LAUSR

Abstract Process-based modelling studies can help to interpret experimental results and gain more insights into CO 2 fluxes in agro-ecosystems. The objective of this study was to evaluate the ability of the DeNitrification-DeComposition (DNDC) model to predict field-measured soil temperature, moisture, and respiration (daily and seasonal) in monoculture versus rotational cropping systems. Field measurements (2003–2005) from monoculture cropping (corn, Zea mays L.; soybean, Glycine max L.; and winter wheat, Triticum aestivum L.), as well as from 2-yr (corn-soybean) and 3-yr (corn-soybean-winter wheat) crop rotations in southwestern Ontario, Canada were used for model calibration and evaluation. Model simulated soil (0–10 cm) temperature ( nRMSE  = 8–9%, d  = 0.97 and EF  = 0.87–0.90) and moisture ( nRMSE  = 15–22%, d  ≥ 0.7 and EF  ≥ 0) agreed well with the growing season field measurements. Predicted daily soil CO 2 fluxes were accurate for corn (under both monoculture and rotation) and winter wheat (under monoculture) ( nRMSE  = 56–70%, d >  0.7, slope = 0.60–0.88, R 2  = 0.36–0.48). The overestimation in daily soil CO 2 fluxes for winter wheat phase under the 3-yr rotation could be addressed by decreasing the root: shoot ratio. The simulated temporal offset in the soil CO 2 fluxes for soybean could be improved by adjusting crop parameters (thermal degree days for crop maturity and root: shoot ratio). The simulated cumulative seasonal soil CO 2 fluxes were not statistically different from the measured CO 2 fluxes for corn, soybean and winter wheat as indicated by the paired- t test ( p  = 0.17–1.00). The impacts of monoculture and rotation cropping on seasonal root autotrophic respiration (R A ), soil heterotrophic respiration (R H ) and R H /R S ratios for corn, soybean and winter wheat were attributed to the different biomass production and residue C input management systems. The R H contributed more to total soil respiration for corn (during the growing season) under monoculture (R H /R S  = 42%) than when it was grown in rotation (26–29%), while ratios of R H /R S were similar among monocultural and rotational soybean (41–49%) and winter wheat (31–40%). The current litter C level and consequently annual R H were driven by the litter C input carried over from previous crop. When carbon in the harvested grain was taken into account in the carbon budget, corn would act as a strong net carbon sink (3 year average of −975, −2675 and −3151 kg C ha −1  yr −1 under monoculture, 2-yr and 3-yr rotations, respectively), while soybean and winter wheat would act as weak net carbon source under both monoculture and rotation. The DNDC model was found to be able to simulate soil temperature, moisture, and soil respiration for corn and winter wheat under both monoculture and rotation in humid southwestern Ontario, Canada but crop parameter adjustments are required to simulate soybean.