MSB-FRA Modeling Invasion Dynamics Across Scales (MIDAS)

外来物种的入侵对几乎所有生态系统都构成了重大威胁。 事实上，大约40%的美国森林已经被外来植物入侵，造成了巨大的经济损失（每年超过1000亿美元）和生态破坏。 不幸的是，目前对大规模入侵模式和分布过程的了解仍然有限。 这个项目，跨尺度入侵动力学建模（MIDAS），旨在了解潜在的生物，地球物理和社会经济因素如何导致大规模入侵模式的出现。 该项目的研究结果将有助于向森林管理者、政策制定者和其他利益相关者提供信息，说明如何通过研讨会、网络研讨会和在线规划工具等多种途径更好地预防和减轻入侵物种对经济和生态的破坏。 该项目将通过研究机会，指导和研讨会对不同群体的学生和研究人员的教育产生重大影响。 它将促进广泛、持久的跨学科和跨机构合作。 此外，该项目将有助于更好地理解“事物”（例如，疾病、技术）传播，从而有助于维持利益的传播，同时限制因社会联系而对生态系统和社会造成的意外伤害。该项目的目标是确定美国大陆森林中两种重要入侵者（植物和害虫）的宏观入侵模式的潜在机制。 该项目将利用一种新的交叉建模方法，结合：（1）基于机器学习的统计模型，旨在捕获受不同空间和时间分辨率挑战的数据之间的关键尺度内和跨尺度相互作用和临界点，以及（2）旨在捕获模型和参数不确定性的贝叶斯模型。 该项目将导致更好的机械理解和更强大的宏观入侵预测。 提出的概念框架，可以通过制定新的入侵理论和巩固各种现有的入侵生态学领域。 该项目还将通过合成和开发多学科数据集以及开发和完善一种非常有用的分析工具来帮助推进宏观系统生物学领域，该工具用于涉及许多变量的调查，其中操纵实验将具有挑战性。

Invasions by exotic species pose a major threat to nearly all ecosystems. In fact, about 40 percent of U.S. forests have been invaded by exotic plants, resulting in significant economic loss (over $100 billion annually) and ecological damages. Unfortunately, current understanding of large scale invasion patterns and distribution processes is still limited. This project, Modeling Invasion Dynamics Across Scales (MIDAS), seeks to understand how the underlying biological, geophysical and socioeconomic factors lead to the emergence of large scale invasion patterns. Findings of the project will help to inform forest managers, policy makers, and other stakeholders on how to better prevent and mitigate the economic and ecological damages of invasive species through multiple venues such as workshops and webinars, and online planning tools. The project will have high impact on the education of a diverse group of students and researchers through research opportunities, mentoring, and workshops. It will facilitate broad, long-lasting interdisciplinary and cross-institutional collaborations. In addition, the project will contribute to greater understanding of the general phenomenon of how "things" (e.g., disease, technology) spread, thus helping to sustain the spread of benefits while limiting unintended harms to ecosystems and society that emerge from social connectivity.The goal of this project is to determine the underlying mechanisms contributing to macroscale invasion patterns for two important taxa of invaders (plants and insect pests) in forests across the continental United States. The project will utilize a novel cross-modeling approach by combining: (1) machine learning-based statistical models aimed at capturing key within- and cross-scale interactions and tipping points among data challenged by different spatial and temporal resolutions, and (2) Bayesian models designed to capture model and parameter uncertainties. This project will result in improved mechanistic understanding and more robust predictions of macroscale invasion. The proposed conceptual framework can advance the field of invasion ecology by formulating new invasion theories and consolidating various existing ones. The project will also help advance the field of MacroSystems Biology by synthesizing and developing multidisciplinary datasets and by developing and refining a highly useful analytical tool for investigations involving many variables in which manipulative experiments would be challenging.