【作者】 谷保静；

【导师】 常杰； 葛滢；

【作者基本信息】 浙江大学 ， 生态学， 2011， 博士

【摘要】 人类活动已经强烈地改变了陆地生态系统氮的生物地球化学循环过程,在增加系统生产力、满足人类需求的同时,也带来了严重的环境和健康问题。因此,人类活动干扰下的全球氮循环不仅是生态学研究的重点之一,也是社会经济和环境可持续发展的核心内容之一。随着人类活动干扰的加强,自然生态系统的结构和功能发生破缺,形成农田、城市、种植园、牧场、工厂、矿山等功能系统,再自组织升级成人类-自然耦合系统(CHANS)。中国目前的社会经济飞速发展,环境氮污染严重,正处于全球CHANS氮循环的“热点”和“热时”。中国的案例研究可为我们更好地理解CHANS的氮循环过程,为合理利用氮循环的正面作用,缓解给人类和地球系统带来的负面影响提供依据。本论文中的CHANS以中国陆地行政边界为水平边界。垂直方向的上边界定在地面以上1千米,不包括大气环流；下边界在岩床表面,不包括资源矿。CHANS的氮循环从非活性的N2被活化为活性氮(Nr)进入系统或者系统外的Nr直接输入系统开始,以Nr氧化／还原为N2或者以Nr直接输出到系统外时终止。系统分为4个功能群：加工者、消费者,移除者和生命支持系统。加工者指将输入的Nr加工为产品的一类功能系统,包括农田、草地、森林、牲畜养殖、水产养殖、工业、以及城市绿地子系统；消费者包括人类及宠物子系统；移除者是指将废Nr进行处理并消除其负面影响的系统,包括污水处理和垃圾处理子系统；生命支持系统包括近地面大气、地表水和地下水子系统。所有从系统外输入的Nr都进入一个或几个上述子系统,然后在系统内子系统间循环流动或输出到系统外。CHANS的氮循环过程存在两个核心问题,即以人类为核心的氮供给-消费和系统可持续发展。本文围绕这两个问题,基于质量平衡法、大气遥感和地理信息系统技术,利用观测、文献、年鉴等来源,编译出中国氮循环的数据集,包括了1980-2008年间超过十万个Nr流通量以及与之相关的气温、降水、土地利用以及社会经济等数据。基于CHANS的理论和假说,利用氮循环数据集和我们构建的NCNA、 URCNC等生态系统模型,对中国近30年人类-自然耦合系统的氮循环过程进行了较全面的量化分析,并在食物氮和工业氮通量方面拓展到了全球尺度进行比较和分析。以下是主要结论：1)1980-2008年间,中国的Nr输入从24.6Tg N yr-1增加到59.6Tg N yr-1,近30年增加了1.4倍,增速为全球平均水平的2倍。以总量计,目前中国在占全球约7%的陆地面积上输入了全球人类源Nr的30%,表明了中国对全球氮循环的巨大贡献。相比自然状态下陆地生态系统氮输入,人类活动使中国陆地生态系统氮输入强度增加了3.6倍,而全球平均仅增加了1-1.5倍。近30年来,中国生物固氮量基本保持不变,这与全球的生物固氮一直增加的趋势不同；而工业固氮在总氮输入中所占的比例从50%增加到69%,高于全球水平。按消费计,工业固氮主要分布在农业发达的华北平原、东北地区、长江中下游地区及四川盆地,以及工业发达的长江三角洲和珠江三角洲等东部沿海地区；化石燃料燃烧带来的NOx-N输入虽然增加了3.7倍,然而其通量较小,在总氮输入中所占的比例维持在4%-8%。因此,食物氮和工业氮的生产-消费是中国氮通量过程的主要驱动因素。2)在中国,近30年来的系统氮输出增量未跟上输入增量,表明伴随人类活动的增强,CHANS趋于走向氮富集。这同之前有研究认为的随着系统人类源Nr输入的增加,Nr主要通过河流或者大气环流加速流失的结论相反。之前研究对系统中工业氮循环过程的忽略可能是造成这种相反结论的主要原因。近30年,中国氮积累增加了3.1倍,2008年达27.0Tg N yr-1,氮积累主要发生在农田(19.8%)、森林(31.3%)、草地(10.5%)、地下水(17.1%)以及人类(19.7%)子系统。与此同时,地表水和大气Nr浓度近30年来一直处于上升趋势,也增加了Nr在人类-自然耦合系统中的积累。3)近30年来,中国加工功能群氮输入增加了1.3倍,从59.4Tg N yr-1增加到137.8Tg N yr-1,其为人类提供的工业氮产品、植物蛋白和动物蛋白分别增加了8.2倍、1.0倍和5.5倍,但是目前人均消费量依然比发达国家低20%-25%。加工功能群的高氮输入强度显著相关于人均GDP (PGDP)、年均温和年降水,因此中国的东部和南部成为高氮输入的热点。近30年来,随着技术进步以及政策管理革新,中国加工功能群氮流失的增加速度慢于系统氮输入的增加速度。这意味着中国加工功能群的氮利用率(NUE)处于上升趋势。4)中国人均食物氮消费水平为5.2kg N yr-1,低于欧美发达国家(6.5kg N yr-1),但已高于全球平均水平(4.5kg N yr-1)。然而中国人均的动物蛋白消费比例仅为33.2%,却低于全球平均水平(38.7%),更远低于欧美发达国家(65.2%)。即便如此,中国农田粮食产出用作饲料的比例也已经高达60%,动物蛋白的生产成为农田氮产出的主要去向。参考发达国家的比例(>70%),中国粮食生产用作饲料的比例仍会增加。考虑到饮食方式的差异,中国动物蛋白消费比例可能不会达到欧美水平,但未来20年动物蛋白消费比例仍会增加50%左右。这将促进中国氮输入进一步增加。5)工业氮是除了食物氮之外人类又一重要Nr生产-消费类型,主要是纤维、房屋、家具等。本研究中首次将工业氮分为人工源(NA,如合成纤维、橡胶等)和生物源(NB,如皮革、棉花等),并估算了全球及中国的工业氮通量。2008年,中国人均工业氮消费量仅为3.4kg N yr-1,低于全球平均水平(4.3kg N yr-1),更远低于欧美发达国家水平(>10kgN yr-1)。人均NA消费量与PGDP和城市化水平显著相关,因此高NA消费主要出现在高城市化水平的发达地区。NB消费量与社会经济参数不具有相关性,因而NB消费量在不同区域之间差异较小。但是不同的文化使NB的消费产生区域差异。6)本文将工业氮产品区分为结构性和非结构性氮两类。其中结构性氮的比例超过70%,导致工业氮倾向于积累在人类居住区。2008年全球工业氮在人类居住区的积累量估算为～21Tg N,这可以解释全球人类源“氮失汇”(-26Tg N yr-1)的81%。结构性工业氮在人类居住区能存留数十年到上百年,直到最终被燃烧或分解释放。这个巨大的时滞缓解了Nr快速释放造成的环境氮污染,然而这种时滞也使垃圾产生量的估测十分困难,政策制定者需要关注这种时滞带来的遗留效应,避免出现“垃圾围城”现象,实现社会经济和环境的可持续发展。7)2008年中国向大气排放的Nr污染物量达到16.9Tg N yr-1,近30年来增加了约1倍,其中NOx排放的增速略快于NH3和N2O。N20的释放量2008年时已达到1.1Tg N yr-1,其温室效应相当于0.2Pg C yr-1CO2当量。上述采用质量平衡法估算的时空尺度上的大气Nr通量变化与大气遥感的反演结果基本一致(R2=0.80-0.94)。华北平原的河南、山东和河北,长江流域的四川和江苏是NH3和N20主要的释放源,而NOx的排放则主要集中在东部沿海地区。NH3和N20的释放主要受农业发展驱动,而NOx排放则受社会经济发展驱动。8)近30年来,农业面源Nr流失贡献了中国地表水氮污染的～60%,生活和工业点源污染排放占30%左右。地表水氮污染主要出现在华北平原以及东北和西北部分地区。南方地区虽然Nr污染通量很大,但是由于地表水资源丰富,稀释作用使水体污染度小于北方。人类自身的氮消费以及相关的人类活动是这些区域地表水Nr污染差异的主要原因,贡献值是人口>城市化>PGDP。中国地下水Nr严重污染,特别是华北平原、东北地区以及长江三角洲地区,平均地下水氮浓度超过20mgN L-。地下水中的氮富集受人口密度、PGDP以及城市化水平影响,而与自然因素与地下水氮富集不相关,证明了人类活动对地下水Nr污染起主导作用。对CHANS氮循环过程未来动态的情景模拟分析发现：大气Nr污染控制的关键因素是技术进步,地下水Nr污染控制的关键因素是政策革新,而地表水Nr污染控制的关键因素则是技术进步和政策革新的共同作用。更多还原

【Abstract】 Human activities have intensively altered the global nitrogen (N) cycle of terrestrial ecosystems, increasing system productivity to meet human needs, but also bringing serious environmental and health problems. Therefore, studies on global N cycling under the disturbance of human activities are not only one of the current research focuses of ecology, but also one of the core contents of socioeconomic and environmental sustainability. With the strengthening of human activity disturbances, the structure and function of natural ecosystems have been breaking, forming cropland, urban, plantations, pasture, industry, mine and other functional systems, further self-organizing and upgrading to a coupled human and natural system (CHANS). China is experiencing rapid socioeconomic development and serious environmental N pollution currently, which is on the’hot spot’and’hot moment’of global N cycle in CHNAS. It provides an ideal case for better understanding the N cycling in CHANS, rational use of the positive role of N, and mitigating the negative impact on human and the Earth system.In this paper the horizontal direction boundary definition of CHANS follows China’s land borders. The upper boundary in the vertical direction is1km above ground surface, not including atmospheric circulation; the lower boundary is on the surface of bedrock, mineral resources not included. N cycling of CHANS starts from the entry of active N (Nr) that activated from N2into the system or Nr direct input to the system from outside-system, and terminate when Nr is oxidized/reduced to N2or directly output to outside-system. CHANS is divided into four functional groups:processors, consumers, remover and life-supporter. Processor can process the fixed N input into the food chain and biomass products, including cropland, grassland, forest, livestock, aquaculture, industrial, and urban greenland subsystems. Consumer includes humans and pets subsystems. Remover refers to a system that can treat the waste Nr and eliminate its negative impact, including wastewater treatment and garbage disposal subsystems. Life-supporter includes near-surface atmosphere, surface water and groundwater subsystem. All Nr imported from outside-system will access to one or several subsystems, recycling among subsystems within the system or output to the outside-system.There are two core questions of N cycle in CHANS:human N supply consumption and system sustainable development. To address these two questions, I complied the dataset of China’s N cycling, including over100thousand Nr fluxes and the related factors of air temperature, precipitation land use and socio-economy, etc. between1980-2008by using of observations, literatures, yearbooks and other sources based on mass balance approach, atmospheric remote sensing and geographic information system (GIS) techniques. On the basis of theories and hypotheses of CHANS and the dataset constructed, I comprehensively conducted the quantitative analysis of China’s N cycling in CHANS during the recent30years, and comparison and analysis on food N and industrial N between China and worldwide by using ecological models we built, e.g., NCNA, URCNC. The main results are as follow:1) China’s reactive Nr input increased from24.6to59.6Tg N yr-1from1980to2008with an increase fold of1.4, which is2times that of global average increase rate. With7%of world’s land area, China consumed30%of global anthropogenic Nr input, indicating China’s great contribution to the global N cycle. Compared to the natural Nr input to terrestrial ecosystems, human activities have increased3.6times of Nr input to terrestrial ecosystems in China; however, this value is only1-1.5on global level. During the recent30years, biological N fixation remained unchanged in China, which is different with the increasing trend globally; proportion of Haber-Bosch N fixation to total Nr input increased from50%to69%, higher than the global level. On the basis of consumption, Haber-Bosch Nr mainly distributed in agriculture developed regions, e.g., North China Plain, Northeast China, Yangtze River and Sichuan basin, and industry developed regions, e.g., Yangtze River Delta, Pearl River Delta; although NOX-N emissions from fossil fuel combustion increased3.7times, its N flux was small and the proportion to total Nr input maintained at4%-8%. Therefore, food and industrial N production-consumption are the main drivers of N cycling in China.2) In China, with the increase of Nr input and proportion of Nr from anthropogenic sources, system Nr output was gradually smaller than input. It means that CHANS tends to accumulate Nr along with the enhancement of human activities. This is opposite to previous findings, which suggest that Nr mainly loses through discharged to ocean or atmospheric circulation to nearby regions with the increase of anthropogenic Nr. The neglect of industrial N by previous studies might be the main reason for this opposite conclusion. During the recent30years, Nr accumulated in China increased3.1times to27.0Tg N yr-1in2008, Nr accumulation occurred mainly in the cropland (19.8%), forests (31.3%), grasslands (10.5%), groundwater (17.1%) and human settlement (19.7%). Meanwhile, the Nr concentration of surface water and atmosphere has being on the risen, also increasing the accumulation of Nr in the system.3) Nr input to China’s processer functional groups increased by1.3-fold, from59.4to137.8Tg N yr-1during the past30years. The industrial N products providing for human as well as plant and animal proteins increased by8.2,1.0and5.5times, but the per capita consumption is still lower than that of developed countries by a factor of20-25%. High Nr input intensity was significantly related to per capita GDP (PGDP), and annual temperature and precipitation, thus, Eastern and Southern China become the "hot spots" of high Nr input. With technological progress and policy innovation over the past30years, Nr loss rate of processer functional group was slower than its input rate in China which means the N use efficiency (NUE) of processer functional group rose.4) China’s per capita food N consumption reached5.2kg N yr-1in2008, lower than the Europe and United States (US)(6.5kg N yr-1), but higher than the global average (4.5kg N yr-1). However, China’s per capita consumption proportion of animal protein was33.2%, lower than the global average (38.7%), and the Europe and US (65.2%). Even though, the proportion of grain used as feed in China had been reduced to60%, and the production of animal protein become the main consumer of grain in China. Referencing this proportion of developed countries (>70%), China’s grain production would increase been used as feed. Considering the differences of human diets, the proportion of China’s animal protein consumption may not reach the level of Europe and US; however, the proportion of animal protein consumption still will increase50%in next20years. It thereby drives China’s Nr input to further increase.5) Industrial N is another important N production/consumption type in addition to food N, mainly referring to fiber, houses, furniture, etc. This is the first attempt to classify the industrial Nr into anthropogenic source (NA, such as synthetic fibers, rubber, etc.) and biological sources (NB, such as leather, cotton, etc.), and to estimate the industrial N flux in China and a global scale. In2008, China’s per capita consumption of industrial N is only3.4kg N yr-1, lower than the global average (4.3kg N yr-1), far lower than that in the Europe and US (>10kg N yr-1). Per capita NA consumption is significantly related to PGDP and urbanization, resulting in high NA flux occurred mainly in high urbanized and developed regions. NB consumption does not relate to socioeconomic development; therefore NB flux varies little across different regions. However, cultural differences can contribute to the variation of NB flux.6) This paper divides industrial N products into two types:structural and non-structural N. The proportion of structural N is over70%, leading to the accumulation of industrial N in human settlements. In2008, global industrial N accumulation is estimated at～21Tg N, which can explain81% of the global’missing anthropogenic N sink’(-26Tg N yr-1). Structural industrial N can be retained in human settlements for decades to centuries, until burned or decomposed. Delayed Nr release from industrial N would mitigate the environmental N pollution caused by rapid release; however, the delays also lead to difficulty in estimating the amount of waste produced. Policy-makers need to focus on the legacy effect introduced by the delay to avoid ’garbage Besieged City’phenomenon and achieve socioeconomic and environmental sustainability.7) In2008, the total Nr emitted to the atmosphere reached16.9Tg N yr-1, doubled the Nr input over the past30years, and China’s NOx emission growth rate was larger than that of NH3and N2O. N2O emission was1.1Tg N yr-1in2008, and its greenhouse effect is equivalent to0.2Pg C yr-1CO2. Atmospheric Nr dynamics estimated by mass balance approach above is consistent with the results retrieved from atmospheric remote sensing inversion (R2=0.80-0.94). Henan, Shandong and Hebei from the North China Plain, and Sichuan and Jiangsu from Yangtze River watershed were the main sources of NH3and N2O, while NOx emissions mainly concentrated in Eastern coastal areas. NH3and N2O emission were mainly driven by agricultural development, while NOx emissions were impacted by socioeconomic development.8) Agricultural nonpoint source pollution contributed～60%, and domestic and industrial point source pollution accounted for～30%of surface water N pollution over the recent30years. Surface water N pollution mainly occurred in North China Plain and the Northeast China and parts of the Northwest China. Although surface water Nr fluxes were large in South China, the sufficient surface water resource made the Nr pollution lightener than North China because of the dilution effect. Human consumption and other related activities are the major causes of the regional differences of surface water Nr pollution, and their contributions rank as follow: population> urbanization> PGDP. China’s groundwater was polluted heavily by Nr, especially in the North China Plain, Northeast China and the Yangtze River Delta region, the average groundwater N concentration exceeding20mg N L-1. High population density, PGDP and urbanization significantly affect the groundwater N concentration, which does not related to natural factors, revealing the dominant role of human activities on groundwater N pollution. Through the scenario analysis of future Nr dynamics in China by system modeling, I found that technologies play a key role in atmospheric Nr pollution control; policies mainly contribute to groundwater Nr pollution control, while technology and policy both work on surface water N mitigation within CHANS.更多还原