CH4, CO2, and N2O flux during oilseed rape and rice seasons
Figure 2 summarizes the seasonal CH4 (Fig. 2a), CO2 (Fig. 2b), and N2O (Fig. 2c) emissions from oilseed rape and rice crops subjected to NT and CT from 2022 to 2023. We observed a clear effect of tillage practices on CH4, CO2, and N2O emissions in oilseed rape-rice rotation crop fields. During the 2022 oilseed rape season (Fig. 2a), CH4 emission flux peaked between 2–3 wk in both treatments and stabilized thereafter. The peak CH4 emission flux for CT treatment (60.2 mg m− 2 h− 1) was significantly higher than that for NT treatment (49.7 mg m− 2 h− 1). Notably, CT registered negative flux values at 8–9 wk, whereas NT registered negative values at 17–18 wk. Subsequently, during the 2022 rice cropping season, CH4 emissions from both NT and CT treatments peaked in the first week, at 242.0 and 286.8 mg m− 2 h− 1, respectively, and then gradually decreased and stabilized without reaching negative values. In contrast, unlike those observed in 2022, CH4 emission fluxes of the NT and CT treatments during the oilseed rape cropping season of 2023 fluctuated throughout the reproductive period. The daily CH4 emission flux during NT treatment gradually stabilized after reaching a peak during the first week of the rice season, whereas the daily CH4 emission flux during CT treatment fluctuated during the fourth week after reaching a peak in the first week. Furthermore, no significant differences in the peak emissions were observed between the two treatments in 2023.
During the oilseed rape and rice cropping seasons of 2022 and 2023, CO2 fluxes fluctuated for both NT and CT. The CT treatment peaked during the fourth week of the rice-growing season in both years, whereas the NT treatment peaked during the eighth and fourth weeks of the respective seasons. Furthermore, the emission peaks for the CT treatments were significantly higher than those for the NT treatments (Fig. 2b). Regarding soil N2O emissions (Fig. 2c), both the NT and CT treatments peaked at week 4 during the 2022 and 2023 oilseed rape cropping seasons, at 91.7 and 85.6 mg m− 2 h− 1, and 73.5 and 103.5 mg m− 2 h− 1, respectively. During both years, both treatments registered the highest N2O emissions in the third week of the rice season, with CT eliciting significantly higher level of emissions than NT.
CH4, N2O and CO2 emissions and GWP and CEE during oilseed rape-rice cropping seasons
To determine the effects of NT and CT on soil GHG emissions, we calculated the total CH4, CO2, and N2O emissions during the oilseed rape-rice cropping seasons. During both oilseed rape cropping seasons as well as during the annual oilseed rape-rice rotation, NT resulted in a significant reduction in total CH4 emissions compared with that observed for CT (Fig. 3a). Specifically, NT led to a reduction in total CH4 emissions of 16.0% and 24.2% in the two oilseed rape cropping seasons, 14.5% and 29.4% in the two rice cropping seasons, and 15.0% and 27.4% in the annual oilseed rape-rice rotation, respectively, compared with that observed for CT. Moreover, the pattern of total soil N2O emissions between the NT and CT treatments was consistent with that of CH4 emissions (Fig. 3b). Specifically, compared with CT, NT significantly reduced N2O emissions in 2022 by 22.1% during the oilseed rape cropping season, by 21.7% during the rice season, and by 21.9% during oilseed rape-rice rotation. Consistently, in 2023, NT treatment significantly reduced N2O emissions by 16.6% during the oilseed rape cropping season, by 19.2% during the rice season, and by 17.9% during the oilseed rape-rice rotation by 17.9%.
Conversely, total CO2 emissions under the NT and CT treatments showed a pattern different from those of CH4 and N2O (Fig. 3c). Specifically, NT treatment had no significant effect on soil N2O emissions in the oilseed rape-cropping season in 2022 but resulted in a significant reduction of 16.7% and 17.8% during the rice and annual rotation seasons, respectively, compared to the CT treatment. In turn, in 2023, CO2 emissions followed the same pattern as CH4 and N2O emissions. Specifically, compared with the CT treatment, the NT treatment resulted in significantly reductions in total N2O emissions of 13.2%, 20.8%, and 17.0% in the rapeseed, rice, and annual seasons, respectively. Similarly, the patterns of CO2 emissions were consistent between the NT and CT treatments for both GWP and CEE (Fig. 3d and 3e).
Carbon fixed during oilseed rape and rice seasons
Table 1 shows the effects of CT and NT treatments on soil carbon fractions and plant responses to carbon fixation. The findings show that NT treatment significantly increased DOC and MBC contents in the soil, compared to CT, regardless of cropping season. In terms of soil ROC, no major difference was found between the NT and CT treatments during either the rapeseed or rice seasons. However, during the 2023 oilseed rape season, the soil ROC content for the CT treatment was 27.9% higher than that for the NT treatment, whereas the opposite trend was observed during the rice season; additionally, soil ROC content of the NT treatment was 15.7% higher than that of the CT treatment. Notably, soil ROC content between the two treatments varied significantly. Similar to the pattern of soil ROC content, SOC content in both NT and CT treatments was not significantly different during the oilseed rape season in 2022. Unlike during the rice season, NT was significantly higher than that in the CT treatment by 12.3%. Meanwhile, NT resulted in a higher SOC content than the CT treatment during both the oilseed rape and rice seasons in 2023, which were significantly higher by 12.3% and 10.8%, respectively. Moreover, plant carbon fixation was significantly higher for the CT treatment than for the NT treatment both in 2022 and 2023. However, the CER for the NT treatment was significantly higher than that for the CT treatment, with increases of 15.9% and 59.4%, respectively.
Table 1
Yield and carbon fixed of oilseed rape and rice seasons.
Treatment | DOC (mg/kg) | ROC (g/kg) | MBC (mg/kg) | SOC (g/kg) | Plant C fixed (Mg ha− 1) | CER |
2022 | Oilseed rape season | NT | 31.73a | 9.92a | 378.17a | 31.73a | | |
CT | 19.98b | 10.07a | 292.85b | 19.98b | | |
Rice season | NT | 27.83a | 11.82a | 357.78a | 27.83a | 6.54b | 4.25a |
CT | 22.99b | 9.62a | 248.02b | 22.99b | 6.86a | 3.66b |
2023 | Oilseed rape season | NT | 27.89a | 6.03b | 239.24a | 17.34a | | |
CT | 22.21b | 7.71a | 199.7b | 15.44b | | |
Rice season | NT | 33.73a | 8.97a | 363.35a | 15.95a | 5.62a | 3.37a |
CT | 26.07b | 7.75b | 245.15b | 14.39b | 4.27b | 2.12b |
Different lowercase letters indicate significantly different values (P < 0.05). |
Relationship between soil GHG emissions and soil carbon fractions
Figure 4 shows the relationship between soil carbon fractions, soil nutrients, and crop yields for the NT and CT treatments. Pearson’s correlation analysis indicated that in the NT treatment (Fig. 4a), there was a significant negative correlation between soil SOC content and MBC and TN. Additionally, MBC content was significantly and positively correlated with ROC and TN contents, whereas ROC content was significantly and positively correlated with yield. In contrast, for the CT treatment, only soil MBC and TN were significantly and positively correlated (Fig. 4b).
Mantel test further revealed that soil CH4 emissions from the NT treatment were positively and significantly correlated with SOC and crop yield (Fig. 4a), whereas CO2 emissions were significantly and positively correlated with MBC, ROC, and yield. Additionally, N2O emissions were significantly correlated with yield. In contrast (Fig. 4b), soil CH4 emissions from CT treatments were significantly and positively correlated with MBC, DOC, TN, and crop yield, and CO2 emissions were significantly and positively correlated with MBC and TN. However, N2O emissions were significantly correlated with yield only.
Effect of NT and CT treatments on soil microbial community
The richness and diversity of the soil microbiota operational taxonomic unit (OUT) level were significantly higher for NT than for CT in regard to the pmoA gene (Figs. 5a, b), whereas no significant difference was observed for soil microbiota richness and diversity between the NT and CT treatments for the mcrA gene (Figs. 5c,d).
Using the NCBI BLAST algorithm, the operational taxonomic units were taxonomically classified into different genera for all treatments. Figure 6a illustrates the top ten most abundant soil microbial genera containing the pmoA genes. These genera included Methylocystis, Methylomicrobium, Methyloparacoccus, Methylobacter, Methylosarcina, Methylomonas, Methylococcus, Methylosoma, Methylogaea, and Methylocaldum. The genera Methylocystis and Methylomicrobium were the most dominant groups, accounting for 74.5% of pmoA sequences under NT treatment. However, under CT treatment, the dominant groups were Methylocystis and Methyloparacoccus, accounting for 69.8% of pmoA sequences. Figure 6b displays the top ten most abundant genera containing the mcrA gene. Methanocella, Methanomassiliicoccus, Methanobacterium, Methanothrix, Methanoregula, Methanosphaerula, Methanosarcina, Luteitalea, Bacillus, and Methanospirillum were the dominant genera identified in this study. Methanocella and Methanomassiliicoccus accounted for 38.1% and 28.4% of mcrA sequences under NT treatment, and 47.1% and 19.1% of mcrA sequences under CT treatment, respectively.
To compare the differences in species composition between treatments and demonstrate the relationship between treatments and trends in species abundance distribution, we performed heatmap compositional analyses (Figs. 6c,d). The top 20 genera, based on their mean abundances, were used for this analysis. Based on the results shown in Fig. 6c, there was a negative correlation between the NT treatment and Methylogaea, Methylosinus, Methylomonas, Methyloparacoccus, Methylococcus, and Methylosoma for the pmoA gene; Methylosarcina, Methylomicrobium, Methylocystis, Methylocaldum, Methylobacter, and Methylomagnum were positively correlated with each other. However, the CT treatment exhibited a correlation opposite to that of the NT treatment. For the mcrA gene (Fig. 6d), both the NT and CT treatments exhibited similar patterns for the pmoA gene. Conversely, CT treatment exhibited the opposite trend. NT treatment displayed a positive correlation with Methanomassiliicoccus, Methanosphaerula, Methanospirillum Methanothrix, Bacillus, Methanoregula, and Luteitalea, while it was negatively correlated with Methanobacterium, Methanocella, and Methanosarcina. In contrast, the CT treatment showed the opposite trend to that of NT.