Isotopic ((13)C) fractionation during plant residue decomposition and its implications for soil organic matter studies.

Carbon isotopic fractionations in plant materials and those occurring during decomposition have direct implications in studies of short-and longer-term soil organic matter dynamics. Thus the products of decomposition, the evolved CO(2) and the newly formed soil organic matter, may vary in their (13)C signature from that of the original plant material. To evaluate the importance of such fractionation processes, the variations in (13)C signatures between and within plant parts of a tropical grass (Brachiaria humidicola) and tropical legume (Desmodium ovalifolium) were measured and the changes in (13)C content (signatures) during decomposition were monitored over a period of four months. As expected the grass materials were less depleted in (13)C (-11.4 to -11.9 per thousand) than those of the legume (-27.3 to -25.8 per thousand). Root materials of the legume were less (1.5 per thousand) depleted in (13)C compared with the leaves. Plant lignin-C was strongly depleted in (13)C compared with the bulk material by up to 2.5 per thousand in the legume and up to 4.7 per thousand in the grass. Plant materials were subsequently incubated in a sand/nutrient-solution/microbial inoculum mixture. The respiration product CO(2) was trapped in NaOH and precipitated as CaCO(3), suitable for analysis using an automated C/N analyser coupled to an isotope ratio mass spectrometer. Significant depletion in (13)C of the evolved CO(2) was observed during the initial stages of decomposition probably as a result of microbial fractionation as it was not associated with the (13)C signatures of the measured more decomposable fractions (non-acid detergent fibre and cellulose). While the cumulative CO(2)-(13)C signatures of legume materials became slightly enriched with ongoing decomposition, the CO(2)-C of the grass materials remained depleted in (13)C. Associated isotopic fractionation correction factors for source identification of CO(2-)C varied with time and suggested errors of 2-19% in the estimation of the plant-derived C at 119 days of incubation in a soil of an intermediate (-20.0 per thousand) (13)C signature. Analysis of the residual material after 119 days of incubation showed little or no change in the (13)C signature partly due to the incomplete decomposition at the time of harvesting. Copyright 1999 John Wiley & Sons, Ltd.