Evaluation of Long-term Agroecosystem Management on Changes in Subsurface vs. Surface Soil Carbon Fractions and Dynamics

Most studies of terrestrial soil organic carbon (SOC) have focused on the upper soil profile (e.g., 0-30 cm), so our knowledge of C dynamics in deeper layers is incomplete. Here, we examine the depth-dependent mechanisms and constraints by which management of the upper soil profile for optimum crop yield in agroecosystems can influence SOC fractions and change in both the surface and subsurface. Our study includes continuous corn systems under long-term conventional tillage (CT) vs no-tillage (NT) at Willsboro, New York (NY) (Kingsbury silty loam soil; 19 y) and Chazy, NY (Raynham silt loam; 38 y), and long-term crop rotation experiments under CT at Algona, Iowa (IA) (Clarion loam; 11 y) and Kanawha, IA (Canisteo clay loam; 57 y). Rotations in IA compared continuous corn to corn rotations with soybean, alfalfa, and/or oats. Cores were collected in 2011 and 2012 at 0-10, 10-20, 20-30, 30-50 and 50-75 cm, and analyzed for bulk density, soil texture, percent organic matter, total C and nitrogen (N), soil inorganic C, and active C (permanganate oxidizable C, POXC). Recent studies have documented that POXC is closely correlated with heavy, small-sized particulate organic C, reflecting a relatively processed and stable pool of labile C that is well-suited to assess land management effects on C dynamics. Overall, cumulative SOC stocks (0-75 cm) in the IA and NY soils ranged from 109.9-168.8 MgC ha-1, and 37.8-104.1 MgC ha-1, respectively. The proportion of total SOC stocks that occurred in the subsurface (30-75 cm) ranged from 39-44% in the IA soils, compared to 16-26% in NY. Across all sites and management we found no examples of statistically significant SOC change below 30 cm, although this may be in part an artifact of greater variability and smaller absolute values of C concentration at depth. SOC data were correlated with POXC measurements, although depth- and site-specific discrepancies in these two measures were observed. For example, POXC was relatively unaffected by tillage treatment in the upper profile while SOC declined with CT compared to NT at the NY sites. Also, in some cases POXC was less variable and more sensitive to management than SOC. The rate of increase in SOC with NT compared to CT over the course of the long-term studies in NY was 0.30 and 1.2 MgC ha-1 yr-1 at the Chazy and Willsboro sites, respectively. In contrast, crop rotation treatments had no significant effect on SOC or POXC at Kanawha, IA, after 57-years of continuous management. At Algona, a small sample size (22 cores, 0.4 ha) collected in 2011 found significantly higher SOC in the conventional corn-soybean rotation compared to an organic system using manure with oats and alfalfa in rotation, but an expanded sampling (200 cores, 24 ha) in 2012 found no significant differences. Our results will be discussed in relation to tillage vs rotation impacts on surface and subsurface dynamics of labile and stable C pools, and POXC:N ratio effects on organic matter decomposition.