河口作为流域最后的生态安全屏障，生态环境敏感脆弱，对于流域健康和水环境安全具有重要的地位和指示作用，已成为流域水生态恢复的瓶颈和焦点。本研究以生态毒理学为基础，应用暴露-响应分析、绘制风险等级图等微观与宏观相结合的方法，通过辨析河口水生态风险源、胁迫因子、GIS生境分析、筛选响应指标、分析暴露-响应路径，构建了风险响应网络模型。以强人为干扰和复合环境污染下的海河流域河口为研究案例，以实验室和野外原位生态监测和模拟数据为基础，确定了响应路径及等级系数，对海河流域河口水生态风险进行了诊断、评价和不确定性分析，揭示了河口水生态风险响应机制和区域风险空间分异规律。同时，提出了流域水生态风险变化过程情景模式，并以单一污染物重金属为例，对流域风险空间分异进行验证分析，为河口生境恢复及流域风险管理部门和决策者提供科学依据。主要研究内容和结论包括：（1）基于河口入海水量减少、水质恶化和生态系统退化交互影响的复杂性，提出了河口水生态风险是指在区域尺度上，综合考虑水质、水量和水生态三个方面，描述和评估环境污染、人为活动或自然灾害等引起的水环境变化对河口生态系统及其组分产生不利作用的可能性和大小的过程。建立了河口水生态风险评价程序：主要包括问题诊断、风险分析、风险表征、不确定性分析和风险管理等5个部分。其中，风险分析阶段主要包括概念模型建立和暴露-危害分析。概念模型用于揭示风险源、胁迫因子、生境及评价终点之间的路径关系，暴露-危害分析主要从河口水质、水量和水生态三个方面展开，定量化风险评价组分之间的暴露-响应系数。风险表征即评估危害作用的大小以及发生概率的过程，是风险评价的核心部分。同时，建立了适用于流域、子流域和河口等不同尺度下水生态风险评价模型与方法；提出了分别对应于自然流域、干扰较弱流域和强人为干扰下3种（累积式、开放式和混合式）流域水生态风险变化过程情景假设。（2）基于现有水资源分区和流域生态水文分区方式，通过辨识风险源、胁迫因子以及生境和终点间的关系，构建了河口风险响应网络模型（RNM）；同时，考虑区域风险源产生胁迫因子可能性的大小，增加了“风险源暴露”系数和“终点-生境”响应系数以减少不确定性。以海河流域河口为案例，选择水质恶化、生态用水量不足、生物多样性减少、生态系统服务价值受损和景观破碎作为生态风险评价终点，构建了反应河口水生态风险源-胁迫因子-生境-终点路径关系的概念模型，针对不同风险源类型、不同胁迫因子、不同生境、不同终点、不同风险小区，应用建立的RNM模型计算了风险值，对水生态风险进行了综合评价。研究结果表明：海河流域河口区域可分为两个低风险区、一个中风险区和两个高风险区，其中最高风险区位于海河河口（天津）区域。城镇化、船舶运输和工业为主要风险源，风险值分别为5.65，4.71 和 3.68；生境破坏、耗氧有机污染物和病原体是主要的胁迫因子；对于生境，开放水域（9.59）和毗邻陆地正面临更大的压力；生态系统服务价值受损（30.54）和生物多样性减少面临最大的风险压力。评价结果通过了蒙特-卡罗模拟法的不确定性检验，证实了该模型对基于流域生态水文分区，并考虑风险源、胁迫因子、生境和评价终点之间的复杂关系的河口水生态风险评价有较好的适用性。评价结果可以表征不同风险小区间相对风险关系，适用于基于风险源、胁迫因子、生境和评价终点之间暴露-响应关系的实验数据和资料相对贫乏的河口生态风险评价。（3）选取典型无机污染物重金属为例，通过布点采样，分析了海河流域河流、湖泊和河口沉积物中重金属分布特征；应用Hakanson 潜在风险指数法，分析了As、Hg、Cr、Cd、Pb、Cu 和 Zn 七种重金属的浓度范围和潜在生态风险；同时与国内外其他水体重金属最大浓度进行了比较分析。结果表明，Hg、Cd 和 Zn 的最大浓度表现出很高的水平，尤其在漳卫南河点位Hg（1135.50 mg/kg）和Cd（195765.83 mg/kg）均出现极高浓度。通过 值分析，Hg和Cd表现出极高的生态风险，7种重金属风险由高到低依次为：Cd＞Hg＞As＞Cr＞ Pb＞Cu ＞Zn。基于RI值，滦河和白洋淀处于中等风险水平，而漳卫南河和河口所有点位均处于高风险水平，表明河口区域单一污染物风险已达高风险水平。评价的典型生态单元生态风险从高到低依次为：漳卫南河 （2278345.68）＞河口 （161914.74）＞滦河 （2278345.68）＞白洋淀（120.95）。（4）根据研究区特点、单一污染物风险与区域风险评价结果，以及结合区域水生态系统健康和风险状况，提出海河流域、滦河流域和海河河口的生态风险安全阈值及其对应的水生态系统健康状态；对海河流域河口区域风险、滦河流域和海河流域总体风险进行了对比分析。结果表明海河河口各小区的相对风险均高于滦河流域各个风险小区的风险；此外，海河流域水环境受强人为干扰，生态风险变化属于混合型变化过程，无论河口是否存在风险，流域内部都可能存在风险；而海河流域的子流域中，滦河流域可认为是自然型流域，属于累积式变化过程；漳卫南河流域内水质、水量和水生态受各种自然和人为胁迫因子综合作用，承受着多种生态风险，属于混合型风险变化过程。最后，基于生态风险变化的式，提出了流域综合生态风险诊断与管理思路和建议。

As a sensitive indicator of the ecological health at the end of basin, estuary is the last barrier of whole watershed. Estuary ecosystem is a giant complex system, a large number of substances, energy and information exchange are accompanied by each of its dynamic changes. However, most estuaries such as Haihe River estuary are facing the water pollution, habitat degradation, coastal erosion, estuarine siltation, and other environmental problems under the strong human activities, which resulting in increasing of water environment risk. Thus estuary risk had become the focus and hot topic of the science and water managers in relevant fields. This paper considered the interactions of many kinds of sources, stressors, habitats and endpoints comprehensively based on the characteristics of estuary. In this study, ecological risk assessment model for estuary water environment was established, and Haihe River Basin Estuary was taken as examples for application research. The main research content includes:(1) Concept of ecological risk assessment for estuary water environment which coupling the water quality, water quantity and aquatic ecosystems was proposed based on the summary of risk assessment theories and methods at home and abroad, also based on the research progress of risk assessment for water environment. Procedure of ecological risk assessment for estuary water environment was developed which included five steps: problem diagnosis, risk analysis, risk characterization, uncertainty analysis and risk management. Among them, the conceptual model in the risk analysis was used to reveal the exposure relations between sources and receptors. The exposure and hazard analysis mainly focused on the water quality, water quantity and aquatic ecosystems. Risk characterization was to evaluate the degree and probability of the hazard effects which was the core of the risk assessment. In the phase of uncertainty analysis, the Monte-Carlo simulation method was adopted in the assessment procedure. At the same time, model and method of risk assessment under different scales (basin, sub-basins and estuary) were proposed. Fuzzy synthetic evaluation method for ecological risk assessment under basin scale, relative risk model method for ecological risk assessment under sub-basin scale, and response network model method for estuary were proposed. Based on the changes process of ecological risk of water environment in watershed, three kinds of scenario assumptions were proposed, which revealed the evolution of water environment in river basin.(2) Based on the relationship of sources, stressor, habitat and endpoint in the region, the response network model was proposed for assessing ecological risk of estuary water environment. In order to decrease the uncertainty in assessment, the probability that sources release stressor was considered through introducing the source-stressor-habitat exposure filter (SSH), the endpoint-habitat exposure filter (EH) and the stressor-endpoint effect filter (SE) to reflect the meaning of exposure and effect more explicitly. Water environment that include water quality, water quantity and aquatic ecosystems was selected as the assessment endpoints. We created a conceptual model which depicting potential affects pathways from source to stressor to habitat to endpoint. Water quality deterioration, eco-water demand shortage, biodiversity reduced, ecological service value damaged and landscape fragmentation were selected as the ecological risk assessment endpoints. Response network model was applied in the ecological risk assessment in Haihe River Basin Estuary. The results showed that there were two low risk regions, one medium risk region and two high risk regions in the Haihe River Basin Estuary. The results also indicated urbanization was the biggest source, the second was shipping and the third was industry, their risk scores are 5.65, 4.71 and 3.68 respectively. Furthermore, habitat destruction was the largest stressor with the risk scores (2.66), the second was oxygen consuming organic pollutants (1.75) and the third was pathogens (1.75). Therefore, these three stressors were the main influencing factors of the pressure in the study area. For habitats, open waters (9.59) and intertidal mudflat were enduring the bigger pressure and should be taken considerable attention. Ecological service values damaged (30.54) and biodiversity decreased were facing the biggest risk pressure.(3) Distribution and characteristics of seven heavy metals in sediment in the rivers, lakes, and estuaries of Haihe River Basin were analyzed through collecting the samples. The approaches of Hakanson potential ecological risk index were used for assessing ecological risk and the concentration scale of As, Hg, Cr, Cd, Pb, Cu and Zn in the eco-units were detected. At the same time, the maxim concentration of heavy metals was compared with other water bodies at home and abroad. The results indicated that the maxim concentration of Hg, Cd and Zn showed a higher level than other water bodies in the world from literatures. Among them, Hg and Cd showed the very high concentration at the site of Zhangweinan River (1135.50 and 195765.83 mg/kg) and Haihe estuary (790.50 and 548.47 mg/kg). According to , Cd and Hg showed very strong ecological risk, and the seven heavy metals of Cd, Hg, As, Cr, Pb, Cu, and Zn at ecological risk levels were in the descending order. Based on RI, Luanhe River and Baiyangdian Lake were in the moderate risk, and every site in Zhangweinan River and Haihe Estuary were in a very high risk levels. That indicated that the ecological risk of single pollutant in estuary region was reached into a relative high level. The sequence of risk from high to low of the representative eco-units was as follows: Zhangweinan River (2278345.68) >estuary (161914.74) >Luanhe River (2278345.68) > Baiyangdian Lake (120.95). (4) Based on the results of ecological risk of single and regional risk, combined with the characteristics of the study area, ecological security threshold and corresponding ecosystem status were proposed. For Luanhe River sub-basin, water quality, water quantity and aquatic ecosystems suffered the relative weaker stress, which can be considered as the cumulative risk change process. However, water quality, water quantity and aquatic ecosystems in Haihe River Basin and Zhangweinan River suffered the relative strong human interference, which can be considered as mix risk change process. Furthermore, risk of the Haihe River estuary was higher than the risk of the Luan River Basin by comparing the regional risk level. As the last ecological barrier at the end of basin, estuary should be given risk management in priority.