OOS 21-2 - Carbon quality as a control on decomposition in tropical forests

Tuesday, August 7, 2012: 1:50 PM
A103, Oregon Convention Center
Stephan Hättenschwiler1, Nicolas Fanin2, Sandra Barantal3, Heidy Schimann4, Johanne Nahmani3, Sylvain Coq3 and Nathalie Fromin5, (1)Centre of Functional Ecology and Evolution, Montpellier, France, (2)UMR 614 Fractionnement des AgroRessources et Environnement, Institut National de la Recherche Agronomique, Reims, France, (3)Centre d'Ecologie Fonctionnelle et Evolutive, CEFE-CNRS, Montpellier, France, (4)INRA-UMR ECOFOG, Kourou, French Guiana, (5)Centre d'Ecologie Fonctionnelle et Evolutive, Centre National de la Recherche Scientifique, Montpellier, France
Background/Question/Methods

Heterotrophic decomposers contribute critically to ecosystem carbon (C) and nutrient cycling. Decomposer abundance and activity are mainly determined by environmental factors and the quality of plant litter as their major food source. Under near optimal temperature and moisture conditions such as in humid tropical forests, high metabolic rates associated with high requirements in energy may lead to decomposer limitation by accessible C substrates. We tested this hypothesis with a series of decomposition experiments in the field of a nutrient-poor Amazonian rainforest in French Guiana and under controlled laboratory conditions including experimental fertilization with C, nitrogen (N), and phosphorus (P).

Results/Conclusions

The high diversity of about 150 co-occurring tree species at our study site resulted in a high variation in tree species-specific leaf litter chemistry at small spatial scales. In line with our hypothesis, interspecific differences in litter decomposition were driven by litter C compounds such as easily accessible non-structural carbohydrates, phenolics or dissolved organic carbon (DOC), but also by inhibiting condensed tannins. N- or P-related litter traits explained no variation in decomposition.

Biomass and the structure and stoichiometry of microbial communities differed substantially among different leaf litter species. This variation was related to litter C quality, but not to litter nutrients or litter C:N:P stoichiometry. Litter identity also influenced microbial activity in the underlying soil, mainly as a result of litter species-specific DOC concentrations.

Fertilization with C (cellulose), N (urea), and P (phosphate) alone had no or little effect on litter decomposition. However, the combined fertilization with N and P increasingly stimulated decomposition with increasing litter N:P ratios. This effect of combined N and P fertilization was enhanced in presence of soil macrofauna that was more abundant in plots fertilized simultaneously with N and P.

Our data suggest that without external resource inputs, decomposer communities and litter decomposition in the studied tropical rainforest is under control of easily accessible C compounds of comparatively low concentrations in leaf litter. Fertilization with C, N, and P did not confirm this conclusion possibly because cellulose may contribute to an increased C availability and priming of litter decay only in the longer term, or because of the difference between proximate (energy) and fundamental (nutrients) limitations.