Oil palm plantations growing on peat soil are associated with high soil CO2 emissions. Oil palm plantations are set up with regular spatial patterns consisting of different surface management microforms: bare soil harvest paths, frond piles, cover plants and drainage ditches. Currently, there is limited understanding about the extent that this spatial variation impacts soil carbon losses, in part due to the challenges of partitioning peat oxidation from total soil respiration. We explored this spatial variation by measuring total soil respiration (Rtot), root density and environmental variables at 210 locations. Measurements were taken along transects going from the base of oil palms into the different microforms. Rtot was partitioned into root respiration (Ra) and heterotrophic respiration (Rh) using two different methods: (i) a "distance from palm" method (which utilizes the fluxes taken from soil with minimal root density) and (ii) a "linear regression" method (which models root density and Rtot, using the regression intercept for Rh). Here, the distance from palm partitioning method gave higher Rh estimates than the linear regression method. Rh varied significantly between the different palms used in the assessment but did not show significant spatial variation aside from this. Rtot and Ra were highest next to the palm and decreased with increasing distance from the palm. Rtot and Ra also showed significant spatial variation between the different surface management microforms, with each giving significantly higher fluxes below the frond piles near the drainage ditches than from below the frond piles near the cover plants. Area-weighted upscaling gave plantation best estimates of Rtot, Rh, Ra of 0.158 +/- 0.016, and 0.130 +/- 0.036 and 0.029 +/- 0.030 g CO2-C m-2 h-1, respectively. We conclude that spatial patterns impact root density, Ra and Rtot fluxes but not Rh fluxes.
View source
