Restoration of marine ecosystems under the impacts of fisheries and climate

我的研究旨在评估和减轻渔业对水生生态系统的影响；对水生食物网的开发破坏了浮游鱼类和底栖鱼类之间的平衡，从而影响了人类渔业的长期可持续性。新的方法、方法和工具被用作假设检验引擎。我的研究特别集中在过去生态系统状态的历史重建如何提供定量的长期管理目标和管理方案评估的模板。这项工作采用了几种类型的动态数值生态系统模型，包括空间模型，允许测试关于渔业如何改变水生生态系统中生物的营养结构、栖息地、物种多样性、关键物种和生活史特征的假设。这种方法使得进行思想实验成为可能，因为真实的实验既不切实际又不道德。历史案例研究使用空间模拟技术，包括枯竭、过度捕捞和污染系统中鱼类和渔业的恢复、禁止捕捞海洋保护区的影响、部门关闭和人造珊瑚礁。这项工作试图模拟浮游生物动力学、滋养动力学、栖息地结构和生活史参数中的重要生态过程。用大量来自调查和实验分析的数据来挑战模型，从而改进模型。参数不确定性由蒙特卡罗技术来解释，而气候波动是利用生态系统模式中选定节点的一系列环境强迫函数来模拟的；使用半贝叶斯蒙特卡罗技术评估风险。我还在进一步开发一种多学科快速评估技术，称为“Rapfish”，它使用非参数排序和模糊逻辑来提供带有置信限的数字分数，以表达对既定管理目标的遵守情况，例如可持续性或对国际行为准则的遵守情况。案例研究来自广泛的温带、热带和极地生态系统。

My research aims to assess and mitigate the impacts of fisheries on aquatic ecosystems; exploitation in aquatic food webs shifts the balance between schooling pelagic fishes and demersal fish, and hence affects the long-term sustainability of human fisheries. New approaches, methodologies and tools are employed as hypothesis testing engines. In particular my research focusses on how historical reconstruction of past ecosystem states provides quantitative long-term management goals and a template against which management options can be evaluated.  The work employs several types of dynamic numerical ecosystem modelling, including spatial models, that allows tests of hypotheses about how fisheries alter the trophic structure, habitat, species diversity, keystone species and life history characteristics of organisms in aquatic ecosystems. This methodology makes it possible to carry out thought-experiments, since real ones are both impractical and unethical. Historical case studies use spatial simulation techniques, including the recovery of fish and fisheries in depleted, overfished and polluted systems, the effects of no-take marine reserves, sectoral closures and human-made reefs. The work attempts to emulate important ecological processes in plankton dynamics, trophodynamics, habitat structure and life history parameters. Models are improved by challenging them with considerable amounts of data from surveys and experimental analyses. Parameter uncertainty is accounted by for Monte Carlo techniques while climate fluctuations are simulated using a range of environmental forcing functions at selected nodes in the ecosystem models; risks are evaluated using semi-Bayesian Monte Carlo techniques. I am also further developing a multi-disciplinary rapid evaluation technique, termed "Rapfish", that uses non-parametric ordination and fuzzy logic to provide numerical scores, with confidence limits, that express compliance with stated management goals, such as sustainability, or compliance with international codes of conduct. Case studies come from a wide range of temperate, tropical and polar ecosystems.