Decomposition of European beech (Fagus sylvatica) litter : Combining quality theory and 15N labelling experiments

A multi-site and multi-age set of in situ 15N-labelled litter decomposition experiments was carried out in various European beech forests. The sites presented contrasted situations of climate and humus type that were representative of northern Europe temperate forests. 15N-labelled leaf litter was produced by urea spraying in an experimental beech plot, and was used in litter decomposition experiments over a 3-6-year period, depending on the site. Schemes of decomposition covered a wide range of situations, some of them not expected, like high decay rate on Moder humus forms or slow decomposition on Mull humus form. The objectives of this study were to determine how nitrogen release and/or incorporation rates could characterize litter decomposition dynamics. A carbon and nitrogen coupled model of decomposition based on the quality theory was used to fit the given data. High correlation coefficients and unbiased fitting were obtained both for carbon ðR2 ¼ 0:96; pp0:0001; n ¼ 392Þ and nitrogen ðR2 ¼ 0:86; pp0:0001; n ¼ 367Þ. The theoretical concept of quality was somewhat lightened and related to decomposer activity: low rates of quality decrease were associated with high decomposition rates, showing that when decomposers are sufficiently supplied with labile carbon from the humus they decompose litter more completely than when they are carbon limited. Provided that the degrading substrate stoichiometry was kept constant (i.e. C loss / N loss ratio), the same litter decomposition could be processed at different rates. Taking seasonal variations of the parameters into account is likely to improve our understanding, but further experiments at more frequent intervals are needed. (Résumé d'auteur)