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 off set 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 off set or magnified by carbon change at depth.