【作者】 苗正红；

【导师】 王宗明；

【作者基本信息】 中国科学院研究生院(东北地理与农业生态研究所) ， 地图学与地理信息系统， 2013， 博士

【摘要】 土壤碳储量的微小变化就能使生态系统发生碳―汇-源‖或―源-汇‖的转变，对全球生物地球化学循环产生重要影响。准确估算土壤有机碳储量、揭示其分布格局及动态将有助于预测陆地生态系统对气候变化的反馈关系。三江平原是我国最大的淡水沼泽湿地分布区，也是近50年来湿地开发最严重的地区，土地利用变化剧烈打破了三江平原生态平衡，影响了三江平原土壤有机碳总储量的变化和分布，本文运用遥感和GIS技术，结合统计学和地统计学方法，估算三江平原1980s年和2010年土壤有机碳储量，确定土壤有机碳密度的控制因素，并且分析其动态变化规律和影响因素。这有助于我们对三江平原土壤有机碳的碳源/碳汇的认识，从而对三江平原湿地保护、政府宏观决策和我国重要农业区的土壤有机碳循环研究具有重要的意义。主要结论如下：1.三江平原土壤有机碳密度及其控制因子。三江平原1980s年和2010年中不同植被类型（0-30cm，0-60cm，0-100cm）土壤有机碳密度的平均值从高到低的顺序均为：沼泽湿地>林地>草地>水田>旱地。1980-2010年三江平原（0-30cm，0-60cm，0-100cm）沼泽湿地土壤有机碳密度平均值均有所降低。土壤质地是影响三江平原土壤有机碳密度空间分布特征的重要环境因素，其中土壤黏粒对其影响最大。年均气温和年均累积降雨量也对土壤有机碳密度有一定的影响。2.三江平原土壤有机碳的垂直分布特征及其影响因素。三江平原1980s年和2010年不同类型的土壤有机碳含量与密度垂直分布特征为自上向下逐渐降低的趋势。土壤有机碳密度和含量更多地集中于表层。1980-2010年间表层土壤有机碳占1m深度总量的比例呈现显著下降趋势。1980s年旱地表层（0-30）土壤有机碳占1m深度总量的比例（54%）显著大于水田表层（0-30）土壤有机碳占1m深度总量的比例（53%），2010年沼泽湿地表层（0-30cm）和中层（30-60cm）土壤有机碳所占1m深度总量的比例相对1980s年均有所下降。三江平原土壤有机碳密度随着土壤深度的增加，其与环境变量的相关性在减弱。表层两期土壤有机碳密度均与年累积降雨量、土壤黏粒、土壤粉粒呈现显著正相关（P <0.01），但与年均温度和土壤砂粒显著负相关（P <0.01）。1980s年和2010年表层土壤有机碳密度占比例与土壤粉粒含量呈显著正相关。3.三江平原土壤有机碳储量时空动态。1980-2010年三江平原土壤有机碳储量均呈现逐渐减少的趋势。三江平原表层土壤有机碳库三十年减少了19.84%。1980s年三江平原1m土壤有机碳库为2930.46TgC，2010年三江平原土壤有机碳库为2324.34Tg C，三十年间减少的土壤有机碳储量为606.11Tg C。两期表层土壤有机碳库约占其当年1m土壤有机碳库的50%。1980s年三江平原不同土地覆被类型中林地的土壤有机碳储量最大，草地最低。经过三十年三江平原土地利用等变化的影响，到2010年林地的土壤有机碳储量虽然最高，但是与旱地土壤有机碳总储量相近。从空间上看，三江平原两期不同深度土壤有机碳密度空间差异明显。三江平原1m土壤有机碳密度研究区约有60%以上的象元中土壤有机碳密度都在减少，东南地区高于西部地区，减少最多值在40kg/m2以上，而三江平原东北地区呈现增加的趋势，增加最多值在10kg/m2以上。4.三江平原土壤有机碳储量动态变化的成因分析。三江平原过去三十年土地利用变化较为剧烈，主要发生在农田和其它土地利用类型之间。三江平原东北部农田增加的面积较大。沼泽湿地开垦、林地开垦和草地开垦三种农田开垦是影响三江平原土壤有机碳储量减少的主要原因，并且研究表明开垦对土壤有机碳的降低主要体现在表层。三江平原由于湿地开垦导致损失的土壤有机碳1m储量为86.84TgC，这也是三江平原东北部地区不同深度土壤有机碳减少的主要原因之一。林地和草地的开垦导致损失的土壤有机碳储量为24.78Tg C和2.01Tg C。退耕还湿和退耕还林均可能增加三江平原土壤有机碳储量。1980-2010年三江平原农田转化成沼泽湿地导致土壤有机碳储量增加19.02TgC。退旱地改沼泽湿地和退旱地改林地均更有助于表层土壤有机碳的增加。化肥施用、水田占农田比重和灌溉方式三种农田管理方式也是影响三江平原土壤有机碳的重要因素。更多还原

【Abstract】 Small changes of soil carbon stock could make the sinks-source "or" source-sink"conversion in ecosystem carbon, it plays a major impact on global biogeochemicalcycles. Accurate estimatting of soil organic carbon and revealing the distributionpattern of soil carbon stock and its dynamics will help us to predict a feedbackrelationship of terrestrial ecosystems to climate change. Sanjiang Plain is the largestfreshwater wetland distribution in China and one of the most development area in thepast50years. The ecological balance of the Sanjiang Plain has been broken, as aresult, the change and the distribution of the total reserves of the soil organic carbonin the Sanjiang Plain are being affected. So we use the remote sensing and GIStechnology, combined with statistical and geostatistical methods to estimate theSanjiang Plain soil organic carbon reserves in1980s and2010. We determined thecontrolling factors of soil organic carbon density, and analyzed the dynamic variationand influencing factors. That helps us understand the carbon source of the SanjiangPlain. It is important for us to protect Wetlands in Sanjiang Plain, help the governmentmacro making decision and research agricultural area of the soil organic carbon cycleresearch. The main conclusions are as follows:1.The density of soil organic carbon and controllable factors in Sanjiang Plain. Theaverage of the different vegetation types (0-30,0-60,0-100cm) soil organic carbondensity in descending order are as follow: wetlands> woodland> grassland> paddy>upland of Sanjiang Plain in1980s and2010. The average marsh soil organic carbondensity in Sanjiang Plain (0-30,0-60,0-100cm) from1980s to2010are reduced. Soiltexture is an important environmental factor that affect the spatial distributioncharacteristics of soil organic carbon density in the Sanjiang Plain, soil clay had thegreatest impact. The average annual temperature and average annual cumulativerainfall impact on soil organic carbon density.2. The vertical distribution characteristics of the Sanjiang Plain soil organic carbonand its influencing factors. The trend of different types of soil organic carbon contentand density of the vertical distribution of characteristics in1980s and2010from thetop to down were decrease. Soil organic carbon density and content were moreconcentrated on the surface. The ratio of surface soil organic carbon accounted for1m depth from1980-2010showed a significant downward trend. Dry farmland (0-30cm)accounted1m depth total soil organic carbon (54%) was significantly greater than inthe paddy field topsoil (0-30) organic carbon accounting the1m depth proportion ofthe total (53%) in1980s. The ratio of marsh in0-30cm and30-60cm accounted1mdepth were decreased in2010, which were50%and28%respectively. Soil organiccarbon density in Sanjiang Plain increased by the soil depth, and its correlation withenvironmental variables was on the wane. It is significant positive correlation (P<0.01) for the topsoil organic carbon density and annual cumulative rainfall, soil claysoil silt, but it is a significant negative correlation (P <0.01) with the average annualtemperature and soil sand. The total proportion of soil silt content of soil organiccarbon density was a significant positive correlation in1980s and2010.3. The space-time dynamic of soil organic carbon storage in the Sanjiang Plain. Itshowed a decreasing trend from1980s to2010. The topsoil organic carbon pool of theSanjiang Plain had decreased19.84%during the three decades. The1m soil organiccarbon pool in1980s in Sanjiang Plain was2930.46Tg C. And it was2324.34Tg Cin2010. It had reduced606.11Tg C of soil organic carbon reserves during the threedecades. Topsoil organic carbon pool accounts for about50%of its1m soil organiccarbon pool in1980s and2010. Woodland was the largest and grass was the minimumin different land cover types of soil organic carbon reserves in1980s. But in2010thesoil organic carbon of woodland was the highest, the value was near dry farmland.Judging from space, it is obvious for the difference between soil organic carbon fortwo periods of time. There are more than60%pixels in the1m soil organic carbondensity of Sanjiang Plain in the study area are reduced, These in the south-east werehigher than the western region, more than40kg/m2. The most increasing area was inin Northeast of Sanjiang Plain and in a maximum of more than10kg/m2.4．Factors of dynamic changes in soil organic carbon of the Sanjiang Plain. Land usewas large changing from1980s to2010which is more dramatic occurs mainlybetween farmland and other land use types. The area of dry farmland in northeastincreased most. The reclamation of marsh、woodland and grassland are the majorreason of the decrease of soil organic carbon stocks. Our studies have shown that thereclamation of soil organic carbon reduction is mainly reflected in the topsoil.1mSanjiang Plain marsh reclamation led to the loss of soil organic carbon reserves of86.84Tg C, it is also one of the main reasons for northeastern in different depth ofsoil organic carbon reduction. The reclamation of woodland and grassland led to the loss of soil organic carbon reserves of24.78Tg C and2.01Tg C. Dry farmland tomarsh and returning farmland to forests may increase the Sanjiang Plain soil organiccarbon stocks. From1980s to2010, dry farmland into wetlands led to the increase by19.02Tg C in soil organic carbon reserves. Dry farmland to marsh and to woodlandare more conducive to the increase of the topsoil organic carbon. Chemical fertilizer,the proportion of paddy in land and irrigation methods affect the Sanjiang Plain soilorganic carbon is also an important factor in three kinds of farmland management.更多还原