保障水生态系统健康、人-水和谐是实现水资源可持续利用、流域高质量发展的必要条件，然而密集的人类活动改变了全球多地水生态系统的水文水质条件和生物栖息地属性，对生态系统造成显著影响。精准的水生态修复技术是保障生态系统健康的有效手段，全面的生物多样性评价可以有效辅助进行水生态修复，进一步提高水生态修复的可靠性，目前关于水生态修复技术和生物多样性评价研究仍待加强，关键驱动因子在水华防治等保障水生态健康举措中研究不够全面，增加了水生态系统健康保护、实现人-水和谐的难度，给水资源安全保障提出了巨大的挑战。本研究以我国第一个水生态文明建设试点城市作为研究区，基于济南市9次水文-水质-水生态一体化大规模采样调查数据，首先构建了确定优先修复区和优先修复因子的水生态修复关键技术，以增加水生态修复工程的针对性；其次，提出了针对济南市的不同尺度生物多样性评价途径，以确定浮游植物物种沿各环境因子梯度的变异性；再次，构建了全面可行的关键驱动因子筛选方法，筛选了水生态修复和生物多样性驱动因子，有助于降低计算复杂度的同时提升水生态修复效果和生物多样性保护措施的可靠性；最后，将驱动因子筛选方法应用于蓝藻防治中，明确了蓝藻群落演变驱动因子的地域分布规律，有助于降低蓝藻水华爆发对水生态系统健康和生物多样性保护产生的威胁。

（1）构建了确定优先修复区和优先修复因子的水生态修复关键技术，包括物种适应性模型（IHSI模型）分析单一生物物种对多个驱动因子适应性；多物种响应度模型（MHSI模型）分析多个生物物种对单一驱动因子响应度；生物群落适应性模型（IHSIA模型）和驱动因子优先度模型（MHSIF模型）确定优先修复区和优先修复因子。研究发现：济南市浮游植物物种的适应性均较高，对环境表现出较强的适应性，小席藻适应性最强；济南市浮游植物物种响应梯度为：WT（水温）为20.9°C-28.4°C，pH为7.8-8.4，DO（溶解氧）为5.4-10.8mg/L，浮游植物对NH₃-N（氨氮）、WT、TP（总磷）等因子变化响应较敏感；济南市优先修复因子为NH₃-N，优先修复区不同季度优先修复因子为NH₃-N、TP。

（2）提出了针对济南市不同尺度生物多样性评价途径，包括构建β_z计算方法确定生物多样性沿各环境梯度变化；耦合α、β_z、γ多维评价生物多样性水平。研究得到：沿TP、NH₃-N浮游植物生物多样性β_z值最大，沿pH、COD（化学需氧量）浮游植物生物多样性β_z值相近，沿WT浮游植物生物多样性β_z值最小；浮游植物群落在α多样性上呈现空间一致性，均呈现山区低、平原高的规律，人口密集的城市地区α生物多样性小，枯水年汛期浮游植物α生物多样性变化幅度明显增大，区域营养物质浓度高低对α生物多样性作用显著；城中区沿环境梯度浮游植物物种丰度和多度的差异最大，南部山区沿环境梯度浮游植物物种丰度和多度的差异最小；研究区总体水生生物多样性水平较高，但呈降低趋势，浮游植物γ生物多样性季节差异显著，呈倒V型分布，夏季最高，秋季次之，春季最低。

（3）构建了可行的关键驱动因子筛选方法，包括通过优势度模型和突变点法确定优势物种；耦合定性分析的CCA方法（典型相关分析）和定量分析的VIP方法（变量重要性投影）综合筛选关键驱动因子。研究表明：济南市浮游植物优势物种为小席藻、小颤藻、肘状针杆藻、尖针杆藻、纤细新月藻、铜绿微囊藻、微小色球藻、细小平裂藻、中华尖头藻、尾裸藻、极分歧羽纹藻、钝顶螺旋藻、四尾栅藻、梅尼小环藻、卵形隐藻、小环藻、颗粒直链藻、银灰平裂藻、绿色裸藻、裸藻等；济南市水生态修复和生物多样性关键环境因子为WD（水深）、RW（河宽）、WT、pH、TP、NH₃-N、COD和DO，与俄罗斯、法国等世界多地研究结论一致。

（4）将关键驱动因子应用至蓝藻防治中，包括蓝藻群落优势物种筛选；蓝藻演变关键驱动因子筛选；关键驱动因子随机森林空间聚类。研究发现：济南市湖泊水库蓝藻优势物种为小席藻，小颤藻，铜绿微囊藻，细小平裂藻，中华尖头藻，银灰平裂藻；济南市湖泊水库蓝藻群落演化进程中COD、DO、TP、NH₃-N、WT、pH值作为蓝藻群落演化关键驱动因子，与蓝藻生长特性相符；在人口较少的山区，DO等驱动因子限制作用下降，蓝藻群落各点位驱动因子空间集聚性高，具有很强的地域分布规律。

本文构建的水生态修复关键技术有助于明确水生态修复的优先区域和环境因子，使修复措施更具针对性，并可以点带面，辐射带动整个研究区水环境改善，提出的不同尺度生物多样性评价途径可有效降低因计算尺度等因素给生物多样性评价带来的不确定性，同时，构建了较完善的关键驱动因子筛选方法，并应用至蓝藻防治中，有助于降低蓝藻水华爆发对水生态系统健康和生物多样性保护产生的威胁。

Ensuring the health of the aquatic ecosystems and the harmony between people and water are necessary conditions for the sustainable use of water resources and the high quality development of the river basin. Accurate water ecological restoration technology is an effective way to ensure the health of the ecosystem. Comprehensive biodiversity assessment can effectively assist water ecological restoration and further improve the reliability of water ecological restoration. At present, research on water ecological restoration technology and biodiversity evaluation still needs to be strengthened. The key driving factors have not been comprehensively studied in measures to protect water ecological health such as water bloom prevention, which increases the difficulty of water ecological system health protection and human-water harmony, and poses a huge challenge to water resources security. This research based on the data of 9 large-scale integrated sampling surveys of hydrology, water quality and water ecology in Jinan, and took the first pilot city for the construction of aquatic ecological civilization in our country as the research area. Firstly, a key technical method for water ecological restoration was established to determine priority restoration areas and priority restoration factors, and increased the pertinence of water ecological restoration projects. Secondly, a biodiversity evaluation system for different scales in Jinan was proposed to evaluate the variability of phytoplankton species along the gradient of various environmental factors. Thirdly, a comprehensive and feasible selection method for key driving factors was constructed to identify aquatic ecological restoration and biodiversity driving factors, which was conducive to reducing computational complexity while improving the effectiveness of aquatic ecological restoration and the reliability of biodiversity protection measures. Finally, the driving factor identification method was applied to the prevention and control of cyanobacteria, and the geographical distribution of driving factors for the evolution of cyanobacteria communities was clarified, which was conducive to reducing the threat of cyanobacteria blooms to the health of aquatic ecosystems and the protection of biodiversity.

(1) A key technical method for aquatic ecological restoration was established to determine priority restoration areas and priority restoration factors, including the IHSI model (species adaptability model) analyzed the adaptability of a single biological species to multiple driving factors; the MHSI model (multi-species responsiveness model) analyzed the responsiveness of multiple biological species to a single driving factor; the IHSIA model (biocommunity adaptability model) and the MHSIF model (driving factor priority model) determined priority restoration areas and priority restoration factors. The study found that the phytoplankton species in Jinan are all highly adaptable, showing strong adaptability to the environment, and Phormidium tenus has the strongest adaptability; the response gradient of phytoplankton species in Jinan: WT (water temperature) was 20.9°C-28.4°C, pH was 7.8-8.4, DO (dissolved oxygen) was 5.4-10.8 mg/L, and phytoplankton was more sensitive to the changes of NH₃-N (ammonia nitrogen), WT, TP (total phosphorus) and other factors; the priority restoration factor in Jinan was NH₃-N, and the priority restoration factors in different seasons in the priority restoration area were NH₃-N and TP.

(2) Proposed a biodiversity evaluation system for different scales in Jinan, including the construction of a calculation method for β_z to determine changes in biodiversity along various environmental gradients; the multi-dimensional evaluation of biodiversity level by coupling α, β_z, and γ. The results showed that the β_z value of phytoplankton biodiversity was the largest along TP and NH₃-N, the β_z of phytoplankton biodiversity was similar along pH and COD (chemical oxygen demand), and it was the smallest along WT; the phytoplankton communities showed spatial consistency in α diversity, with the pattern of low in mountainous areas and high in plain areas. The α biodiversity was small in urban areas with densely populated area, α biodiversity of phytoplankton changes significantly during dry seasons, and the concentration of regional nutrients has a significant effect on α biodiversity; the difference in phytoplankton species richness and abundance along the environmental gradient was greatest in the central area, and the difference in phytoplankton species richness and abundance along the environmental gradient was the smallest in the southern mountainous area; the overall level of aquatic biodiversity in the study area is relatively high with a decreasing trend. The γ biodiversity of phytoplankton has a significant seasonal difference, showed an inverted V-shaped distribution, and it is the highest in summer, following by autumn, and the lowest in spring.

(3) A comprehensive and feasible selection method for key driving factors was constructed, including the determination of dominant species through dominance model and mutation point method; the combination of the CCA method (canonical correlation analysis) of qualitative analysis and the VIP method of quantitative analysis (variable importance projection) to identify key driving factors comprehensively. The results showed that the dominant species of phytoplankton in Jinan were Phormidium tenus, Oscillatoria tenuis, Synedra ulna, Synedra acus, Closterium gracile, Microcystis aeruginosa, Chroococcus sp, Merismopedia tenuissima, Raphidiopsis sinensia, Euglena caudate, Pinnularia sp, Nostoc sp2, Scenedesmus quadricauda, Cyclotella meneghiniana, Cryptomonas ovate, Cyclotella sp, Melosira granulate, Merismopedia glauca, Euglena viridis, and Euglena; the key environmental factors for the restoration of water ecology and biodiversity in Jinan were WD (water depth), RW (river width), WT, pH, TP, NH₃-N, COD and DO, which were consistent with the research conclusions of Russia, France and many other places in the world.

(4) Apply key driving factors to the control of cyanobacteria, including the selection of dominant species in cyanobacteria community, identification of key driving factors for the evolution of cyanobacteria, and random forest spatial clustering of key driving factors. The results showed that the dominant species of cyanobacteria in lakes and reservoirs of Jinan were Merismopedia glauca, Merismopedia tenuissima, Microcystis aeruginosa, Oscillatoria tenuis, Phormidium tenus and Raphidiopsis sinensia; COD, DO, TP, NH₃-N, WT, and pH were the key driving factors for the evolution of the cyanobacteria community, which were consistent with the growth characteristics of the cyanobacteria during the evolution of the cyanobacteria community in the lakes and reservoirs of Jinan; the limiting effect of DO and other driving factors has declined in mountainous areas with a small population, and the driving factors at each point of the cyanobacteria community were highly spatially agglomerated and with a strong regional distribution pattern.

A key technical method for water ecological restoration constructed in this paper is conducive to clarifying the priority areas and environmental factors for water ecological restoration, which make the restoration measures more targeted, and can be used to improve the water environment of the entire study area. The biodiversity evaluation system of different scale proposed in this paper can effectively reduce the uncertainty caused by factors such as calculation scales in biodiversity evaluation. At the same time, this research has constructed a more complete identification system of key driving factors and applied it to the prevention and control of cyanobacteria, which is conducive to reducing the threat of cyanobacteria blooms to the health of aquatic ecosystems and the protection of biodiversity.