Résumé
Nitrogen movement among microsites is thought to be an important control on patterns of ecosystem-level N cycling. In particular, N transfer between decomposing leaves may explain why litter mixtures sometimes decompose differently than would be predicted from the decomposition dynamics of each species separately. We evaluated how N moves between leaves of differing N status in leaf-pair microcosms. We collected litter from six species of trees from French Guiana (three with high N concentration, three with low) and
15N-labeled the microbial communities growing on each species. We then established microcosms with one labeled and one unlabeled leaf in a fully factorial design (each species with every species,
15N on each species) and measured
15N transfer over 28 days. There was substantial transfer of the
15N label in all cases, averaging between 15% and 30% of the
15N originally on the labeled leaf. Net N transfer from high-N to low-N leaves resulted from greater gross
15N transfer from high-N to low-N leaves than in reverse. Gross
15N transfer was controlled entirely by the N status of the source leaf, rather than by the difference in N-status of the leaves or by the characteristics of the sink leaf. For example, as much
15N was transferred from a high-N leaf to another high-N leaf as to a low-N leaf. These results support the assumption from N mineralization theories that microbes at a specific site have first access to that N and therefore control how much N is available to move to other microsites in the soil system. The strength of the gradient between microsites may then control the rate at which available N moves, but not how much N is available to move. If N transfer among different litter species is important for synergistic effects on decomposition of litter mixtures it would not be driven by the N gradient as is often hypothesized, but by the characteristics of the source leaf.