Simultaneous estimation of actual litter enzymatic catalysis and respiration rates with a simple model of C dynamics in Sphagnum - dominated peatlands . C dynamics

SG wrote the article, developed and implemented the model and participated in the design of the experiment, the data acquisition and the discussion. NG was in charge of the day-today management of the experiment, the regular data acquisition, participated in the design of the experiment and the analysis of the data. FLD participated in the design of the experiment and the discussion. A-JF contributed to the writing of the article, participated in the discussion and the improvement of the modelling. FD participated in the implementation of the carbohydrate and the SUVA280 analyses and the discussion. NL participated in the implementation, the quantification and discussion of the carbohydrate analyses. Abstract C dynamics in soils plays an important role in the interactions between climate and terrestrial ecosystems. Given the importance of the soil C pool in future climate scenarios, accurate models are required which can then be integrated into global models. On the one hand, models of soil C dynamics can be very simplistic in that only solid organic matter (OM) is taken into account with constant rates. Such models can miss important features of soil C dynamics such as enzymatic activity and CO 2 production. On the other hand, some models are too complicated to be experimentally calibrated, tested and widely used. We propose a model of soil C dynamics that (i) integrates all C fluxes from the solid to the gaseous form, and (ii) can be easily implemented experimentally. Because peatlands are important C stores that are experiencing vegetation changes, the model was tested on experimental results obtained with peatland litter: Sphagnum cuspidatum (autochtonous vegetation), Molinia caerulea and Betula spp (both invasive species). Our model was able to accurately describe the early stages of C dynamics in litter especially when only one pool of OM was used and when catalysis and respiration rates were allowed to decrease with time. Our model is efficient in assessing the entire actual catalysis rate. This is a notable achievement as usually, enzyme activity is measured on specific enzymes in conditions often far from actual ones. Our model appeared to be sufficiently robust and worthy of development, keeping in mind that it should remain experimentally testable.