Effect of grazing exclusion on nitrous oxide emissions during freeze-thaw cycles in a typical steppe of Inner Mongolia

•High pulse N2O emissions were found during spring freeze-thaw cycles (S-FTCs) in a typical grassland of Inner Mongolia.•Short-term grazing exclusion (GE) could increase N2O emissions during FTCs, while long-term GE had no significant effects.•Soil water content was a predominate factor that caused increased N2O emissions.•Differences in soil water content during FTCs were associated with the snow cover depending on vegetation biomass. Grazing exclusion of grasslands has traditionally been considered to reduce nitrogen inputs and decrease nitrous oxide (N2O) emissions. However, recent studies have shown that grazing exclusion may increase N2O emissions because of freeze-thaw cycles. In this study, year-round N2O fluxes (including high-frequency measurements during the freeze-thaw cycles) were measured in three sites that included long-term grazing exclusion (ungrazed since 1979, UG79), short-term grazing exclusion (ungrazed since 1999, UG99), and continuously grazed (CG), in Inner Mongolia grassland. The results showed that: 1) across the entire period of observations, mean annual N2O fluxes were the highest at UG99 (12.2 μg m−2 h-1), followed by CG (8.6 μg m−2 h-1) and UG79 (7.5 μg m−2 h-1); 2) during spring freeze-thaw cycles, cumulative N2O emissions were significantly higher at UG99 than UG79 and CG (P < 0.05); 3) freeze-thaw cycles had significant effects on N2O emissions with strong pulse emissions exhibited at the beginning of the spring thaw period; and 4) structural equation model analysis indicated that soil water content was a predominant environmental factor that caused increased N2O emissions, while other factors tended to affect N2O emissions indirectly. The results confirmed that short-term grazing exclusion could increase N2O emissions owing to the high N2O concentration during freeze-thaw cycles, associated with high quantities of snowmelt water. However, the effects of long-term grazing exclusion on N2O emissions showed no differences compared with grazing management. This suggests that grazing exclusion does not always result in an increase in N2O emissions during freeze-thaw cycles as it is more controlled by soil liquid water content. Thus the new finding of the event emissions of N2O should come up with more reliable estimates of soil‐atmosphere trace fluxes.