Changes in soil carbon under long-term maize in monoculture and legume-based rotation

Legume-based cropping systems could help to increase crop productivity and soil organic matter levels, thereby enhancing soil quality, as well as having the additional benefit of sequestering atmospheric C. To evaluate the effects of 35 yr of maize monoculture and legume-based cropping on soil C levels and residue retention, we measured organic C and 13C natural abundance in soils under: fertilized and unfertilized maize (Zea mays L.), both in monoculture and legume-based [maize-oat (Avena sativa L.)-alfalfa (Medicago sativa L.)-alfalfa] rotations; fertilized and unfertilized systems of continuous grass (Poa pratensis L.); and under forest. Solid state 13C nuclear magnetic resonance (NMR) was used to chemically characterize the organic matter in plant residues and soils. Soils (70-cm depth) under maize cropping had about 30-40% less C, and those under continuous grass had about 16% less C, than those under adjacent forest. Qualitative differences in crop residues were important in these systems, because quantitative differences in net primary productivity and C inputs in the different agroecosystems did not account for observed differences in total soil C. Cropping sequence (i.e., rotation or monoculture) had a greater effect on soil C levels than application of fertilizer. The difference in soil C levels between rotation and monoculture maize systems was about 20 Mg C ha-1. The effects of fertilization on soil C were small (~6 Mg C ha-1), and differences were observed only in the monoculture system. The NMR results suggest that the chemical composition of organic matter was little affected by the nature of crop residues returned to the soil. The total quantity of maize-derived soil C was different in each system, because the quantity of maize residue returned to the soil was different; hence the maize-derived soil C ranged from 23 Mg ha-1 in the fertilized and 14 Mg ha-1 in the unfertilized monoculture soils (i.e., after 35 maize crops) to 6-7 Mg ha-1 in both the fertilized and unfertilized legume-based rotation soils (i.e., after eight maize crops). The proportion of maize residue C returned to the soil and retained as soil organic C (i.e., Mg maize-derived soil C/Mg maize residue) was about 14% for all maize cropping systems. The quantity of C3-C below the plow layer in legume-based rotation was 40% greater than that in monoculture and about the same as that under either continuous grass or forest. The soil organic matter below the plow layer in soil under the legume-based rotation appeared to be in a more biologically resistant form (i.e., higher aromatic C content) compared with that under monoculture. The retention of maize residue C as soil organic matter was four to five times greater below the plow layer than that within the plow layer. We conclude that residue quality plays a key role in increasing the retention of soil C in agroecosystems and that soils under legume-based rotation tend to be more “preservative” of residue C inputs, particularly from root inputs, than soils under monoculture. Key words: Soil carbon, 13C natural abundance, 13C nuclear magnetic resonance, maize cropping, legumes, root carbon