Simulating UK plant biodiversity under climate change to aid landscape decision making

景观由多种生境以及其中的生物多样性组成，是物种、气候、地理和人类利用之间相互作用的产物。它们提供许多生态系统服务，例如提供食物和水，调节气候和碳循环，这对我们的社会，经济，健康和福祉的稳定未来至关重要。植物是所有陆地生态系统的基础，是提供这些生态系统服务的基础。因此，景观决策应该得到能够预测植物对全球变化和景观管理的反应的工具的支持。然而，目前模拟植物物种分布的方法在这方面存在缺陷，因为它们侧重于单个或少数物种，忽略了物种之间的相互作用;或只模拟少量的植物功能类型。一个系统的方法将被用来解决目前现实世界的景观决策支持的这一重大差距，通过开发工具来预测（包括不确定性量化）目前和未来的分布，所有~1,800英国植物物种的方式，占物种之间的竞争相互作用。这将有助于有效评估景观决定和/或气候变化的影响，例如在特定地点或对泥炭地等重要生境类型的影响。入侵的非本地植物被认为是对生态系统服务的日益增长的威胁，通过扩展到全球约20万种植物物种，该工具还可以评估入侵的非本地植物物种对当前和未来英国景观的影响。害虫和疾病也是一个重大挑战，该项目开发的工具将成为管理植物健康景观的宝贵资源，例如，通过提供风险种群的分布-即任何感兴趣的疾病或害虫的植物宿主的分布。未来的工作可以探索潜在的关键反馈之间的动态的植物群落分布和传播的害虫和疾病的耦合models of these processes.This project built on a existing coarse spatial scale model for all plant biodiversity on Earth and an ongoing NERC funded project developing a higher resolution version for UK plant species.后一个项目利用了英国更详细的气候、土地利用和植物覆盖记录。然而，还需要进一步完善，以适当量化这一框架内的结构和过程的不确定性。如果没有这样的工作，从这些模型中出现的气候变化和土地利用决策的影响的预测可能是误导。目前生态位偏好参数化的观测数据没有不确定性评估。就结构不确定性而言，通过确定每个物种如何生长和繁殖（其功能类型）来更好地解释物种间异质性至关重要。因此，基于自然历史博物馆现有的数字化专业知识，我们建议从现有的分类描述中提取相关的功能类型信息，为所有英国本土和非本土植物物种创建一个更广泛的性状数据库。这项工作本身就是一种宝贵的资源，可扩展到所有全球植物记录，它将在项目中用于增强模拟模型，以捕捉物种之间在生长、竞争和扩散方面的相对差异。与基于有限数量功能类型的当前模型进行比较，将突出结构复杂性的作用和非线性对模型输出的影响。我们还将开发工具，利用现有的植物物种分布数据量化这些模型中的不确定性，以便我们能够正确地捕捉计划和预期的土地利用和气候变化的影响，并最终指导未来的景观决策。

Landscapes are composed of multiple habitats as well as the biodiversity that resides within them, and are a product of interactions between species present, climate, geography and human use. They provide many ecosystem services, such as provision of food and water, regulation of climate and carbon cycling, which are vital for a stable future for our society, economy, health and wellbeing. Plants form the basis of all terrestrial ecosystems and are fundamental to providing these ecosystem services. Landscape decisions should therefore be underpinned by tools that enable prediction of plant responses to global change and landscape management. However, current approaches to modelling plant species distributions are deficient for this purpose as they focus on individual, or a small number of, species; ignore interactions between species; or only model a small number of plant functional types.A systems approach will be used to address this significant gap in current real-world landscape decision support by developing tools to predict (including uncertainty quantification) current and future distribution of all ~1,800 UK plant species in a manner that accounts for competitive interactions between species. This will enable effective assessment of the impacts of landscape decisions and/or climate change, e.g. in specific locations or on important habitat types such as peatlands. Invasive non-natives are considered a growing threat to ecosystem services and through extension to ~200,000 plant species worldwide this tool also enables assessment of the impact of invasive non-native plant species on current and potential future UK landscapes. Pests and diseases also represent a significant challenge and tools developed by this project will be a valuable resource for managing landscapes for plant health, for example, by providing distributions of at-risk populations - i.e. the distribution of plant hosts for any disease or pest of interest. Future work could explore the potentially critical feedbacks between the dynamics of plant community distributions and the transmission of pests and diseases by coupling models of these processes.This project builds on an existing coarse spatial scale model for all plant biodiversity on Earth and an ongoing NERC-funded project developing a higher resolution version for UK plant species. The latter project makes use of the more detailed climate, land use and plant coverage records available for the UK. However, further refinements are needed to properly quantify structural and process uncertainty within this framework. Without such work predictions of the effect of climate change and land use decisions that emerge from these models could be misleading.Currently niche preferences are parameterised by observational data with no uncertainty assessment. In terms of structural uncertainty, it is critical to account for between-species heterogeneity better by establishing how each species grows and reproduces (its functional type). Building on existing digitisation expertise at the Natural History Museum we therefore propose to extract relevant functional type information from existing taxonomic descriptions to create a more extensive trait database for all UK native and non-native plant species. As well as being a valuable resource in its own right and extensible to all global plant records, this work will be used within the project to enhance the simulation model to capture the relative differences in growth, competition and dispersal between species. Comparison with the current model based on a limited number of functional types will highlight the role of structural complexity and the impact of non-linearities on model output. We will also develop tools to quantify uncertainty in these models using available plant species distribution data so that we can correctly capture the impact of planned and expected land use and climate change, and ultimately guide future landscape decision making.