Meta-analysis on effects of residue retention on soil N2O emissions and influence factors in China

Abstract: Scientists have paid more and more attention to the nitrous oxide (N2O) emissions due to the great contribution of N2O to global warming and stratospheric ozone destruction. Agricultural N2O emissions have been estimated to more than 60%-75% of calculated annual atmospheric N2O. Thus, it is critical to identify the effects on N2O emission responding to residue retention (RR) in understanding how RR contributes to mitigation of climate change. In this study, a meta-analysis based on 31 published peer-reviewed papers published before 2015 was conducted to investigate the effects of RR on soil N2O emission and 79 available comparisons were compiled into the dataset. Natural log of response ratio was used as the effect size in the random effect meta-analysis. Only field scale studies were included in this study. Original documented information, including seasonal accumulative N2O emission flux, standard deviation, replicates, and types of land-use were obtained from each study. If seasonal accumulative N2O emission fluxes were not directly provided, these values were calculated by multiplying time and mean N2O emissions by the measurement period. To identify the difference from site-specific conditions, a categorical meta-analysis were adopted by dividing the comparisons into groups to assess the effects of N2O emission responding to RR under different cropping systems, soil properties, land use types, and regions. The results presented herein showed that differences in N2O emission under RR among regions were obvious in the categorical meta-analysis. For example, RR significantly reduced N2O emission by 17% in the Eastern China (P<0.05), while significantly increased the emission by 95% and 34% in the Central and North China, respectively (P<0.05). Furthermore, the natural log of response ratio of N2O emission increased with increasing N fertilization input rate but was limited, and the effect of RR on N2O emission gradually changed from suppression to promotion. In addition, when N fertilization input rate was more than 240 kg/hm2, the effect size about N2O emission was negative under RR. The N2O emission were increased significantly by 21% when N fertilization input was between 180-240 kg/hm2 and decreased significantly by 16% when N fertilization was between 240-300 kg/hm2. In addition, the results showed soil pH had no significant influence on the N2O emission. Results indicated that N2O was significantly reduced by 29% (P<0.05) when the mass fraction of clay was less than 15%, but without significant influence when more than 15%. However, the weighted mean effect size showed a decreasing trend after increasing with pH value and mass fraction of clay. Thus, crop residue retention may increase the emission of N2O in neutral soil, or decrease in acidity and alkalinity soil. Results also showed that the N2O emission could increase with the increase of the amount of crop residue retention. This may due to the characteristic of residue as crop fertilizer. The C:N ratio of crop residues also affected the emission of N2O in despite of no significant influence in this study. The retained residue with high C:N ratio could decrease N2O emission. Significant differences in N2O emissions were observed within the same land use type along with different cropping systems. For example, a significant decrease by 12% was observed in rice season for a rice-wheat system but a significant increase by 10% was found in rice season of rice-rape system, respectively (P<0.05). The effects of RR on N2O emission were affected by many site-specific conditions and managements. Therefore, further research on the reduction of soil N2O emissions with residue retention should be targeted and to specific biophysical conditions, e.g., soil type, cropping system and farming management.