Forest succession (FS) and no-till (NT) agriculture are generally assumed to have a beneficial effect on surficial soil organic C (SOC) stocks compared with conventional tillage (CT) management; however, land use effects to depths >30 cm remain uncertain. In this research we compared SOC contents and composition to 2 m under CT, NT, and FS at the 30-year Horseshoe Bend agroecosystem experiment in Athens, Georgia, USA. Soils from 0 to 2 m were fractionated into particulate organic C (POC) (53-2000 m) and fine C (<53 m) fractions, and bulk soil delta 13C signatures were determined. Soils from 0 to 28 cm were dry- and wet-sieved to estimate aggregate stability. Soil solutions were also collected at 0, 15, and 100 cm for dissolved organic C (DOC) analysis. Full-profile (0-2 m) SOC storage is 52 Mg ha -1 in CT, 60 Mg ha -1 in NT, and 62 Mg ha -1 in FS. Significant differences are limited to 0-5 cm and are linked to enhanced aggregate stability under NT and FS. Increases in subsoil POC under FS and changes in soil delta 13C and C/N ratio indicate that substantial subsoil C cycling has occurred. DOC fluxes at 0 cm were significantly greater under NT (200 kg ha -1 year -1) and FS (210 kg ha -1 year -1) than under CT (80 kg ha -1 year -1). DOC fluxes at 15 cm are estimated to be 20 kg ha -1 year -1 under CT and NT and 40 kg ha -1 year -1 under FS. At 100 cm, DOC fluxes are 2 kg ha -1 year -1, regardless of land use. An increase in FS POC of 2 Mg ha -1 from 15 to 100 cm outweighs cumulative differences in DOC input to this layer, implicating deep forest rooting and bioturbation as active mechanisms in subsoil C change. Whereas differences in SOC content were concentrated near the surface, dynamic changes in C cycling extend well below the plow layer.
