Carbon turnover (delta C-13) and nitrogen mineralization potential of particulate light soil organic matter after rainforest clearing

Soil samples from under a rainforest, a papaya plantation, a pure Brachiaria humidicola pasture and a B. humidicola/Desmodium ovalifolium sward from the humid tropical Bahia, Brazil, were analyzed for fertility parameters, litter quality, particulate light fraction [PLF>100/53-100 mu m, < 1.8 g cm(-3) density (sodium polytungstate)] and delta(13)C signatures. Under the grass pasture total soil C was reduced to 80 and 95% of the initial forest value after 18 years in the surface 2 cm and the 5-15 cm horizons, respectively, largely due to the sharp decline in rainforest-derived C after deforestation. Losses of rainforest-derived C amounted to 73% (0-2 cm) and 40% (5-15 Cm) of the initial soil C content. The contribution of newly added B. humidicola-derived soil C was estimated to be 24 and 5 mg C g(-1) soil and accounted for 68 and 37% of total C in the surface and lower soil layer, respectively, after 18 years. The C content of the particulate light fraction (PLF-C) was largest in the rainforest and smallest in the papaya plantation. The PLF-C accounted for 12-36% in the surface soil and 10-16% in the 5-15 cm layer. The size of the PLF was not only influenced by the quantity of C inputs, but also strongly reflected the quality of incoming litter. According to its delta(13)C signature the large (> 100 mu m) PLF was of younger age than the finer PLF or whole soil samples. The age of the 53-100 mu m PLF was very close to that of stabilized organic matter as indicated by C-13 data, its low C-to-N ratio (15-17:1) and its turnover time was about 30 y. The PLF appeared not to consist of one uniform pool and thus may have to be subdivided to be useful for modelling purposes (e.g. active, passive pools and charcoal). The contribution of PLF and above ground litter to total soil N mineralization (anaerobic incubation) was highest in the undisturbed rainforest (12/16% for the PLF/litter, respectively, in the 0-2 cm layer), but much less in the other systems. However, mineralization of PLF from different sources was unpredictably altered by the density agent. Introducing D. ovalifolium into these pastures increased, significantly, the amounts of N in PLF, decreased the C-to-N ratio of PLF and tended to increase the mineralization potential despite its high lignin and polyphenol content. The combination of delta(13)C analysis and size/density separations, of surface samples in particular, allowed sensitive detection of changes in soil organic matter dynamics and soil fertility. (C) 1997 Elsevier Science Ltd.