环境扰动下传粉生态服务可持续性的多粒度预测模型与验证研究

Ecological service of pollination system has been critical to human survival and development. Understanding the resilience, sustainability and ecological role of pollination system subject to environmental disturbances has theoretical significance and application value in predicting and deterring the service decay. However, researches on unveiling evolution patterns and key factors that drive pollination system have not shown due to the system’s complicated traits such as nonlinear, heterogeneity, adaptivity and multiple stable states. By incorporating into complex adaptive system perspectives and interdisciplinary knowledge, e.g., pollination ecology, environmental science as well as information science, and by taking advantage of pollination ecology big data and the multigranular model that represents the system feedbacks, this project aims to discover key variables that determine the trajectory of pollination system dynamics and the confidence interval of those variables that help the system maintaining its resilience. In addition, the project investigates the possible decaying routes and critical conditions under which resilience can be maintained when disturbances, such as climate change, pollination insects decline, and niche destruction occur in various amplitudes and frequencies. The primary innovation and contribution of the project will be that the new research paradigm is going to be established for studying pollination ecology with multigranular model; and the scientific evidence, strategies and feasible early warnings will be provided in order to protect pollination ecological system.

传粉生态服务与人类生存密切相关，认知传粉生态系统在环境扰动下的调节能力，理解传粉生态服务可持续性的形成机制和生态学意义，对于预测和减缓其退化进程、应对生态灾难具有重要意义。然而，针对传粉生态系统的非线性、异质性、自适应性和多重稳态等复杂特征，如何解码其演化模式并发现驱动系统的关键因素，目前国内外尚缺乏具体研究。本项目将有机结合传粉生态学、环境科学与信息科学等多学科知识，以复杂适应系统为视角，以传粉生态大数据为基础，以构建能够表达传粉生态系统复杂特征的多粒度预测模型为工具，探明决定传粉生态系统演化轨迹的关键变量，估计使系统保持弹性的阈值；并结合对野生蓝莓传粉系统的预测和验证，探索在气候变化、传粉昆虫锐减和生境破碎等环境因素及其不同幅度和频度的扰动下，传粉生态系统维持其服务的临界条件和可能的退化路径，建立以多粒度模型研究传粉生态演化的新范式，为保护传粉生态系统提供科学依据和可行的技术预警手段。

项目在传粉生态学、环境科学与信息科学等多学科背景下，以复杂适应系统为视角，以传粉生态大数据为基础，以弹性理论为指导，以构建能够表达传粉生态系统复杂特征的多粒度预测模型为工具，探索了结合元模型全局参数敏感性度量和非线性动力学分析对传粉生态系统稳态结构、失稳临界条件和可能演化路径进行量化的新方法，探明了决定传粉生态系统演化轨迹的关键变量，估计使系统保持弹性的阈值；通过对野生蓝莓传粉系统的预测和验证，探索了在气候变化、传粉昆虫锐减和生境破碎等环境因素及其不同幅度和频度的扰动下，传粉生态系统维持其服务的临界条件和可能的退化路径，建立了以多粒度模型研究传粉生态演化的新范式，为保护传粉生态系统提供科学依据和可行的技术预警手段。重要贡献在于通过全局敏感性度量发现对系统稳定性具有显著性影响的因子（慢变量），揭示因子的贡献度及阈值，为确定环境扰动下传粉系统状态所属态势及其临界条件提供较为精确的参数估计，以此为依据构建了生态大数据分析和面向深度学习的生态过程模式发现的理论框架。在基金委的资助下，本项目在生态学、信息科学等领域TOP期刊发表SCI论文19篇，高水平国际会议论文7篇，其中2篇获得最佳论文奖，获得授权发明专利2项，毕业博士2人，硕士10人，获得省部级科技进步二等奖1项。