In this study we performed a 1-year litter-manipulation experiment to examine how soil CO2 efflux was altered by the amount of fresh litter. Three treatments were applied: litter exclusion (E), control (C, natural amount: 486 g m−2) and litter addition (A, twice the natural amount: 972 g m−2).
Litter decomposition rate was not affected by fresh litter amount. However, the addition or exclusion of fresh litter quickly increased or decreased total soil CO2 efflux (FS) significantly, but the relative contribution of fresh litter to total soil respiration remained unchanged between the C and A treatments, as determined by laboratory measurements. Variation in FS among treatments was not related to modification of its temperature sensitivity which was not affected by fresh litter amount (Q10: 3.5 for E, 3.2 for C, 3.6 for A). While litter exclusion was the main cause of the FS decrease in the E treatment, only 68 % of FS was directly attributable to litter addition in the A treatment. The remaining 32 % of FS in the A treatment was related to a real priming effect that appeared to be a long-lasting phenomenon. This priming effect lasting over 1 year may be related to a continuous release of organic compounds from litter to soil because of the progressive decomposition of leaf litter. Q10 estimates and isotopic data lead to the hypothesis that the priming effect corresponded to the activation of the whole soil system.
As a consequence, the increase in ecosystem productivity may lead, via an increase in the amount of litter, to an increase in carbon turnover in soil. Further labelling experiments involving high-frequency carbon stable isotope measurements of CO2 efflux would help to clarify the relative importance of bulk soil and rhizosphere in the priming effect.