Bridging the gap from individual-based processes to joint species distributions in a forest landscape

植物物种在一个地区的分布是自然和人为干扰、环境条件、每个物种的繁殖和扩散能力以及同一物种或不同物种的个体之间相互作用(竞争、促进)的累积结果。联合物种分布模型是用来描述物种分布如何依赖于环境以及物种范围如何在环境变化下发生变化的主要工具之一。由于这些模型在时间上是静态的，它们的解释假设物种范围处于某种稳定状态，这可能会限制它们预测群落将如何对突然变化做出反应的能力。以个体为基础的模型，如森林模拟器，更直接地描述了影响不同物种个体的人口、扩散和相互作用过程。其中一些模型还纳入了景观尺度上的干扰(例如火灾)的影响。然而，将这些模型参数化需要具有多个采样时间的大量数据集。因此，有必要采用新的方法，在联合物种分布模型的框架内纳入物种改变其活动范围的动态过程。在接下来的五年里，我将执行三个项目，它们都有共同的目标，即将扩散过程纳入多种物种分布的模型，但侧重于不同规模的森林(Obj.1)到复杂的森林景观(Obj.2)到更大的区域(Obj.3)。1.为长期研究的新热带森林中种子的传播和幼苗的建立建立多物种模型。2.模拟魁北克西部北部针叶林中枯木苔藓植物群落的动态。3.基于社区科学观测重建水生入侵物种沿道路和水文网络的扩散。特别是目标2和目标3的实现将为加拿大北方针叶林的管理提供有用的信息。苔藓植物关键类群的扩散能力(Obj.2)将部分确定保留在景观中的老生长林的数量和安排，以促进这些生物及其相关生态系统服务在收获后的回归。入侵物种传播的社区级模型(Obj.3)将有助于根据与已有入侵物种在特征上的相似性来预测新入侵物种的传播。

The distribution of plant species across a region is the cumulative result of natural and anthropogenic disturbances, environmental conditions, the reproduction and dispersal capacities of each species as well as interactions (competition, facilitation) between individuals of the same or different species. Joint species distribution models are one of the main tools used to describe how the distribution of species depend on the environment and how species ranges may shift under environmental change. Since these models are static in time, their interpretation assumes a certain steady state for species ranges, which may limit their capacity to predict how the community would respond to sudden change. Individual-based models, such as forest simulators, describe more directly the demographic, dispersal and interaction processes affecting the individuals of different species. Some of those models also incorporate the effects of disturbances (e.g. fire) at the landscape scale. However, parametrizing these models requires extensive datasets with multiple sampling times. There is thus a need for new methods to incorporate the dynamic processes through which species change their range within the framework of joint species distribution models. Within the next five years, I will pursue three projects that all share the goal of including dispersal processes into models for the distribution of multiple species, but focus on different scales, from a forest (obj. 1) to a complex forest landscape (obj. 2) to a larger region (obj. 3). 1. Develop multi-species models for the dispersal of seeds and the establishment of seedlings in a long-studied Neotropical forest. 2. Model the dynamics of communities of deadwood-inhabiting bryophytes across the boreal forest of western Québec. 3. Reconstruct the dispersal of aquatic invasive species along road and hydrological networks based on community science observations. The realization of objectives 2 and 3 in particular will provide useful information for the management of the Canadian boreal forest. The dispersal capacity of key groups of bryophytes (obj. 2) will determine in part the amount and arrangement of old growth forests to retain in a landscape to promote the return of these organisms and their associated ecosystem services following a harvest. Community-level models of invasive species spread (obj. 3) will assist in predicting the spread of new invasive species, based on similarity in traits to those already here.