Data from: Developing state and transition models of floodplain vegetation dynamics as a tool for conservation decision-making: a case study of the Macquarie Marshes Ramsar wetland

1. Floodplain vegetation states (communities) exhibit spatiotemporal dynamics in vegetation structure and composition, which reflect unique hydrological and connectivity patterns. Shifts in inundation regimes can drive succession and establish new stable states, determined by the magnitude and duration of the hydrological perturbation. 2. We aimed to develop a modelling approach that is able to capture ecosystem dynamics, identify and quantify the main drivers of change, and provide a tool for conservation decision-making. We developed state and transition models for floodplain vegetation states based on surveys in 1991 and 2008 in the Macquarie Marshes (Australia), a Ramsar wetland of international importance. We used a Bayesian logistic regression approach to model state and transitions between vegetation states and investigated how flood frequency, distance to stream and fire frequency were associated with vegetation dynamics during this period. 3. During 1991–2008, significant transitions have occurred towards drier states. Semi-permanent wetland vegetation had the lowest persistence probability (ppsis = 0·456) and a significant threshold response of transitioning to terrestrial vegetation (ptran = 0·505). Transition to drier states was driven by lower inundation probabilities followed by increased fire probability, and distance to nearest stream. 4. Using developed models, we predicted persistence probabilities of vegetation states under an unregulated (i.e. no dams or diversions) and regulated water availability system. Under a regulated system, semi-permanent wetland vegetation had an average persistence of ppsis = 0. 67 and 0·08 in the northern and southern sections of the nature reserve, respectively. Under an unregulated system, the predicted persistence of semi-permanent wetland vegetation was considerably higher: ppsis = 0·87 and 0·38, respectively. 5. Synthesis and applications. Developing quantitative models of state transitions significantly improved our understanding of ecosystem dynamics, identifying sensitive indicators for monitoring and thus supporting conservation decision-making. This helps managers understand potential trajectories of change in ecosystems in response to management options. For example, increasing environmental flows in the Macquarie Marshes is predicted to shift the community towards more of a wetland than the terrestrial state, resulting from river regulation. State and transition models identified how key ecological assets respond to drivers of change, particularly where these can be managed. This is critical for ensuring that all ecosystem components are managed and that these do not shift into undesirable states.1. 洪泛平原植被状态（群落）的植被结构与组成呈现出时空动态特征，该特征反映了独特的水文格局与连通性模式。淹水节律的变化可驱动植被演替并形成新的稳定状态，其取决于水文扰动的强度与持续时长。 2. 本研究旨在构建一种能够捕捉生态系统动态、识别并量化主要变化驱动因子，并为保护决策提供工具的建模方法。基于1991年与2008年对澳大利亚麦夸里沼泽（Macquarie Marshes）——一处具有国际重要性的拉姆萨尔湿地（Ramsar wetland）——的野外调查，本研究构建了洪泛平原植被状态的状态与转换模型。研究采用贝叶斯逻辑回归（Bayesian logistic regression）方法对植被状态及其间的转换进行建模，并探究了该时段内洪水频率、距溪流距离以及火灾频率与植被动态的关联。 3. 1991年至2008年间，植被群落发生了向更干旱状态的显著转变。半永久性湿地植被的存续概率最低（ppsis=0.456），且向陆生植被转换时呈现出显著的阈值响应（ptran=0.505）。向干旱状态的转换主要由更低的淹水概率驱动，其次是火灾概率升高以及距最近溪流的距离增加。 4. 基于构建的模型，本研究预测了无调控（即无大坝或引水工程）与有调控水资源供给系统下的植被状态存续概率。在有调控系统下，自然保护区南北片区的半永久性湿地植被平均存续概率分别为ppsis=0.67与0.08；而在无调控系统下，半永久性湿地植被的预测存续概率显著更高，分别为ppsis=0.87与0.38。 5. 总结与应用。构建植被状态转换的定量模型显著提升了我们对生态系统动态的认知，同时识别出用于监测的敏感指示因子，进而为保护决策提供支撑。该模型可帮助管理者了解生态系统在响应不同管理方案时的潜在变化轨迹。例如，麦夸里沼泽的生态流量（environmental flows）提升措施，在河流调控的背景下，预计将使植被群落向更偏向湿地的状态而非陆生状态转变。状态与转换模型明确了关键生态资产对变化驱动因子的响应机制，尤其是在这些驱动因子可被人为调控的场景中。这对于确保所有生态系统组分得到合理管理、避免其向不良状态转变至关重要。