【作者】 莫志鸿；

【导师】 李玉娥；

【作者基本信息】 中国农业科学院 ， 气象学， 2012， 硕士

【摘要】 天然草原作为陆地生态系统的重要组成部分，在吸收和固定大气CO₂过程中发挥着重要作用，但由于超载放牧、管理不当、气候变化、投入不足等原因已造成我国北方草原生态系统发生不同程度退化、生产力下降、生物多样性减少以及生态功能弱化。探究草原生态系统碳收支分量对气候变化的响应反馈对于科学评估其源汇贡献、减缓与适应气候变化等方面具有重要的现实意义。本文以高寒草甸、温性草甸草原、温性草原和温性荒漠为研究对象，利用CENTURY模型模拟分析1961-2010年4类草原生态系统净初级生产力（NPP）、异养呼吸（R_h）和土壤有机碳（SOC）对气候变化的动态响应特征，进一步模拟研究考虑和不考虑大气CO₂浓度增加的直接效益（简称“CO₂增益”）2020s（2011-2040年）、2050s（2041-2070年）、2080s（2071-2100年）时段草原生态系统NPP、R_h和SOC对气温和降水量变化的响应反馈，得到以下主要结论：（1）1961-2010年草原生态系统NPP、R_h和SOC的动态特征及其相互关系： NPP具有明显的年际波动特征，NPP的平均值表现为高寒草甸>温性草甸草原>温性草原>温性荒漠，其中高寒草甸NPP呈极显著的波动增加趋势，其余变化趋势不显著。 R_h的波动特征与变化趋势与NPP一致，两者存在同步变化关系。总SOC的年际波动较小，且主要由活性SOC和缓性SOC的波动变化引起，其中温性荒漠总SOC呈极显著的下降变化趋势。草原生态系统中NPP、R_h和SOC之间具有内在的相关性，其中R_h与NPP同步变化关系是通过活性SOC作为载体因子关联起来，总SOC是NPP与R_h动态平衡的综合反映。（2）不考虑大气CO₂增益作用，A2和B2情景下2020s、2050s、2080s时段草原生态系统NPP、R_h和SOC对气温和降水量变化的响应反馈结果： NPP的平均增幅（较1961-1990年基准时段，下同）分别为3.6%、9.6%、14.8%和1.6%、4.6%、6.5%，其中温性草原增幅最大，其次是温性荒漠，而温性草甸草原和高寒草甸变幅稍小。 R_h的平均增幅分别为3.4%、9.2%、14.3%和1.9%、4.3%、6.7%，其中温性草原增幅最大，其次是温性荒漠，而温性草甸草原和高寒草甸相对变幅稍小。 SOC的平均减幅分别为-1.0%、-2.4%、-4.4%和-1.3%、-2.6%、-4.1%，其中温性草甸草原减幅最大，其次是高寒草甸和温性荒漠，而温性草原减幅较小。（3）考虑大气CO₂增益作用，A2和B2情景下2020s、2050s、2080s时段草原生态系统NPP、R_h和SOC对气温和降水量变化的响应反馈结果： NPP的平均增幅分别为19.0%、29.9%、42.4%和7.7%、12.0%、16.0%，较不考虑大气CO₂增益作用显著增加，其中温性草原增幅最大，其次是高寒草甸和温性荒漠，而温性草甸草原增幅稍小，表明大气CO₂增益能显著提高草原生态系统生产力。 R_h的平均增幅分别为15.3%、21.6%、32.6%和5.9%、9.5%、12.9%，较不考虑大气CO₂增益作用显著增加，其中温性草原增幅最大，其次是温性荒漠和温性草甸草原，而高寒草甸增幅较小，表明大气CO₂增益可以刺激增强草原生态系统的呼吸作用。 SOC的平均减幅分别为-5.4%、-8.5%、-11.9%和-2.2%、-3.7%、-5.5%，较不考虑大气CO₂增益作用减幅更大，其中温性草甸草原减幅最大，其次是温性荒漠和温性荒漠，而温性草原减幅稍小，主要原因在于大气CO₂增益作用能刺激增强草原土壤呼吸作用对土壤有机碳产生负反馈效应。更多还原

【Abstract】 Natural grassland as an important component of terrestrial ecosystem plays an essential role in theprocess of absorption and fixation atmospheric CO₂. However, as the result of overgrazing,misadministration, climate change and lack of investment, grassland ecosystem has been suffering withgrassland degradation, declining productivity, biodiversity reduction and weakening ecological function.Therefore, understanding the carbon balance of grassland ecosystem responding to climate change ispractically meaningful for scientific evaluation on the contributions to the sources or sinks ofatmospheric CO₂, and its adaptation to the climate change.Based on the calibrated and validated CENTURY model, dynamic characteristics of net primaryproduction （NPP）, heterotrophic respiration （R_h）, soil organic carbon （SOC） in four types of grasslandecosystems, namely alpine meadow （AM）, temperate meadow steppe （TM）, temperate steppe （TS） andtemperate desert （TD）, were simulated and analyzed by using observed climatic data （1961-2010） andprojected climatic data （1961-2100） by PRECIS model with or without considering the enhancedatmospheric CO₂concentration under A2and B2scenarios, respectively. These main research resultsand conclusions achieved as follows:（1） In the past50years, dynamic characteristics of NPP, R_hand SOC in four types of grasslandecosystems showed that: NPP in four types of grassland ecosystems showed large inter-annualvariation and the coefficient of variation was in sequence by TD>TS>TM>AM. However, the averageof NPP in four types of grassland ecosystems was AM>TM>TS>TD, and the trend of NPP in AMascended significantly. R_hhas the same characteristics of fluctuation and trend as NPP in four types ofgrassland ecosystems. Actually, they were in the synchronous changing relationship. Total SOC infour types of grassland ecosystems showed a little inter-annual variation and the coefficient of variationwas mainly affected by active and slow SOC. The average of total SOC in four types of grasslandecosystems was AM>TS>TM>TD. The trend of total SOC in TD declined significantly because of thedecrease in slow and passive SOC, and that trend of slow or passive SOC in AM, TM and TS alsodeclined significantly. According to the analysis of partial correlation, NPP, R_hand SOC areinterrelated in four types of grassland ecosystems. The synchronous changing relationship between NPPand R_his reached through the crucial factor of active SOC, while the results of dynamic balancebetween NPP and R_hcan be reflected by total SOC.（2） Without considering the effects of elevation of atmospheric CO₂concentration, the dynamicfeedback results of four types of grassland ecosystems responding to the changes of temperature andprecipitation in2020s,2050s and2080s under A2and B2scenarios showed that: The averagechanges of NPP in four types of grassland ecosystems in2020s,2050s and2080s compared to baseline（1961-1990） would increase by3.6%,9.6%,14.8%and1.6%,4.6%,6.5%under A2and B2scenarios,respectively. Among them, the increase rate of NPP would be TS>TD>TM>AM. The averagechanges of R_hin four types of grassland ecosystems in2020s,2050s and2080s would increase by3.4%, 9.2%,14.3%and1.9%,4.3%,6.7%under A2and B2scenarios, respectively. Among them, the increaserate of R_hwould be the same as NPP in sequence by TS>TD>TM>AM. However, the averagechanges of SOC in four types of grassland ecosystems in2020s,2050s and2080s would decrease by-1.0%,-2.4%,-4.4%and-1.3%,-2.6%,-4.1%under A2and B2scenarios, respectively. Among them,the decrease rate of SOC would be TM>AM>TD>TS.（3） With considering the effects of elevation of atmospheric CO₂concentration, the dynamicfeedback results of four types of grassland ecosystems responding to the changes of temperature andprecipitation in2020s,2050s and2080s under A2and B2scenarios showed that: Changes of NPP in2020s,2050s and2080s would be19.0%,29.9%,42.4%and7.7%,12.0%,16.0%, which wouldsignificantly increase by15.5%,20.3%,27.6%and6.1%,7.3%,9.4%compared to without CO₂enhanced effect under A2and B2scenarios, respectively. Thus, NPP of grassland ecosystems wouldbenefit from atmospheric CO₂enhancement. Changes of R_hin2020s,2050s and2080s would be15.3%,21.6%,32.6%and5.9%,9.5%,12.9%, which would significantly increase by11.9%,12.4%,18.3%and3.9%,5.1%,6.3%compared to without CO₂enhanced effect under A2and B2scenarios,respectively. Thus, heterotrophic respiration of grassland ecosystem would be stimulated by CO₂enhancement. Changes of SOC in2020s,2050s and2080s would be-5.4%,-8.5%,-11.9%and-2.2%,-3.7%,-5.5%, which would significantly decrease by-4.4%,-6.1%,-7.5%and-0.9%,-1.1%,-1.3%compared to without CO₂enhanced effect under A2and B2scenarios, respectively. The main reasonsfor causing more decrement of SOC would be that R_hof grassland ecosystem would be increased underhigher atmospheric CO₂concentration.更多还原