Abstract
In addition to the effect of litter quality (LQ) on decomposition, increasing evidence is demonstrating that carbon mineralisation can be influenced by the past resource history, mainly through following two processes: (i) decomposer communities from recalcitrant litter environments may have a wider functional ability to decompose a wide range of litter species than those originating from richer environments, i.e. the functional breadth (FB) hypothesis; and/or (ii) decomposer communities may be specialized towards the litter they most frequently encounter, i.e. the home-field advantage (HFA) hypothesis. Nevertheless, the functional dissimilarities among contrasting microbial communities, which are generated by the FB and the HFA, have rarely been simultaneously quantified in the same experiment, and their relative contributions over time have never been assessed. To test these hypotheses, we conducted a reciprocal transplant decomposition experiment under controlled conditions using litter and soil originating from four ecosystems along a land-use gradient (forest, plantation, grassland and cropland) and one additional treatment using 13C labelled flax litter allowing us to assess the priming effect (PE) in each ecosystem. We found substantial effects of LQ on carbon mineralisation (more than two-thirds of the explained variance), whereas the contribution of the soil type was fairly low (less than one-tenth), suggesting that the contrasting soil microbial communities play only a minor role in regulating decomposition rates. Although the results on PE showed that we overestimated litter-derived CO2 fluxes, litter-microbe interactions contributed significantly to the unexplained variance observed in carbon mineralisation models. The magnitudes of FB and HFA were relatively similar, but the directions of these mechanisms were sometimes opposite depending on the litter and soil types. FB and HFA estimates calculated on parietal sugar mass loss were positively correlated with those calculated on enzymatic activity, confirming the idea that the interaction between litter quality and microbial community structure may modify the trajectory of carbon mineralisation via enzymatic synthesis. We conclude that although litter quality was the predominant factor controlling litter mineralisation, the local microbial communities and interactions with their substrates can explain a small (< 5%) but noticeable portion of carbon fluxes