Long-term effects of crop rotations and fertilization on soil C and N in a thin Black Chernozem in southeastern Saskatchewan

Carbon sequestration in soil is important due to its influence on soil fertility and its impact on the greenhouse gas (GHG) phenomenon. Carbon sequestration is influenced by agronomic factors, but to what extent is still being studied. Long-term agronomic studies provide one of the best means of making such assessments. In this paper we discuss and quantify the effect of cropping frequency, fertilization, legume green manure (LGM) and hay crops in rotations, and tillage on soil organic carbon (SOC) changes in a thin Black Chernozemic fine-textured soil in southeastern Saskatchewan. This was based on a 50-yr (1958-2007) crop rotation experiment which was initiated on land that had previously been in fallow-wheat (Triticum aestivum L.) (F-W), or F-W-W receiving minimum fertilizer for the previous 50 yr. We sampled soil in 1987, 1996 (6 yr after changing from conventional tillage to no-tillage management and increasing N rates markedly) and again in 2007. The SOC (0-15 cm depth) in unfertilized F-W and F-W-W appears not to have changed from the assumed starting level, even after 20 yr of no-till, but SOC in unfertilized continuous wheat (Cont W) increased slightly [not significant (P>0.05)] in 30 yr, but increased more after 20 yr of no-till (but still not significant). No-till plus proper fertilization for 20 yr increased the SOC of F-W, F-W-W and Cont W in direct proportion to cropping frequency. The SOC in the LGM-W-W (unfertilized) system was higher than unfertilized F-W-W in 1987, but 20 yr of no-tillage had no effect, likely because grain yields and C inputs were depressed by inadequate available P. Soil organic carbon in the two aggrading systems [Cont W (N+P) and F-W-W-hay(H)-H-H (unfertilized)] increased significantly (P<0.05) in the first 30 yr; however, a further 20 yr of no-tillage (and increased N in the case of the Cont W) did not increase SOC suggesting that the SOC had reached a steady-state for this soil and management system. The Campbell model effectively simulated SOC changes except for Cont W(N+P), which it overestimated because the model is ineffective in simulating SOC in very fertile systems. After 50 yr, efficiency of conversion of residue C inputs to SOC was negligible for unfertilized F-W and F-W-W, was 3 to 4% for fertilized fallow-containing systems, was about 6 or 7% for Cont W, and about 11% for the unfertilized F-W-W-H-H-H systems.