生物入侵己成为当前国际社会公认的重要生态环境问题，是本世纪人类所面临的巨大挑战和生态威胁之一。在全球变化背景下，越来越多的生物被有意或者无意地引入或转移到其原产地以外的区域，并呈现爆发式的扩增，对当地的生态环境和经济社会产生了一系列的负面影响。全球滨海湿地是遭受外来植物入侵最为严重的生态系统之一，为有效地缓解或消除外来植物入侵对本地生态系统产生的负面影响，深入探讨入侵地中外来植物的入侵和扩散机制已成为入侵生态学研究的重要内容和热点话题。滨海盐沼生态系统地貌景观因子所主导的地貌过程是众多物理（如水文要素波动等）、化学（如生物地球化学循环等）和生物过程（如植物种群建立、营养级联等）的载体和基础，对维持盐沼结构与功能的稳定具有重要的意义。潮沟系统作为滨海盐沼生态系统中最显著、最基本、最活跃的地貌景观要素/单元，影响和控制着盐沼与外部环境的水文连通过程以及以水为媒介的物质交换、能量流动和信息传递，其在发挥重要连通功能的同时，也可能成为外来入侵植物向盐沼内部扩散的“帮凶”。然而，当前关于盐沼中外来植物入侵扩散机制的研究却往往忽视了潮沟系统这一重要地貌景观要素/单元带来的潜在影响，以至于难以精准地预测外来植物的入侵扩散格局和生态影响。因此，从盐沼地貌景观要素与外来植物入侵关联的角度，系统解析滨海湿地潮沟系统影响下外来植物的入侵扩散格局和多重机制，对管控外来植物入侵和维护盐沼生态安全具有十分重要的意义。

作为国际重要湿地，黄河口滨海湿地近年来也受到严重的外来植物互花米草（Spartina alterniflora）入侵扩散的威胁，其不仅占据了盐沼低、中潮滩区域，并沿着潮沟系统向盐沼内部不断扩大其入侵范围。因此，为了更好地预测互花米草在黄河口滨海盐沼的入侵扩散格局以及认识典型地貌景观要素对外来植物入侵扩散的影响机制，本研究以黄河口典型潮沟系统和入侵植物互花米草为研究对象，围绕“滨海湿地地貌景观要素/特征如何影响外来植物入侵”这一关键核心问题，按照“辨识格局-揭示机理-调控管理”的整体思路，综合运用遥感解译与空间分析、野外调查与长期监测、原位与室内控制实验、大数据整合分析等多元研究方法，在景观尺度上，定量解析了潮沟系统结构和功能属性特征与互花米草入侵格局的配置关系；在局地尺度上，基于繁殖体压力、入侵“机会窗口”、生境可入侵性、生态系统抵抗力、入侵植物“生态陷阱”等相关生态学理论或假说，系统阐明了潮沟系统中关键非生物和生物过程对互花米草入侵扩散关键过程的多重影响机制，并将其归纳整理成“一张图”；同时，基于全球海量文献的数据挖掘，定量评估了全球米草入侵对滨海湿地动物的生态影响以及不同移除方案对本地动物的恢复效果。最后，基于上述研究结果，提出了黄河口滨海湿地互花米草入侵防控的综合管理策略。其主要结果如下：

（1）潮沟系统属性特征与互花米草入侵格局的时空演变及其配置关系

       就潮沟系统的功能属性特征而言，2013年至2020年间，整个研究区的潮沟长度和面积在不断的增加，8年间分别增加了79.4%和34%，表明潮沟系统处于不断的发育状态。潮沟曲率和网络复杂度在整体上变化不大，而排水密度和运行效率呈现明显的增加趋势，分别增加了66%和85%，其潮水平均归槽流路长度也呈现显著的下降趋势，这表明潮沟系统的排水效率及其对潮滩的服务功能在不断地增加。就互花米草的入侵过程而言，研究区内互花米草面积呈递增趋势，8年间增加了近5.4倍。在互花米草的初期入侵阶段（2013-2015年），其主要以外部离散式扩张为主，其次为边缘式扩张；在扩张中期阶段（2015-2017年），互花米草开始以潮沟驱动式扩张为主，外部隔离式和边缘式扩张所占比例下降；在互花米草扩张的后期阶段（2017-2020年），潮沟驱动式扩张依然占据主要地位，外部离散式扩张持续下降，而边缘式扩张模式有所回升。此外，互花米草入侵面积与潮沟长度、潮沟曲率呈显著正相关，而与网络复杂度无明显的相关关系，表明潮沟的结构属性对互花米草的陆向入侵具有显著的促进作用，互花米草在长度较长、结构弯曲的潮沟系统中入侵潜力更大。互花米草入侵面积与潮沟排水密度和运行效率均呈显著正相关，与潮水平均归槽长度呈显著负相关，表明潮沟系统的功能属性特征对互花米草的陆向入侵具有显著的促进作用，即互花米草在排水效率和运行效率较高的潮沟集水区具有较强的入侵潜力。

（2）潮沟系统关键非生物要素对互花米草入侵扩散的多重影响机制

       潮沟系统的水动力扰动是影响互花米草初期幼苗定植的关键制约因子。相对于主级潮沟边缘带而言，末级潮沟边缘带更容易遭受互花米草的有性繁殖体入侵，其主要通过以下两个重要的机制和过程：（i）末级潮沟边缘带中较高的繁殖体压力与低盐度和高含水率的环境条件相结合，为互花米草的幼苗定植奠定了基础；（ii）末级潮沟边缘带中持续低强度的水动力扰动环境为定植互花米草幼苗的存活和生长提供了“机会窗口”。潮沟系统的弯曲地貌结构对互花米草入侵具有“庇护所”效应，特别是凸侧潮沟结构是互花米草入侵过程中的“垫脚石”，为互花米草沿潮沟系统向陆入侵提供了空间窗口，其主要通过以下两个重要的机制和过程：（i）凸侧边缘带中相对较低的流速能够促进潮水携带的互花米草种子在此处搁浅，从而增加了繁殖体压力；相反地，凹侧边缘带较高的水动力扰动环境不利于繁殖体的搁浅，从而在源头上抑制了互花米草的成功入侵；（ii）凸侧边缘带中持续低强度的水动力扰动为互花米草幼苗的定植和生长提供了“机会窗口”，而凹侧高水动力扰动产生的强底床剪切力，进一步降低了互花米草幼苗定植的成功性。

潮沟介导的水盐-繁殖体压力梯度是潮沟边缘带中互花米草横向扩张的主要驱动力，其在潮沟纵向和横向距离梯度上变异规律决定了互花米草在不同盐沼区位潮沟边缘带中横向扩张的边界。在潮沟纵向高程梯度上，位于盐沼低、中潮滩的潮沟边缘带面临较高的繁殖体压力和较弱的水盐胁迫，导致互花米草在该区域的扩张强度较大；而在潮沟横向距离梯度上，位于盐沼中、高潮滩的潮沟边缘带中，距离潮沟较远的生境面临较小繁殖体压力和较强的水盐胁迫，导致互花米草在该区域入侵强度较低。总体而言，较高的繁殖压力与潮沟边缘带中的空间入侵窗口（即具有适宜环境条件的微生境）相结合，为互花米草的定植、生长、有性繁殖和无性扩张奠定了坚实的基础，从而在很大程度上促进了互花米草在潮沟边缘带的横向扩张。从更广义的角度，一旦生态系统中存在空间和时间上的“机会窗口”，比如地貌景观特征创造的空间上的适宜物理环境和外界胁迫波动形成的时间上的无/低扰动期，便为外来植物入侵种群的建立创造了先决条件，而在机会窗口期间增加繁殖体压力往往可以增强外来植物的入侵强度。

（3）潮沟系统关键生物过程对互花米草入侵扩散的多重影响机制

      本地蟹类（Helice tientsinensis）的植食作用显著增强了遭受人为干扰的盐沼生态系统对互花米草入侵的抵抗力。野外实验表明，盐沼高潮滩人为潮沟挖掘活动和本地食草蟹类的耦合作用对互花米草初期幼苗定植的影响过程呈阶段式。首先，人为挖掘干扰引发的上行促进作用有利于互花米草幼苗的定植，其主要通过两个独立的机制来实现：（i）捕获或截留大量的互花米草种子来增强繁殖压力；（ii）降低或改变生态系统非生物抵抗（即高盐和干燥胁迫）的限制，为互花米草种子萌发、幼苗初期定植和生长提供适宜的环境条件。其次，本地蟹类对定植互花米草幼苗的植食作用调节了互花米草的最终入侵格局。如果蟹类植食强度较高，则无法成功入侵；反之，如果蟹类植食强度较低，则入侵成功。

     除了蟹类植食外，来自盐沼邻近陆地高地的哺乳类食草野兔（Lepus capensis）能够大量取食互花米草的茎秆、叶片、分蘖苗和种子，增强了本地盐沼对互花米草入侵的抵抗力。在时间尺度上，随着时间的推移食草野兔对互花米草的植食强度不断增强；在空间尺度上，其觅食范围也逐渐从盐沼高潮滩潮沟边缘带拓展到互花米草分布面积更大的中潮滩。由于互花米草的营养价值与本地植物相差无几，且适口性较高（互花米草茎秆和种子中具有较高的可溶性总糖、水分、粗纤维和粗蛋白）、生长周期较长，能为本地食草野兔越冬提供稳定多样和高质量的能量供应，因而已成为其主要食源。此外，本地食草野兔在入侵生境中获益的同时，也可能陷入了互花米草构筑的“生态陷阱”。因为互花米草入侵生境中也吸引了大量捕食者黄鼬（Mustela sibirica）前来觅食，从而增加了其被捕食的风险，带来潜在的适合度损失，也在一定程度上减弱了食草野兔对互花米草入侵扩散的抑制作用。本地食草野兔从“生态陷阱”中获益还是损失，取决于其如何在“陷阱吸引力”和“适应成本”之间进行权衡，而这种权衡也最终决定了本地盐沼对互花米草入侵的生物抵抗力。

     潮沟边缘中母代互花米草产生的生物地貌反馈塑造了潮沟边缘带的凹凸地貌（即扩张带内形成凸地，在扩张带边缘形成凹地），从而对后代互花米草的定植和扩张产生影响，且这种代际间的种内作用具有阈值依赖性。其具体机制为：（i）由于边缘带凹地靠近种子源区，具有较高的种子可获得性和到达率，增加了繁殖体压力；（ii）边缘带凹地通过滞留潮水对子代互花米草产生阈值性影响，即当母代互花米草塑造的凹地深度在阈值之内时，其能够促进后代实生苗的定植、生长和无性扩张；反之超过阈值时，则发生淹水胁迫对后代幼苗产生抑制作用。此外，这种生物地貌反馈所塑造的凹地对本地物种盐地碱蓬产生了显著的负面效应，最终在潮沟边缘带形成本地物种盐地碱蓬的生长范围/生存空间不断被“挤压”，而外来入侵物种互花米草不断地向盐沼内部横向扩张的分布格局。

（4）全球米草属植物入侵及其移除对滨海湿地动物的生态影响

      基于文献数据挖掘的整合荟萃分析表明：在全球尺度上，米草属植物入侵显著降低了滨海湿地动物群落生物多样性（如物种丰富度等），这可能是由入侵植物形成的同质化栖息地而导致。然而，本地动物群落的种群丰度（如个体数量、密度等）和适合度（如生物量等）在一定程度上呈增加趋势，这可能是由于入侵米草的“定向生态过滤效应”对某些底栖动物产生了促进作用而导致。上述生态影响具有情境依赖性，其影响模式因栖息地类型、动物分类群、营养级和摄食类型而异。尽管不同移除措施对本地动物的恢复效果存在差异性，但总体上有利于滨海湿地动物群落的恢复。值得提出的是，长期水淹对本地动物群落的恢复具有较大的负面影响，因而在具体工程实践时要慎重考虑；而本地植物群落的重建显著提升了本地动物群落的适合度，对于生态系统功能的恢复至关重要。综上所述，米草属植物入侵作为强扰动重组了滨海湿地动物群落，只有通过合理的人工措施将其清除，本地动物群落才可能被部分恢复。更重要地，入侵植物的根除并非入侵地生态恢复的终点，在经济投入和维护管理等方面制定切实可行的解决方案来实现其结构和功能的恢复，对退化湿地的保护和修复至关重要。

（5）黄河口滨海湿地互花米草入侵扩散的综合防控策略

      为了将上述对过程和机理的探讨更好的服务于滨海湿地中互花米草入侵的防控管理，本研究基于上述归纳的潮沟系统下互花米草入侵扩散机制和全球米草属植物入侵及其移除对本地动物生态影响的整合分析结果，从适应性管理的视角提出了“加强关键节点生态监测，重点防范优先入侵区域；严格限制人为挖掘活动，维持盐沼湿地自然地貌；合理实施水文连通调节，关闭盐沼湿地入侵窗口；充分利用本地动物植食，提升盐沼湿地生物抵抗；关注生物地貌反馈效应，规避地貌上行不利影响；权衡植物移除措施利弊，加强盐沼湿地功能恢复”的黄河口滨海湿地互花米草入防控综合管理策略，从而为当地相关调控方案的制定和管理策略的实施提供借鉴和参考。

      综上所述，本研究不仅从地貌景观特征/要素与外来植物入侵关联的角度丰富了相关的入侵生态学理论，还为黄河口滨海湿地的生态保护和外来物种入侵防控提供了数据支撑和理论指导，并为全球河口滨海湿地的保护、修复和管理提供了借鉴和经验。

Biological invasion has long been recognized as an important ecological and environmental problem and it is one of the great challenges and ecological threats faced by human beings in this century. In the context of global changes, more and more organisms are introduced or transferred to areas outside their origin, both intentionally and unintentionally, and present explosive expansion, which has produced a series of negative impacts on the local ecological environment, economy and society. Coastal wetlands are one of the mostseverely impacted ecosystems by exotic plant invasions worldwide. In order to effectively mitigate or eliminate the negative effects of exotic plant invasion on local ecosystems, it has become an important content and hot topic to deeply explore the spread mechanisms of invasive plants in their invaded regions. The geomorphic process dominated by the geomorphic landscape factors of the coastal salt marshes is the carrier and basis of many physical (e.g., hydrological element fluctuations, etc.), chemical (e.g., biogeochemical cycles, etc.) and biological processes (e.g., plant establishment, tropic cascades, etc.), which is of great significance for maintaining the stability of ecosystem structure and functioning. The tidal channel system, which is the most significant, basic, and active geomorphic landscape element/unit in the coastal salt marshes, affect and control the hydrological process between the salt marsh and the external environment, as well asthe water-based material exchange, energy flow and information transmission. It not only playing an important connectivity function, but also may become an “accomplice” of invasive plants spreading into the salt marshes. However, current researches on the invasive and spreading mechanisms of alien plants in salt marshes often ignore the potential impacts of the tidal channel systems, an important geomorphic landscape element/unit, so that it is difficult to accurately predict the invasion patterns and ecological impacts induced by exotic plants. Therefore, from the perspective of the association between the geomorphic landscape elements/features and alien plant invasions, the objective of this thesis is to systematically analyze the patterns and multiple mechanisms of alien plant invasion and spread under the influence of coastal tidal channel systems. The results will be of great significance for controlling alien plant invasion and maintaining the ecological security of salt marshes.

The Yellow River Delta, which is one of important wetlands in the world has been threatened by the invasion and spread of the exotic plant (Spartina alterniflora, hereafter S. alterniflora) in recent years. S. alterniflora not only occupies the low and middle salt marshes, but also expands its invasion-ranges landward along the tidal channels into upper zones of salt marshes. Therefore, in this thesis, I used tidal channel systems and invasive S. alterniflora as the study system, in order to better predict the spread pattern of invasive S. alterniflora in the Yellow River Estuary. The main question of this thesis is “how does coastal wetland landscape element/feature affect exotic plant invasion?”. Using multiple research methods such as remote sensing analysis, field survey & long-term monitoring, in-situ & indoor control experiments, and big data meta-analysis, in accordance with the overall research idea of “identifying pattern - revealing mechanism - regulating management”, this thesis has finished the following works. On the landscape scale, the configuration relationship between the attributes of tidal channel systems (both structural and functional) and the invasion pattern of S. alterniflora was quantitatively analyzed. On the local scale, based on the relevant ecological theories/hypotheses, such as propagule pressure, invasive “windows of opportunity”, habitat invasibility, ecosystem resistance, and “ecological traps” formed by invasive plants, the multiple mechanisms of key abiotic and biotic factors that affecting the key ecological processes of S. alterniflora invasion and spread in tidal channel systems were systematically elucidated, and were also summarized into a "synthesizing map". Additionally, the ecological impacts of global Spartina species invasions on coastal faunal communities and the outcomes of Spartina species removals on faunal community recovery were also quantitatively evaluated, based on the big data meta-analysis extracting from global peer-reviewed literature. Finally, based on our research results, a comprehensive management strategy for the control of invasive S. alterniflora in the coastal wetlands of the Yellow River Estuary was proposed. The main results of this thesis are concluded as follows:

 (1) Spatial and temporal evolution of the attibutes of tidal channel systems and invasion pattern of the exotic S. alterniflora, and their configuration relationships

In terms of the functional attributes of the tidal channel systems, from 2013 to 2020, the length and area of tidal channels in the entire study area have increased by 79.4% and 34% respectively, indicating that the tidal channels are in a state of continuous development. The sinuosity and network complexity of the tidal channels have not changed much on the whole, while the drainage density and efficiency have shown an obvious upward trend, increasing by 66% and 85% respectively. The average overmarsh path length (OPL) also showed a significant downward trend. This shows that the drainage efficiency of the tidal channel systems and their service function to tidal flats are constantly increasing. With regard to invasion process of S. alterniflora, the area of S. alterniflora within our study area has increased 5.4 times over the past 8 years. During the initial invasion stage (2013-2015), S. alterniflora mainly expanded by the way of external isolated expansion, followed by marginal expansion. In the mid-stage of invasion (2015- 2017), the expansion of S. alterniflora was mainly driven by tidal channel systems, while the proportion of external isolated expansion and marginal expansion decreased. In the later stage of S. alterniflora expansion (2017-2020), tidal channels-driving expansion still dominated, external isolated expansion continues to decline, while the way of marginal expansion picked up.

Importantly, the invasion area of S. alterniflora was positively correlated with the tidal channel length and sinuosity, but has no obvious correlation with the network complexity. This indicates that some structural attributes of tidal channel systems play an important role in facilitating S. alterniflora landward invasion. For promoting effect, S. alterniflora has greater potential to invade in the tidal channel systems with longer length and high sinuosity. The invasion area of S. alterniflora was significantly positively correlated with the drainage density and efficiency of tidal channels, and also negatively correlated with the average overmarsh path length (OPL), indicating that the functional attributes of tidal channels have a significant facilitation to the landward invasion of S. alterniflora. In other words, S. alterniflora has strong invasion potential into tidal channel systems with higher drainage efficiency.

(2) Multiple influence mechanisms of key abiotic processes in the tidal channel systems on the invasion and spread of S. alterniflora

The hydrodynamic disturbance of tidal channel systems is the key factor affecting the initial seedling establishment of S. alterniflora. The margins of secondary tributaries were more susceptible to the invasion of S. alterniflora than the margins of main tidal channels due to two important mechanisms and processes: (i) higher propagule pressure combined with suitable salinity-moisture conditions of soils (i.e., low salinity and high moisture) laid a firm foundation favoring the seedling establishment of S. alterniflora, and (ii) low-intensity hydrodynamic disturbance provided windows of opportunity for the seedling survival and growth of S. alterniflora. Additionally, the meandering geomorphic structures of the tidal channels could generate “shelter effects” on S. alterniflora invasion, especially the convex-side margins of tidal channels play as a “stepping stone” during the plant invasion process, thus providing a spatial window for the landward invasion of S. alterniflora along with the tidal channel margins. This is mainly through the following two vital mechanisms and processes: (i) the relatively low flow velocity of the convex margins allowed more S. alterniflora seeds carried by the currents/tides to be stranded there, thereby increasing the propagule pressure. In contrast, the higher hydrodynamic disturbance of the concave margins was not beneficial to the stranding of propagules, thus inhibiting the invasion success of S. alterniflora at the source;(2) the continuous low hydrodynamic disturbance in the convex margins provided windows of opportunity for the establishment, colonization and growth of S. alterniflora seedlings, while the strong bed shear stress caused by the high hydrodynamic disturbance in the concave margins further reduced the success of the seedlings establishment.

The gradient of physical stresses and propagule pressure mediated by tidal channels is the main driver of the lateral range-expansion of S. alterniflora in the tidal channel margins, and their variation in the longitudinal and lateral distance gradients determines the expansion boundary of S. alterniflora in the tidal channel margins of different elevational marsh zones. In the longitudinal elevation gradient, the tidal channel margins at low and middle marshes faced higher propagule pressure and lower physical stresses (i.e., salinity and inundation), resulting in a higher intensity of range-expansion within this region. However, in the lateral distance gradient along tidal channel margins, the microhabitats far away from the tidal channels in the middle and high marshes, faced lower propagule pressure and higher physical stresses, which led to the low intensity of Spartina range-expansion in this region. In summary, higher propagule pressure integrated with the suitable physical conditions of the microhabitats that tidal channels structurally formed lays a firm foundation benefiting the establishment, growth, colonization and sexual reproduction of S. alterniflora, which substantially contributes to the rapid range-expansion of S. alterniflora along the tidal channel margins. In a broad sense, once a geomorphic landscape feature creates spatial invasion windows (e.g., the spatially suitable physical environments created by the landscape features and the temporal no/low disturbance period formed by external stress fluctuations) that allow the initial invasive populations to be established successfully, the community structure and ecosystem functioning can in evitably be damaged by the invasive plant gradually, and even completely destroyed when the invasion reaches a certain threshold. Importantly, increasing propagule pressure during the invasion windows can often enhance the intensity of plant invasion.

(3) Multiple influence mechanisms of key biotic processes in the tidal channel systems on the invasion and spread of S. alterniflora

The native herbivorious crab (Helice tientsinensis) significantly enhances the biotic resistance of anthropogenically disturbed salt marshes to S. alterniflora invasion. Our field experiments showed that the invasion processes of S. alterniflora to high marshes triggered by anthropogenic ditching disturbances appeared stage-by-stage. First, the bottom-up facilitation triggered by anthropogenic ditching disturbance favored seedling establishment of S. alterniflora, which is mainly achieved through the two independent mechanisms: (i) trapping a large number of S. alterniflora seeds to enhance propagule presseure, and (ii) reducing the limit of the ecosystem abiotic resistance (i.e., high salinity and drought stress) to provide suitable environmental conditions for seed germination, seedling establishment and plant growth of S. alterniflora. Second, the top-down inhibition resulting from native herbivores regulated the final pattern of S. alterniflora invasion. If the herbivorous intensity of the crabs was high, S. alterniflora could not successfully invade. Conversely, if the intensity of crab herbivores was low, the invasion could be successful.

In addition to the native crab herbivory, native mammalian herbivorous hare (Lepus capensis) from the neighboring terrestrial uplands could also enhance the biotic resistance of salt marshes to S. alterniflora invasion, as many plant tissues/organs of invasive S. alterniflora such as stems, leaves, ramets and seeds were heavily consumed by the native hares. On the temporal scale, the herbivorous intensity of native hares on invasive S. alterniflora has significantly increased over the time. On the spatial scale, the foraging range of native herbivorious hares has gradually expanded from the tidal channel margins of high marshes to the middle marshes, in which S. alterniflora has a larger distribution area there. Nowadays, S. alterniflora has become the main food source for the native herbivorous hares, because of its similar nutritional value to native plants, high palatability (i.e., the stems and seeds of S. alterniflora have higher total soluble sugar, moisture content, crude fiber and protein) and long growth cycle, which could provide stable, diverse and high-quality energy supply for the overwintering of native hares. Other than the native hares, surprisingly, it also has been observed that their predators (weasels, Mustela sibirica) frequently haunt around invasive Spartina meadows. Therefore, despite that invasive S. alterniflora could be of beneficial for the native hares by providing abundant food resources, the invasive plants yet increase the risk of the attracted hares being preyed by weasels, and in this way, act as an ecological trap for the native mammalian grazers. Whether the native mammals benefit or lose from ecological traps relies on how they make a trade-off between the two contrasting characteristics of the traps formed by invasive S. alterniflora (i.e., trap attractiveness vs. fitness cost), and this trade-off ultimately determines the biotic resistance of native salt marshes to S. alterniflora invasion.

The bio-geomorphic feedbacks generated by the maternal S. alterniflora shaped the concaveconvex geomorphology of the tidal channel margins (i.e., convex geomorphology was formed within the S. alterniflora expansion belt, and concave geomorphology was formed at the edge of plant expansion belt), and thus influenced the establishment and expansion of the offspring of S. alterniflora. The above-mentioned intergeneration intraspecific interaction was thresholddependent, and the specific mechanisms are as follows: (i) as the marginal depressions were close to the seed source, the high availability and arrival of S. alterniflora seeds increased the propagule pressure; (ii) the marginal depressions showed a threshold effect on the offspring of S. alterniflora through retaining the tidal waters. Specifically, when the depth of depressions shaped by the maternal S. alterniflora was within the threshold, it could facilitate the establishment, growth and asexual expansion of the offspring seedlings. In contrast, when the depth exceeded the threshold, hyper inundation stress would inhibit the offspring seedlings. Furthermore, the marginal depressions also showed significant negative effect on the native dominant species Suaeda salsa. Eventually, the growth range/living space of Suaeda salsa would be continuously "squeezed", while the invasive S. alterniflora continued to laterally expand its range into salt marshes.

(4) The ecological impacts of global Spartina species invasions and their removals on coastal faunal communities

Our meta-analysis revealed that Spartina invasions have generated differential displacement on native faunal communities at a global scale. The overall diversity (e.g., species richness) of native faunal communities was significant decreased, which may due to the homogeneous habitats that invasive Spartina formed. However, the overall abundance (e.g., individual number) and fitness (e.g., biomass) of native faunal communities were increased to some extent, this could be explained by the directional ecological filtering effects that invasive plants may facilitate specific species of benthic animals. Moreover, the pattern of invasion impacts is context-dependent, depending on habitat types, faunal taxa, trophic levels, and feeding types. Furthermore, the results of our meta-analysis demonstrated invasive Spartina removals were overall beneficial for the recovery of native faunal communities, even though the responses were varied among different control measures. Importantly, the implementation of long-term waterlogging should not be encouraged for controlling invasive Spartina species as this measure has strong negative impacts on the recovery of native faunal communities. Overall, these findings indicate that Spartina invasions could act as a press perturbation that reorganizes the coastal faunal communities, and the native faunal communities could be partly recovered, only if the invaders have been removed with artificial control measures. Thus, invasive plant eradication is not the endpoint of ecological restorations. Developing practical solutions that are logistically and economically feasible for the structural and functional recovery of invaded ecosystems will be of great significance for future ecosystem conservation, restoration and management.

(5) Comprehensive management strategy of invasive S. alterniflora in coastal wetlands of the Yellow River Estuary

Based on the above-summarized comprehensive mechanisms of S. alterniflora landward invasion along the tidal channel systems and the results of global meta-analysis, this thesis proposed the following strategies from the perspective of adaptive management: (i) strengthening the ecological monitoring and focusing on the priority areas of S. alterniflora invasion; (ii) strictly restricting anthropogenic ditching activities to maintain the natural geomorphology of salt marshes; (iii) reasonably implementing hydrological connectivity regulations to close the invasion windows of salt marshes; (iv) making full use of native herbivores to enhance the biotic resistance of salt marshes; (v) paying attention to the bio-geomorphic feedbacks induced by invasive plants to avoid the negative bottom-up impacts; (vi) balancing the pros and cons of invasive plant removal measures and strengthening the functional restoration of salt marshes. These strategies could provide references for the formulation and implementation of relevant local invasive plant control programmes.

In summary, this thesis not only enriches the relevant ecological theories from the perspective of the relationship between geomorphic landscape features/elements and exotic plant invasions, but also provides data support and theoretical guidance for the ecological conservation of the coastal wetlands of the Yellow River Estuary and the scientific control of exotic plant invasion. Additionally, it also provides references and insights for the conservation, recovery, restoration and management of global coastal wetlands.