NUTRIENT CYCLING THROUGH MICROBIAL BIOMASS UNDER RICE-PASTURE ROTATIONS REPLACING NATIVE SAVANNA

An area of native savanna on an acid, strongly P-sorbing Oxisol in the Eastern Plains of Colombia was opened and sown to Various rotations o f grass pasture with rice, grass-legume pasture with rice or rice mono crop. After 4.5 y, the soil under each cropping system was sampled and analysed for total organic matter, microbial C, N and P content, and mineralization rate of C and N. Microbial biomass C did not vary much among treatments, whereas the N and P contents of the biomass were con siderably lower in rice monocrop than in any crop-pasture treatment. B iomass P was also low under native savanna. The contribution of microb ial N or P to soil-organic-matter N or P was lowest under the rice mon ocrop and highest-under the grass-legume pasture: Microbial C-to-N rat ios fell in ranges commonly reported, but C-to-P ratios were rather wi de (34-50), indicating that the microbes may have adapted themselves t o the low-P conditions of these soils. The contribution of microbial P to soil-organic-matter P, however, was about the same as usually foun d in soils of much higher P fertility. As microbial biomass nutrients cycle relatively rapidly and P availability in these strongly P-sorbin g soils is low, the microbial biomass may play an important role in su pplying P to plants growing in these soils. A similar conclusion was r eached for N. The microbial respiration rate per unit of microbial bio mass C (qCO(2)), as determined during an Ii-day incubation, was higher under rice-pasture rotations-particularly in the presence of a legume -than under rice monocropping. This suggests that a greater fraction o f microbes was active under rice-pasture rotations, probably because o f a more continuous and higher input of fresh organic matter. The frac tion of organic matter mineralized during the incubation was also high est for the rice-pasture treatments. (C) 1997 Published by Elsevier Sc ience Ltd.