基于增强回归树的红壤旱坡花生地气态氮损失影响因素分析

    Boosted Regression Tree-based Identification of Environmental Factors Influencing Gaseous Nitrogen Loss in Peanut Field on Red Soil Arid Hillside

    • 摘要: 为探究红壤旱坡花生地气态氮(NH3和N2O)排放特征及其影响因素,采用通气法和密闭式静态暗箱-气相色谱法,原位监测翻耕和免耕条件下红壤旱坡花生地土壤NH3挥发和N2O排放的动态变化特征,并基于增强回归树(boosted regression tree,BRT)方法识别关键影响因子。结果表明:(1)整个花生生育期翻耕和免耕处理NH3挥发速率(以N计)变化范围分别为0.02~1.55和0.02~1.05 kg·hm-2·d-1,氨挥发累积量(以N计)分别为(17.19±8.56)和(18.38±7.41)kg·hm-2,占总施氮量的(11.77±5.86)%和(12.59±5.08)%,热点时段主要集中在施基肥后15 d内;翻耕和免耕处理N2O排放通量(以N计)变化范围分别为0.07~2.90和0.02~3.97 mg·m-2·d-1,累积量(以N计)分别为(0.81±0.27)和(0.68±0.10)kg·hm-2,占总施氮量的(0.55±0.17)%和(0.46±0.06)%,N2O排放通量热点时段不明显;两种耕作条件下NH3挥发无明显差异,但免耕降低了16.05%的N2O排放累积量;氨挥发是红壤旱坡花生地氮素气态损失的主要途径。(2)采用增强回归树(BRT)分析发现,土壤铵态氮含量、施肥后天数、花生生育期、前3天降水量和土壤硝态氮含量是红壤旱坡花生地氨挥发的关键影响因子,贡献率分别为47.92%、14.78%、8.21%、7.44%和5.91%;而N2O排放的关键影响因子分别为土壤含水量、土壤铵态氮含量、地温、前3天降水量、土壤硝态氮含量、施肥后天数和气温,相对贡献率分别为24.67%、20.34%、12.26%、9.93%、9.91%、9.64%和8.51%。上述研究结果表明,施肥是影响红壤花生地氮肥气态氮损失的重要因子,气态氮(NH3和N2O)损失最高可占施肥量的18.35%,微生物硝化过程及其环境控制因子可能在土壤氮损失过程中发挥了重要作用。这些结果可为提高南方红壤区坡地氮肥利用率、减轻环境污染提供理论依据。

       

      Abstract: To clarify the characteristics and influencing factors of gaseous nitrogen (N2O and NH3) loss in peanut field on red soil arid hillside, aeration method and closed static opaque chamber-gas chromatography (GC) methods were adopted. In situ monitoring was carried out to reveal the dynamic change characteristics of NH3 volatilization and N2O emission of the peanut field on red soil arid hillside under conditions of plowing and no-tillage. Their key influencing factors were identified based on boosted regression tree (BRT) method. The results show that: (1) The variation range of NH3 volatilization rate (measured in N) during the whole peanut growth period under plowing and no-tillage conditions was 0.02-1.55 and 0.02-1.05 kg·hm-2·d-1, respectively, the cumulative volatilization was (17.19±8.56) and (18.38±7.41) kg·hm-2, respectively, accounted for (11.77±5.86)% and (12.59±5.08)% of the N application, respectively, and the NH3 volatilization was concentrated within 15 d after base fertilizer application. The variation range of N2O emission flux under plowing and no-tillage conditions was 0.07-2.90 and 0.02-3.97 mg·m-2·d-1, respectively, the cumulative emission was (0.81±0.27) and (0.68±0.10) kg·hm-2, respectively, accounted for (0.55±0.17)% and (0.46±0.06)% of the N application, respectively, and the hot period of N2O emission flux was not obvious. No significant difference in NH3 volatilization between two tillage treatments were found, while no-tillage reduced cumulative N2O emission by 16.05%. Ammonia volatilization was the main way of nitrogen gas loss in peanut field of red soil arid hillside. (2) The BRT analysis findings, soil ammonium-nitrogen content, days after fertilization, peanut growth period, rainfall in the previous 3 days, soil nitrate-nitrogen content were identified as the key factors affecting ammonia volatilization of the peanut filed on red soil arid hillside. The contribution rates were 47.92%, 14.78%, 8.21%, 7.44% and 5.91%, respectively. However, the key influencing factors of N2O emission and their relative contributions were soil moisture (24.67%), soil ammonium-nitrogen content (20.34%), ground temperature (12.26%), rainfall in the previous 3 days (9.93%), soil nitrate-nitrogen content (9.91%), days after fertilization (9.64%), and temperature (8.51%). The above results show that fertilization is an important factor affecting the gaseous loss of nitrogen fertilizer in peanut field of red soil, and the highest loss of gaseous nitrogen (NH3 and N2O) accounted for 18.35% of the N applied. Microbial nitrification process and its environmental control factors may play an important role in soil nitrogen loss. These results could provide a theoretical basis for improving nitrogen utilization rate and mitigating environmental pollution in hillside agriculture in red soil region of South China.

       

    /

    返回文章
    返回