【作者】 夏文建；

【导师】 周卫； 梁国庆；

【作者基本信息】 中国农业科学院 ， 植物营养学， 2011， 博士

【摘要】 长江中下游稻麦轮作制中过量施氮较为普遍,这不仅导致氮肥利用率降低,而且引起环境风险。因此,田间尺度下原位定量化研究化肥氮的去向和氮素循环过程,以及区域尺度下评估氮肥管理的农学和环境效应是优化氮肥管理,提高氮肥利用率的关键。为研究优化施氮下长江中下游稻麦轮作农田氮素循环特征,于2007年至2008年在湖北省开展田间试验和15N示踪的微区试验,监测不同氮肥运筹下氨挥发、N₂O排放、无机氮动态分布和15N的去向。田间试验和15N示踪的微区试验均设置4个处理,（1）习惯施氮（小麦和水稻氮肥用量分别为225kgN/hm²和210kgN/hm²）,（2）氮肥减量（小麦和水稻氮肥用量分别为157.5kgN/hm²和147kgN/hm²）,（3）优化施氮（基于作物不同生长阶段氮素需求增加施肥比例和次数,氮肥用量同处理2）,和（4）对照（不施氮肥）。习惯施氮和氮肥减量处理小麦季基肥和拔节肥各占1/2,水稻季基肥和分蘖肥各1/2。优化施氮处理,小麦季基肥,拔节肥和孕穗肥各占1/3;水稻季基肥,分蘖肥和孕穗肥各占1/3。采用密闭室连续抽气法测定氨挥发,密闭箱技术原位监测N₂O排放,连续流动注射分析仪测定土壤NO₃^--N和NH₄+^-N含量。基于观测值,检验脱氮-分解模型（DNDC）田间尺度模拟预测氮素循环的可靠性和灵敏度。同时,以湖北潜江市为例,采用DNDC模型结合地理信息系统（GIS）技术对区域尺度优化施氮下氮素循环与平衡进行估算。研究取得以下主要进展:1.田间试验结果表明,三个施氮处理间小麦、水稻的籽粒产量和植株吸氮量差异不显著,优化施氮处理略高于习惯施氮处理。与习惯施氮处理相比,优化施氮小麦和水稻季作物表观回收率分别提高了12.5和7.96个百分点。整个小麦和水稻生长季,习惯施氮、氮肥减量和优化施氮处理氮素表观损失量分别为180、115和80.8kgN/hm²,占氮素总输入量的41.3%、37.8%和26.5%。2.小麦季施肥后氨挥发大约持续7-10天,水稻季大约持续5-7天。在整个稻麦轮作周期肥料氮氨挥发损失主要发生在水稻季,占整个轮作周期氨挥发总量的74.1%-78.6%。与传统施氮相比,优化施氮氨挥发量减少了N 26.3kg/hm²,氨挥发降低了32.9%,可见优化施氮是降低氨挥发较理想的施肥方式。3.稻麦轮作体系中土壤N₂O排放具有明显的季节性变化规律,小麦季和水稻季N₂O通量分别与土壤和田间水中无机氮含量呈正相关。N₂O排放排放量随施氮量增加而增加。小麦生育期土壤N₂O排放量范围为N₂O 2.43-4.84kg/hm²,肥料氮通过N₂O排放的损失率为0.54%-0.74%。水稻季土壤N₂O排放量为N₂O 0.89-2.45kg/hm²,肥料氮通过N₂O排放的损失率为0.39%-0.47%。与习惯施氮相比,优化施氮土壤N₂O排放减少了N₂O 1.74kg/hm²,减少了23.9%。4. 15N示踪试验结果表明,与习惯施氮相比,小麦季和水稻季优化施氮分别增加氮回收率6.19和3.07个百分点。土壤残留率小麦季增加4.76个百分点,水稻季增加2.49个百分点。肥料氮损失率小麦季和水稻季分别减少11.0和5.57个百分点。与习惯施氮相比,优化施氮处理小麦季和水稻季累积氨挥发量分别降低41.3%和54.0%,N₂O排放损失总量减少35.2%和35.3%。5.田间尺度上,采用DNDC模型模拟的土壤氨挥发速率和N₂O排放通量与田间实测结果较为吻合,氨挥发通量模拟值与实测值相关系数为0.688,N₂O排放通量模拟值与实测值相关系数为0.528,均达极显著水平,显示DNDC模型预测农田土壤氮素具有较高可信度。模拟结果显示,气温和氮肥用量是影响作物产量和吸氮量的关键因素;土壤氨挥发主要受氮肥品种影响,并随氮肥用量增加而增加;土壤N₂O排放主要受温度、土壤pH值、土壤有机碳含量影响。6.区域尺度上,通过DNDC模型与潜江市空间地理信息数据相连接,设置了习惯施氮和优化施氮两种氮肥管理情形,发现潜江市习惯施氮和优化施氮条件下氮素盈余量为897和620t。习惯施氮条件下,氮氨挥发损失量为91.2t,氮淋失量达497t,硝化反硝化损失量为485.88t;优化施氮可降低氨挥发80.7%,氮淋失38.8%,硝化反硝化62.5%。土壤养分和环境因子的空间异质性导致了区域氮输出的空间变异。更多还原

【Abstract】 Excessive nitrogen fertilization generally exists in rice-wheat rotation system in middle and lower reaches of Yangtze river, which cause not only the decrease of nitrogen use efficiency but also the increasing of the risk of environmental pollution. Therefore, identifying the N cycling processes in field scale as well as in regional scale has become an essential need for optimizing N application and improving the N use efficiency in agricultural soils in middle and lower reaches of Yangtze river. To gain better insight into the nitrogen cycling in rice-wheat rotation system under optimized nitrogen management in middle and lower reaches of Yangtze river, field plot trials combined with micro-plot experiment with 15N tracing were established in Hubei province from 2007 to 2008 to monitor ammonia volatilization （AV）, denitrification and N₂O emission, NO₃--N accumulation and 15N fate under different nitrogen application strategies. For both field trail and micro-plot experiment with 15N tracing, four N treatments included: （1） conventional N application (225 and 210kgN·ha^-1 for winter wheat and rice, respectively), （2） reduced N application （157.5 and 147kgN·ha-1 for winter wheat and rice, respectively）, （3） optimized N application （increase ratios of dress N to base N fertilizer and split application based on crop N uptake at different growth stage, same to N rate in reduced N application）, and（4） control with no nitrogen applied. For the treatments of conventional and reduced N application, 1/2 of nitrogen fertilizer applied as basal and 1/2 as topdressing at elongating stage for winter wheat, and 1/2 as basal and 1/2 as topdressing at tillering stage for rice. For optimized N application, 1/3 nitrogen fertilizer applied as basal, and 1/3 as topdressing at elongating stage and 1/3 at ear bearing stage for winter wheat, and 1/3 as basal, and 1/3 as topdressing at tillering stage and 1/3 at ear bearing stage for rice. A continuous airflow enclosure method was adopted to determine the ammonia volatilization （AV） loss, N₂O emission was analyzed using the static chamber-gas chromatograph method, and soil NO₃^--N and NH₄⁺-N was determined by flow injection analysis. Based on above observation, the reliability and sensitivity of Denitrification-Decomposition simulation model （DNDC） for forecasting nitrogen cycling and balance were tested in field scale, and influencing factors of various N cycling pathways were identified. Meanwhile, Qianjiang city of Hubei province was taken as the study area, DNDC coupled with geographic information system （GIS） was used to estimate the nitrogen cycling and balance in rice-wheat rotation system in regional scale. The main findings obtained are summed up as follows:1. Field plot trials showed that grain yield and N uptake for wheat and rice in the treatment of optimized N application were slightly higher than those under conventional N application, there is no significant difference within three N fertilizer application treatments, except no N fertilizer application. Compared to conventional N fertilization, the apparent N recovery under optimized N fertilization improved by 12.46% and 7.96% for wheat and rice, respectively. During whole growing season of wheat and rice, the apparent loss of N in the treatments of conventional, reduced and optimized N application were 179.7, 115.2 and 80.8kgN/ha respectively, accounting for 41.3%, 37.8% and 26.5% of the total N fertilizer input.2. The AV loss generally increased with the urea application rate, and affected by the temperature and soil water condition. In wheat growing season, the AV from fertilizer N lasted for 7-10 days after fertilization; while lasted for 5-7 days in rice growing season. Majority of the AV loss occurred in the rice season, occupied by 74.08%-78.65% of total AV loss in the entire rotation system. Compared to conventional N fertilization, the AV loss reduced N 26.28kg/ha, the loss rate reduced 32.90% under optimizing N application, showing that optimized N application is a rational and practicable N fertilization mode for reducing AV in rice-wheat rotation system.3. There were significant seasonal variations of N₂O emission flux during rice-wheat growing season. Total N₂O emission generally increased with the N application rate, and majority of N₂O emission occurred during the wheat season. In wheat growing season, N₂O emission varied from 2.43 to 4.84 kgN₂O/ha which accounted for 0.54% to 0.74% of applied N; While in rice growing season, N₂O emission varied from 0.89 to 2.45 kgN₂O/ha which ranged from 0.39% to 0.47% of applied N. Compared to conventional N application, N₂O emission loss rate reduced 23.87%, under the optimizing N application, indicating that optimizing N application associated with increasing ratios of dress N to base N fertilizer and times of split application based on plant N requirement at different growth stage could obviously reduce the N₂O emission under the rice-wheat rotation system.4. Micro-plot experiment with 15N tracing showed that N recovery for wheat and rice under optimizing N application was improved by 6.19% and 3.07%, respectively, the residual rate in soil increased by 4.76% and 2.49% in wheat and rice season respectively, the loss rate of fertilizer 15N decreased by 10.95% and 5.57%, respectively, in comparison with conventional N application. Cumulative AV amounts from fertilizer 15N under optimizing N application reduced by 41.27% and 54.04%, and denitrification losses reduced by 35.18% and 35.29% during wheat and rice growing seasons, respectively, compared to conventional N application.5. In field scale, a higher reliability was found between soil ammonia volatilization and N₂O emission simulated by DNDC model and field measured results, suggesting that the DNDC model could be used for describing N loss in soils under different nitrogen fertilizer management measures. According to sensitivity test on the DNDC model, both grain yield and crop N uptake were significantly affected by temperature and N application rate, AV was significantly influenced by fertilizer N type and the nitrogen application rate, N₂O emission significantly affected by temperature, soil pH and soil organic carbon.6. In regional scale, the regional N surpluses were 896.84 and 620.08t under the conventional N application and optimizing N application respectively. Compared to the conventional N application, AV, the nitrate leaching, and nitrification and denitrification loss reduced by 80.7%, 38.80% and 62.51% under optimizing N application respectively. The spatial heterogeneity in soil nutrients and environmental parameters result in the spatial variance of regional N outputs.更多还原