Quantifying carbon sequestration in a minimum tillage crop rotation study in semiarid southwestern Saskatchewan

Scientists and the agricultural community require methods of quantifying C sequestration in soils. This is important in assessing the impact of crop management practices on emission of greenhouse gases and for "C trading". Using simulation models may be a more effective method of quantification as compared with in situ measurements. A 17-yr crop rotation experiment being conducted on a medium-textured Orthic Brown Chemozem at Swift Current, Saskatchewan, in which soil organic C (SOC) was being monitored periodically, was used to assess the effect on C sequestration of cropping frequency, wheat class, legume green manure (LGM), flexible cropping based on available water, and regrassing of crop land. Prior to the study, the experimental site had been cropped to fallow-wheat (F-W) for the previous 60 yr. Crop management in this experiment involved minimum tillage, snow trapping, and N + P fertilization based on soil tests. Three models [Century, the Introductory C Balance model (ICBM), and the Campbell model] were tested for their effectiveness in simulating SOC trends. Because growing season precipitation was average to above average, yields, and thus C inputs from residue, were also above average, and consequently SOC increased in most systems for the first 10 yr before reaching a new steady state. SOC gains (kg ha(-1) yr(-1)) in the 0- to 15-cm depth in 17 yr were directly proportional to cropping frequency (F-W-W = 135, F-W-W-W = 332, and Cont W = 441); LGM-W-W gained SOC at a much higher rate than F-W-W (329 vs. 135 kg ha(-1) yr(-1)); Canada Western Red Spring (CWRS) wheat (Triticum aestivum L.) although it yielded 26% less than Canada Prairie Spring (CPS) wheat, gained SOC at a higher rate than CPS wheat (135 vs. 0 kg ha(-1) yr(-1)). Further, 2 yr of conventionally-tilled fallow in 17 yr (flexible system) markedly suppressed SOC gain by 46% compared with Cont W (441 vs. 236 kg ha(-1) yr(-1)). There was a 282 kg ha(-1) yr(-1) gain in SOC under crested wheatgrass (Agropyron cristatum L.) (CWG) but most of this gain occurred in the last 7 yr. Though having their inherent weaknesses, the ICBM and Campbell models performed equally well in simulating SOC trends (r(2) = 0.55**), but Century was less effective (r(2) = 0.21*), in part because of its limited ability to simulate yields. Because C input, and thus yield, is one of the main factors influencing SOC gains, and since measured yields are used in the ICBM and Campbell models, while simulated yields are used by Century, the ICBM and Campbell models have an advantage over the Century model in this comparison. Efficiencies of conversion of input C to SOC increased with cropping frequency, and were higher for LGM-W-W than for F-W-W, and for systems with CWRS wheat rather than CPS wheat. Efficiency of conversion was 8% for F-W-W, 15% for LGM-W-W and 21% for Cont W.