Agricultural Management and Soil Carbon Storage in Surface vs. Deep Layers

Soil C sequestration research has historically focused on the top 0 to 30 cm of the soil profile, ignoring deeper portions that might also respond to management. In this study we sampled soils along a 10‐treatment management intensity gradient to a 1‐m depth to test the hypothesis that C gains in surface soils are offset by losses lower in the profile. Treatments included four annual cropping systems in a corn ( Zea mays )‐soybean ( Glycine max )‐ wheat ( Triticum aestivum ) rotation, perennial alfalfa ( Medicago sativa ) and poplar ( Populus x euramericana ), and four unmanaged successional systems. The annual grain systems included conventionally tilled, no‐tillage, reduced‐input, and organic systems. Unmanaged treatments included a 12‐yr‐old early successional community, two 50‐yr‐old mid‐successional communities, and a mature forest never cleared for agriculture. All treatments were replicated three to six times and all cropping systems were 12 yr post‐establishment when sampled. Surface soil C concentrations and total C pools were significantly greater under no‐till, organic, early successional, never‐tilled mid‐successional, and deciduous forest systems than in the conventionally managed cropping system ( p ≤ 0.05, n = 3–6 replicate sites). We found no consistent differences in soil C at depth, despite intensive sampling (30–60 deep soil cores per treatment). Carbon concentrations in the B/Bt and Bt2/C horizons were lower and two and three times more variable, respectively, than in surface soils. We found no evidence for C gains in the surface soils of no‐till and other treatments to be either offset or magnified by carbon change at depth.