【作者】 江长胜；

【导师】 王跃思； 郑循华；

【作者基本信息】 中国科学院研究生院（大气物理研究所） ， 大气物理与大气环境学， 2005， 博士

【摘要】 本研究主要以位于四川省盐亭县内中国科学院盐亭紫色土农业生态实验站为实验平台,2002.11-2005.5连续两年半时间在田间原位对该地区四种常见耕作制度下农田CH₄、N₂O和CO₂的排放同时进行了研究。实验方法为静态暗箱-气相色谱法。研究了该地区不同耕作制度下农田CH₄、N₂O和CO₂的排放特征及其影响因素,作物植株的参与和N肥施用对农田CH₄和N₂O排放的影响,并对不同耕作制度下农田生态系统NEE和CH₄、N₂O和CO₂排放所产生的综合增温潜势（GWP）分别进行了初步估算。研究表明,川中丘陵区冬灌田CH₄排放通量较高,水稻生长季平均通量值为21.4±1.8 mg m^-2·h^-1,该值比以前同类地区研究报道值低得多,且全年CH₄排放量主要集中在水稻生长季,冬灌田休闲期虽然时间很长,但在此期间CH₄排放量平均只占全年CH₄总排放量的24.2%。水稻植株对冬灌田CH₄排放具有明显的促进作用,水稻生长季水稻种植区CH₄排放量约是无水稻种植区的3倍。冬灌田CH₄排放通量与5cm深土壤温度呈极显著正相关（P<0.001）,而与水层深度的对数值却呈极显著负相关关系（P<0.01）。CH₄累积排放量与水稻地上部分生物量呈极显著正相关（P<0.001）。尿素态N肥的施用量从0提高到150 kgN·hm-2时,对冬灌田CH₄排放的影响不显著。采用不同的耕作制度,对川中丘陵区农田CH₄排放具有重要影响。在水稻生长季,相对于常规耕作冬灌田,采用冬灌田强化栽培、稻-麦轮作和稻-油轮作耕作方式稻田CH₄排放量平均分别下降了33.2%、54.5%和57.8%。无论冬灌田还是水-旱轮作稻田,在水稻分蘖盛期、拔节孕穗期和抽穗扬花期都是控制CH₄排放的最有效时期。农田在旱作阶段,通气良好的旱地土壤都为大气CH₄的弱汇,但这种吸收量非常有限,全旱季CH₄吸收量都低于5 kg·hm^-2,远远小于水稻季CH₄的排放量。旱地土壤中尿素N肥施用量的提高会降低土壤对CH₄的吸收量。研究表明,川中丘陵区冬灌田也是N₂O的排放源,且主要集中在水稻生长季,更多还原

【Abstract】 To understand of methane （CH₄） and nitrous oxide （N₂O） emissions from different agro-ecosystems and carbon dioxide （CO2） exchange between the cropland and atmosphere of Central Sichuan hilly areas of Southwest China, a field experiment was conducted in situ for two and a half years using the static opaque chamber method in Yanting Experimental Station of the Agricultural Ecology of Chinese Academy of Sciences （CAS）. In this study, we investigated the effects of N, crop plant and other environment factors on CH₄ and N₂O emissions from local prevailing four agro-ecosystems and CO2 exchange between the cropland and atmosphere, and also the global warming potential （GWP） of the emission of CH₄, N₂O and CO₂ under the four agro-ecosystems were assessed in an integrated manner. The results showed that the average fluxes of CH₄ from the permanently flooded rice field with a single middle rice crop and flooded with no winter crop （hereafter referred to as PF） were 21.4±1.8 mg m^-2·h^-1 and 3.8±1.0 mg m^-2·h^-2 during the rice-growing and non-rice-growing stages, respectively, where both values were much lower than those of many previous reports from similar regions in Southwest China. The CH₄ missions from PF were intensive in the rice-growing stage, being only 24.2% of the total annual CH₄ emission emitted during the non-rice-growing stage, though the latter occupied two thirds of the year. Rice plant stimulated CH₄ emission from PF remarkably, the total CH₄ emission from rice-involved plot was ca. 3 times than that from rice-uninvolved plot during the rice-growing stage. Correlation analysis demonstrated that CH₄ emissions from PF were significantly positively correlated with the soil temperature under 5cm deep soil （P<0.001）, and were significantly negatively correlated with the water depth of log-transformed （P<0.01）, and the aggregate CH4 emissions from PF were significantly positively correlated with the above-ground dry weight rice biomass （P<0.001）. The CH4 emissions from PF had no remarkable effect when the application of urea-based N fertilizer was increased from the rate of 0 to 150 kgN?¤ha-1 （P>0.1）. Cultivation systems have important effects on CH4 emission from croplands of the Central Sichuan hilly areas of Southwest China. After implementing SRI model （an innovatory method of rice planting）, rice-wheat rotation （RW） and rice-oilseed rape rotation （RR） from the PF, the CH4 emissions were reduced substantially, being only 66.8%, 45.5% and 42.2% of those of the PF treatment, respectively. Whether for the cultivation system of the permanently flooded rice fields or of the paddy rice-upland crop rotation systems, the optimum stages of developing mitigation options of methane are when rice is in later tillering stage, stem elongation stage or heading and flowering stage. When croplands are in upland crop season, dry aerobic soils can absorb CH4 from atmosphere and become a weak CH4 sink. But the amount of CH4 absorption from atmosphere by soil during the upland crop season was less than 5 kg?¤ha-1?¤yr-1, and was much lower than that of CH4 emission from paddy field during the rice growing stage. Increasing the application rate of urea-based N fertilizers decreased CH4 absorption by soil from atmosphere in upland crop season. The results showed that PF was also a N2O emission source, and the N2O emissions from PF were intensive in the rice-growing stage, being only 22.1% of the total annual N2O emission emitted during the non-rice-growing stage. Rice plant also stimulated N2O emission from PF remarkably, the total N2O emission from rice-involved plot was ca. 1.7 times than that from rice-uninvolved plot during the rice-growing stage. Applying urea-based N fertilizers at a rate of 150 kgN?¤ha-1 simulated the seasonal N2O emission in comparison with the no N application treatment.Cultivation systems have insignificant effects on N2O emission from croplands during the rice-growing stage. After implementing SRI model, RW and RR systems from the PF, the N2O emissions were enhanced by 0.7%, 11.6% and 17.5% as compared to the PF treatment, respectively. The mean N2O flux was higher in upland crop season than in the rice-growing season. As for the paddy rice-upland crop rotation systems, the largest peak of N2O emission in the whole year occurred after the application of basal fertilizers during the early wheat or oilseed rape growing season. It was calculated that the amounts of N2O emission in RW and RR during the following twenty days after the wheat or oilseed was sowed accounted for 22.1% and 24.8% of the respective annual of N2O emission in RW and RR systems. The crop plant of the upland crops （such as wheat, oilseed） can stimulate the N2O emission from cropland, but the root of maize plant could restrain the N2O emission from soil. There is a wide uncertainty of 0.50%².64% for the annual estimates of direct N2O emission factor （Efd） of the Central Sichuan hilly areas of Southwest China, and the Efds are changing with the cultivation systems, the application rate of N fertilizer and different years. There was a big difference in a year-round N2O emission amounts between cultivation systems. After implementing RW, RR and permanent upland crop field （PU） systems from the PF, the N2O emissions were enhanced by 191.0%, 117.2% and 190.6% as compared to the PF treatment, respectively. Soil-crop total ecosystem respiration rate in crop planted plot and the soil heterotrophic respiration rate in no crop planted plot of different cultivation systems all were showed an evident seasonal variations, which were mainly controlled by air temperature and crops?ˉ life processes. Different types of crop had different dynamics of above-ground biomass and root/shoot ratio and NEE （net ecosystem CO2 exchange） with the crop growth. The amount of carbon sequestration from atmosphere to crop ecosystem was followed this relationship: RW>PF>RR>PU, and carbon sequestration mainly occurred in the rice-growing stage. The GWPs of the emission of CH4 and N2O under four cultivation systems wereassessed in an integrated manner. With 100-year spans, the integrated GWPs of the mean annual CH4 and N2O emission followed the relationship: PF>RR??RR>PU, where the GWP of PF is 1.8, 1.9 and 5.5 times than that of RW, RR and PU, respectively.更多还原