Spatial controls of topsoil and subsoil organic carbon turnover under C3-C4 vegetation change

Soil organic carbon (SOC) is often heterogeneously distributed in arable fields and so is probably its turnover. We hypothesized that the spatial patterns of SOC turnover are controlled by basic soil properties like soil texture and the amount of rock fragments. To test this hypothesis, we cultivated maize as a C-4 plant on a heterogeneous arable field (155 x 60 m) that had formerly been solely cultivated with C-3 crops, and monitored the incorporation of isotopically heavier maize-derived C into SOC by stable C-13 isotope analyses. To obtain a homogeneous input of C-4 biomass into the C-3 soil across the field, we chopped the aboveground maize biomass after harvest in autumn and re-spread it uniformly over the field. Subsequently, the soil was grubbed and then ploughed in the next spring. In addition, we assessed the spatial patterns of SOC stocks, amount of rock fragments and texture, as well as potential soil organic matter (SOM) degradability by ex-situ soil respiration measurements. Heterogeneity of maize growth was monitored as a covariate using laser scanning and satellite images. After two years, maize C had substituted 7.4 +/- 3.2% of SOC in the topsoil (0-30 cm) and 2.9 +/- 1.7% of SOC in the subsoil (30-50 cm). Assuming that monoexponential decay mainly drove this SOC substitution, this resulted in mean residence times (MRT) of SOC in the range of 30 12 years for the topsoil and of 87 45 years for the subsoil, respectively. Variation in topsoil MRT was related to potential CO2 release during soil incubation (R-2 = 0.51), but not to basic soil properties. In the subsoil, in contrast, the variation of maize C incorporation into the SOC pool was controlled by variations in maize yield (R-2 = 0.44), which also exhibited a pronounced spatial variability (0.84 to 1.94 kg dry biomass m(-2)), and which was negatively correlated with the amount of rock fragments (R-2 = 0.48, p < 0.001). We assume that heterogeneous input of belowground root biomass blurs the causal interactions between the spatial heterogeneity of soil properties and the related patterns of SOC turnover, and conclude that spatial patterns of SOC turnover are not easily predictable by standard soil analyses.