A modelling method to quantify in situ the input of carbon from roots and the resulting C turnover in soil

Decomposition experiments, using C-14 and N-15 labelled plant materials in porous soilbags buried in contact with roots, were laid out in the Venezuelan paramo and the Bolivian puna. The tracers were previously used to calibrate a new version of Momos, a microbial biomass (MB) and decomposition model. MB is growing by ingestion of labile (VL) and stable (VS) plant materials as well as labile (HL) and stable (HS) humus, and is decreasing by respiration (CO2 output) and mortality (input into HL). The fallow production model Faprom estimates biomass production, standing dead mass, and soil C input from plant mortality and potential root exudation. This work aimed to address three objectives: (1) use parameter values calibrated on C-14 transformations to predict total C changes in soils, (2) quantify by simulations the C storage and turnover from new C-12 inputs from roots, (3) split inputs from root death into VL and VS, direct root exudation into MB. These objectives were attained except (2) in the humic soil of the paramo, for which total C or MB-C predictions were not significantly improved by root-C input. The method described here is a new tool to predict in situ C input from roots (even when a production model is lacking) and the resulting C turnover for soils with moderate C stock. Experimental data could be improved by optimizing contact between roots and soilbags and enhancing the repeatability of MB-C measurements. Momos predicts half-lives of MB-C and HL-C labile compartments to be about 2 years in the puna and 35-80 y in the paramo. The corresponding half-lives of the stable HS-C compartment were 300-500 y in the puna and 800-900 y in the paramo