Controls on soil carbon storage and turnover in German landscapes

Abstract

Soil organic carbon (OC) storage across regions is influenced by climate and parent materials, which determine soil properties like clay content and mineralogy. Within homogeneous soil regions, land use and management practices are further important controls for soil OC contents and turnover. Here, we studied the impact of study region, land use (forest, grassland), forest management (spruce and beech forest under age-class management, unmanaged beech forest), and grassland management (meadow, mown pasture, unmown pasture) on stocks and turnover (based on Δ14C values) of soil OC in density frations of topsoil horizons. Samples were taken from 36 plots in the regions Hainich–Dun (HAI) and the Schwabische Alb (ALB) in Germany. They were separated into two light fractions (free light fraction (LF1), occluded light fraction (LF2)) and the mineral-associated organic matter (MOM) fraction using sodium polytungstate with a density of 1.6 g cm−3. Overall most soil OC was stored in the MOM fraction (73%). Soil OC concentrations and stocks in the MOM fraction differed between study regions, probably due to larger amounts of pedogenic Al- and Fe-oxides in the ALB than in the HAI region. Within each region, forest soils stored significantly higher proportions of total OC in the two LF (33±1.9 %) than grassland soils (20±2.3 %). Different management of forests and grasslands affected the C:N ratio of density fractions, but not OC storage. While modelled soil OC turnover in the MOM was longest of all fractions, all fractions had average Δ14C values above atmospheric values, suggesting a significant fast-cycling component in all of them. Different from stocks, turnover of OC in the MOM fraction were not affected by study region or contents of pedogenic oxides. Radiocarbon contents in the LF were higher for forest than for grassland sites, indicating faster turnover of OC at grassland sites. However, some of the observed difference could originate from different average lifetimes of roots in forests and grasslands. Applying different lag-times for OC input for forests and grasslands significantly reduced the differences in modelled turnover times. Lower Δ14C values of mown pastures than pasture soils in both regions suggest a management effect on soil C turnover in grasslands.We conclude that OC storage in the MOM of topsoil layers is more affected by regional differences in soil texture and mineralogy than by land use and management, while its turnover could not be explained with the studied soil properties. Soil OC storage and turnover in the two LFs is influenced by land use (forest or grassland) and management, but ecosystem specific lag-times have to be considered for modelling OC turnover in these fractions.