and soluble phenols concentrations in the initial litter, and reduced litter-driven losses of starch and enhanced litter-driven losses of soluble phenols. While decomposition indices based on initial litter quality provide a broad explanation of quantitative and qualitative decomposition, they can only be taken as first approximations. Indeed, the microbial ATP activity, the litter N loss and resulting litter quality, were strongly altered irrespective of the compounds' initial concentration and by means of processes that occurred independently of the initial litter-qualitative changes. The experimental design was valuable to assess litter- and ecosystem-driven decomposition pathways simultaneously or independently. The ability to separate these two drivers makes it possible to attest the presence of litter-qualitative changes even without any litter biochemical determinations, and shows the screening potential of this approach for future experiments dealing with multiple plant species.