磷是限制湖泊中藻类及其他生物生长的一种关键性营养元素。近几十年来，化石燃料与生物质燃烧、汽车尾气排放以及化肥施用等人类活动改变了磷的全球循环，导致湖泊磷负荷显著增加。湖泊磷负荷增加可能加速湖泊富营养化，改变湖泊浮游植物群落组成和生物多样性。因此为了解富营养化的成因和危害，并对湖泊富营养化进行预防、控制和管理，国内外学者对富营养化湖泊磷生物地球化学进行了比较深入的研究。与富营养化湖泊相比，目前对贫营养化湖泊中磷生物地球化学关注相对不足。高原湖泊通常属于贫营养化湖泊，一般受人类活动的影响较轻，磷的主要来源为自然源一般为大气沉降和地表径流输入。近年来的研究表明，高原贫营养化湖泊对磷等营养物质的变化反应敏感，磷入湖通量的微小变化可能会引起其生态效应的较大改变。因此，探究高原贫营养化湖泊中磷的生物地球化学过程具有重要的实际意义。

      本研究以我国最大的内陆高原堰塞咸水湖泊青海湖为主要研究区域，采集了青海湖水、悬浮颗粒物、沉积物、沉积物孔隙水样品，布哈河水、悬浮颗粒物样品，入湖河流下游表层土壤、土壤剖面样品，大气干湿沉降样品；分析样品基本理化性质、磷浓度和形态、保守示踪元素浓度，揭示了青海湖多介质磷分布特征。开展了沉积物吸附磷动力学和热力学实验，估算了沉积物和土壤磷吸附解吸平衡浓度（EPC₀）。采用磷与保守元素比值方法，解析了大气湿沉降磷来源；采用气团轨迹反演方法，阐明了大气干沉降磷的来源。结合青海湖输入输出分析，全面揭示了青海湖磷生物地球化学行为与过程。本研究的主要结果和结论如下：

      青海湖湖水中溶解态磷与颗粒态磷浓度（7.69 μg L^-1）接近，而河水中磷以颗粒态（44.98 μg L^-1）为主。青海湖水、布哈河水以及青海湖沉积物孔隙水中溶解态磷平均浓度分别为6.89 μg L^-1，5.17 μg L^-1 和171.69 μg L^-1。布哈河支流河水中悬浮颗粒态磷浓度基本低于干流河水。布哈河口附近湖区溶解态磷浓度较低，而悬浮颗粒态磷浓度较高。青海湖南岸以及海心山附近沉积物孔隙水中溶解态磷浓度相对较低。青海湖不同深度湖水中溶解态磷浓度间不存在显著差异。河水中P/Cl 摩尔比值高于湖水，一方面表明河流输入的溶解态磷存在输出途径；另一方面，从历史演化的过程并不排除磷在湖水中的富集。

      沉积物中总磷含量范围为172.59-605.23 mg kg^-1，剔除离群值后，平均浓度和中位值浓度分别为490.43 和477.81 mg kg^-1。表层土壤总磷含量范围为457.69-786.04 mg kg^-1，平均浓度和中位值浓度分别为608.31 和605.20 mg kg^-1。沉积物中碳酸盐含量是土壤的2 倍多，高达33.75%。碳酸盐沉淀累积干沉积物中导致的固体稀释效应是沉积物磷及主要元素K、Na、Al、Fe、Mn 含量低于土壤的主要原因。沉积物中磷以碳酸盐结合态和有机质结合态为主，分别占总量的56.99%和42.36%。土壤经风蚀和水蚀输入湖泊后有机结合态磷向碳酸盐结合态磷转化。湖中部区域沉积物磷含量通常较高，而南岸湖区沉积物磷含量较低。表层土壤磷含量空间差异小。土壤中磷含量、有机磷含量沿土壤剖面随深度增加逐渐下降，与人类活动排放的磷逐渐增加和植物富集磷有关。

       沉积物和土壤吸附磷的动力学过程表现为表面扩散的快速吸附阶段（0-8 h）和孔隙扩散的慢速吸附阶段（8-36 h），且以快速阶段为主。沉积物和土壤吸附磷的热力学过程符合Langmui 等温吸附模型，沉积物和土壤吸附磷的最大量为465.37-679.51 mg kg^-1。沉积物磷吸附解吸平衡浓度（EPC₀）（0.011-0.019 mg L^-1）低于土壤（0.043-0.073 mg L^-1），且两者均高于湖水中磷浓度。因此，沉积物以及土壤通过风蚀水蚀途径输入湖泊后均存在解吸释放磷的倾向。低盐度下的阳离子架桥作用促进了沉积物和土壤吸附磷，而高盐度下的硫酸根等阴离子的竞争吸附降低了沉积物和土壤吸附磷。

       青海湖大气湿沉降中溶解态氮体积加权平均浓度为3.33 mg L^-1；其中NH₄⁺-N、NO₃^--N、WSON 分别占59.16%、16.51%和23.12%。具有NH₄⁺-N 浓度雨季高于旱季，而NO₃^--N 浓度雨季低于旱季的特点。雨季较高的气温和农牧业活动导致的氨挥发和旱季气溶胶空中滞留时间长导致的NO₃^--N 浓缩是其季节浓度变化的主要原因。青海湖大气湿沉降中WSIP 和WSOP 体积加权平均浓度分别为0.30 mg L^-1 和0.07 mg L^-1，从春季到秋季呈现出逐渐上升的趋势。大气干沉降中水溶态氮和非水溶态氮质量加权平均浓度分别为1.88 mg g^-1 和1.82 mg g^-1，而水溶态磷和非水溶态磷质量加权平均浓度分别为0.044 mg g^-1 和0.756 mg g^-1。干沉降水溶态NH₄⁺-N、NO₃^--N 和WSON 质量加权平均浓度分别占干沉降总浓度的23.17%、16.43%和9.70%。干沉降各形态氮和WSOP 质量加权月均浓度为夏秋季高于冬春季，而WSIP 质量加权月均浓度为冬春季高于夏秋；其原因可能与各物质的来源以及气团来源有关。

       青海湖水溶态氮和非水溶态氮大气干沉降通量分别为3.92 kg ha^-1 yr^-1 和4.06 kg ha^-1 yr^-1，水溶态磷和非水溶态磷大气干沉降通量分别为0.095 kg ha^-1 yr^-1 和1.623 kg ha^-1 yr^-1。表明大部分大气干沉降氮可被生物利用，而极小部分大气干沉降磷可被生物利用。主要受降水量和降尘量影响，4 至8 月各形态氮磷湿沉降通量逐渐升高，到9 月有所降低；各形态氮磷干沉降通量3、4 月最高。青海湖大气湿沉降中氮磷主要来源于人为活动，其中农业生产、畜牧业、燃油及燃煤是湿沉降中氮的主要来源，煤及生物质燃烧是湿沉降中磷的主要来源。干沉降氮磷主要来源于内蒙古西部和甘肃北部的土壤。青海湖大气沉降水溶态氮磷摩尔比值为23.54；其中湿沉降比值为20.01，干沉降比值为86.80。该比值均高于湖泊生物营养限制因子氮磷比值（16:1），因此，目前青海湖大气沉降水溶态N:P 摩尔比值可维持青海湖浮游植物生长为P 限制型。

      青海湖磷输入通量为1.58 kt yr^-1。其中，河流输入通量仅占4%，而且3%以悬浮颗粒态形态输入湖泊。磷大气湿沉降通量为0.791 kt yr^-1，占其输入通量的50%；磷大气干沉降通量为0.691 kt yr^-1，占其输入通量的46%。因此，大气沉降是青海湖磷的主要来源，大气降水为青海湖浮游植物提供了绝大部分的磷。青海湖磷沉积通量为0.927 kt yr^-1，远低于磷输入通量，输入通量与沉积物通量的比值为1.70。导致磷输入通量与沉积通量差异的因素包括：（1）鱼等水产品捕捞输出部分磷；（2）鸟类体内及鸟粪蓄积磷的输出；（3）湖泊浪花溅入大气中的磷；（4）磷在湖水中积累。从湖泊长期演化视角，受大气沉降尤其是大气湿沉降的驱动，青海湖水中溶解态磷浓度存在升高的潜势。青海湖磷沉积物通量高于颗粒态磷输入通量，表明输入湖泊的部分溶解态磷通过吸附、与碳酸盐共沉淀、生物残体等途径积累于沉积物中。

       钙镁沉积物通量远高于输入通量，表明全球大气CO₂ 浓度升高加速青海湖水中钙镁沉淀，并积累于沉积物中。其对青海湖磷生物地球化学过程的影响有待深入研究。

   Phosphorus (P) is a critical nutrient that limits the productivity of lake ecosystems. In recent decades, human activities, such as fossil fuel combustion, vehicle exhaust emission, and application of fertilizer, have altered the global cycles of P, which has led to an increased loadings of P to lake ecosystems. Long-term and high-level P loadings may cause eutrophication, which would further decrease the biodiversity of the aquatic ecosystem. Therefore, in order to understand the causes and hazards of eutrophication and to prevent, control and manage it, scientists around the world have widely studied the geochemical behaviors and processes of P in lake ecosystems. However, compared with the studies on P in eutrophic lakes, the behaviors and processes of P in oligotrophic lakes has not been paid enough attention. Alpine lakes are typically nutrient poor and generally less affected by human activities. These lakes receive proportionally a larger fraction of P from the atmosphere than lowland lakes. Recent studies have shown that, alphine lake ecosystems respond rapidly to P alterations, particularly in alpine oligotrophic lakes where a small change in absolute concentration can mean a large change in relative availability. Therefore, the behaviors and biogeochemical processes of P in alpine oligotrophic lakes are critical to protect them.

    Lake Qinghai is China’s largest inland closed-basin saline lake, with a surface elevation of 3194 m. It is characterized with typical alpine oligotrophic lakes. The objectives of this thesis were to study the distributions and geochemical behaviors and processes of P in the multimedia of Lake Qinghai. We collected water, suspended particle, sediment, and sediment porewater samples of the lake, water and suspended particle samples of Buha River, topsoil and soil core samples around the lake, and atmospheric wet and dry deposition samples. Physiochemical properties, P concentrations and chemical forms, conservative tracer elements were determined for these samples. Adsorption dynamic and isotherm of P on sediment and soil were conducted. Conservative tracer method and backward airmass trajectories method were used to identify the souces of P in atmospheric wet and dry depsotions, respectively. P input and output fluxes of Lake Qinghai were estimanted. Major results and conclusions were obtained as follow:

    The concentrations of water-soluble P were close to those of particulate P (7.69 μg L^-1) in the lake water, while particulate P was the main form of P in the river water (44.98 μg L^-1). The average concentrations of water-soluble P in the lake, river and pore water were 6.89, 5.17 and 171.69 μg L^-1, respectively. The concentrations of particulate P in the tributary river were basically lower than those in the main river. The concentrations of dissolved P were lower while the concentrations of particulate P were higher in the lake area near the mouth of Buha River. The concentrations of dissolved P in the pore water were relatively low on the south bank of Qinghai Lake and near Haixin Mountain. There were no significant differences in dissolved P concentrations in different depths of Lake Qinghai. The P/Cl molar ratios in river water were higher than those in lake water. On the one hand, it shows that there is an output route of dissolved P from the river. On the other hand, the process of historical evolution does not exclude the accumulation of P in the lake water. The total phosphorus (TP) concentrations in the sediments ranged from 172.59 to 605.23 mg kg^-1. After removing the outliers, the average and median concentrations of P in the sediment of Lake Qinghai were 490.43 and 477.81 mg kg^-1. The TP concentrations in the topsoils ranged from 457.69 to 786.04 mg kg^-1, the average and median concentrations of P in the topsoils of Lake Qinghai were 608.31 and 605.20 mg kg^-1. The carbonate concentrations in the sediments were more than twice those of the topsoils, as high as 33.75%. The solid dilution effect caused by carbonate deposition in sediments was the main reason that the P, K, Na, Al, Fe and Mn in sediments lower than those in soils. Ca-P and organic P were the main forms of P (accounted for 54.76% and 44.14% of TP) in the sediments; organic P and Ca-P were the main forms of P (accounted for 56.99% and 42.36% of TP) in the topsoils. The organic P changed to carbonate P after the soil was transported into the lake by wind or water weosion. The P concentrations in the sediments in the middle of the lake were higher, while the P concentrations in the south lake were lower. The spatial differences of P concentrations in topsoils were small. The concentrations of P in soil profiles decreased with the increase of soil depth, which was related to the gradual increase of P emitted by human activities and the ebrichment of P by plants.

The dynamic processes of P adsorption in sediments and soils were characterized by the rapid adsorption stage of surface diffusion (0-8 h) and the slow adsorption stage of pore diffusion (8-36h). The kinetic curves and equilibrium isotherms of phosphorus sorption onto the sediments and soils of the Lake Qinghai could be described well by the Langmuir model. The maximum amounts of P adsorbed by sediments and soils were ranged from 465.37 to 679.51 mg kg^-1. The equilibrium phosphorus concentrations (EPC₀) for the sediments were lower than those for the soils, and the value ranges were 0.011-0.019 mg L^-1 and 0.043-0.073 mg L^-1, respectively. The EPC₀ for the sediments and soils were all significantly high than the P concentrations in the overlying water, indicating the sediments and soils in Lake Qinghai held a trend of releasing P. Salinity is a main factor influencing P sorption behaviors on sediments and soils of Lake Qinghai. Under low salinity, the cationic bridging can adsorb P in sediments and soils, while under high salinity, the competitive adsorption of anions such as sulfate can reduce the adsorption of P in sediments and soils.

    The annual volume-weighted mean (VWM) concentration of water-soluble total nitrogen (WSTN) in the atmospheric wet deposition was 3.33 mg L^-1. NH₄⁺-N, NO₃^--N and WSON accounted for 59.16%, 16.51% and 23.12% of WSTN, respectively. The monthly VWM concentration of NH₄⁺-N was higher in the wet season and low in the dry season, while it was low in the wet season and high in the dry season for NO₃^--N. These seasonal or monthly changes in NH₄⁺-N and NO₃^--N concentrations were due to ammonia volatilization caused by high temperature and agricultural/animal husbandry activities in the wet season and nitrate nitrogen accumulation due to long air residence of aerosol in dry season. The annual VWM concentrations of water-soluble inorganic phosphorus (WSIP) and water-soluble organic phosphorus (WSOP) in the atmospheric wet deposition were 0.30 and 0.07 mg L^-1, respectively. From spring to autumn, there was a gradual upward trend. The annual mass weighted mean concentrations of water-soluble phosphorus and particulate phosphorus in the atmospheric dry deposition were 0.044 and 0.756 mg g^-1, respectively. NH₄⁺-N, NO₃^--N and WSON accounted for 23.17%, 16.43% and 9.70% of WSTN, respectively. The mass weighted monthly average concentration of nitrogen and WSOP in atmospheric dry deposition was higher in summer and autumn than in winter and spring, and WSIP was higher in winter and spring than in summer and autumn. The reason may be the source of the matter and the air mass.

    The atmospheric dry deposition fluxes of water-soluble N and particulate N were 3.92 and 4.06 kg ha^-1 yr^-1, the atmospheric dry deposition fluxes of water-soluble P and particulate P were 0.095 and 1.623 kg ha^-1 yr^-1. It shows that most of the atmospheric dry deposition nitrogen can be bioutilized, and a very small part of the atmospheric dry deposition P can be bioutilized. Mainly affected by precipitation and dust fall, the wet deposition fluxes of nitrogen and phosphorus in various forms increased gradually from April to August, and decreased to some extent in September. The nitrogen and phosphorus dry deposition fluxes of all forms were highest in March and April. Nitrogen and phosphorus in the atmospheric wet deposition in Tiebujia site of Lake Qinghai mainly come from human activities. Agricultural production, animal husbandry, fuel oil and coal burning were the main sources of nitrogen in the wet deposition, while coal and biomass combustion were the main sources of phosphorus in the wet deposition. Dry deposition nitrogen and phosphorus mainly come from soils in western Inner Mongolia and northern Gansu. The N:P ratios in annual total (dry + wet), wet and dry atmospheric deposition flux in this study were 23.54, 20.01 and 86.80, respectively. This result indicated that the aquatic ecosystems in Lake Qinghai, fed principally by atmospheric nutrients, may tend toward P-limitation.