Abstract
Uncertainty in soil carbon (C) fluxes across different land-use transitions is an issue that needs to be addressed for the further deployment of perennial bioenergy crops. A large-scale short-rotation coppice (SRC) site with poplar (Populus) and willow (Salix) was established to examine the land-use transitions of arable and pasture to bioenergy. Soil C pools, output fluxes of soil CO (2), CH (4), dissolved organic carbon (DOC) and volatile organic compounds, as well as input fluxes from litter fall and from roots, were measured over a 4-year period, along with environmental parameters. Three approaches were used to estimate changes in the soil C. The largest C pool in the soil was the soil organic carbon (SOC) pool and increased after four years of SRC from 10.9 to 13.9 kg C m(-2). The belowground woody biomass (coarse roots) represented the second largest C pool, followed by the fine roots (Fr). The annual leaf fall represented the largest C input to the soil, followed by weeds and Fr. After the first harvest, we observed a very large C input into the soil from high Fr mortality. The weed inputs decreased as trees grew older and bigger. Soil respiration averaged 568.9 g C m(-2) yr(-1). Leaching of DOC increased over the three years from 7.9 to 14.5 g C m(-2). The pool-based approach indicated an increase of 3360 g C m(-2) in the SOC pool over the 4-year period, which was high when compared with the -27 g C m(-2) estimated by the flux-based approach and the -956 g C m(-2) of the combined eddy-covariance + biometric approach. High uncertainties were associated to the pool-based approach. Our results suggest using the C flux approach for the assessment of the short-/medium-term SOC balance at our site, while SOC pool changes can only be used for long-term C balance assessments.