Collaborative Research: Effects of Changing Temperature on the Gulf of Alaska Ecosystem

这项研究有可能改变我们对气候如何影响海洋生态系统的理解，并改善基于生态系统的渔业管理工作。研究人员将分析现有数据，以确定气候特性随时间的变化可能如何影响阿拉斯加湾（GOA）具有重要商业价值的鱼类。几十年来，海表温度（SST）与果阿地区大量海洋生物丰度之间的重要关系为将环境数据纳入渔业管理战略提供了一些希望。然而，这些统计联系中的许多在20世纪80年代末突然恶化，同时GOA海温与气候条件其他方面之间的相互作用也发生了变化。本研究将验证这样一个假设，即20世纪80年代大尺度气候变率的转变导致了GOA大气和海洋特性之间关系的重组，而这反过来又导致了温度和海洋生物学之间关系的变化。这些类型的“无模拟气候状态”在古生态学中得到了很好的认识，并已被预测为气候变化的潜在结果，但很少有关于这种转换的生态响应的研究。该项目将支持一名博士后科学家，以及研究生和本科生研究人员。它将促进三所公立大学、一个非营利研究实验室和一个联邦管理机构的科学家之间的合作，以便将开展这项研究所需的各种专业知识结合起来。这项研究有可能改变我们对气候如何影响海洋生态系统的理解。初步分析表明，阿拉斯加湾海温与社区状态之间的统计关系似乎是非平稳的，1988/89年前后驾驶员-反应关系明显不同。初步分析还表明，在1988/89年前后，海温与其他一些果阿气候参数之间的相关性有显著差异。此外，北太平洋海温异常的主导模态在1988/89年前后的相对重要性发生了变化，在过去30年里，第二模态（北太平洋环流振荡，或NPGO）比第一模态（太平洋年代际振荡，或PDO）解释了更多的变异。这种变化与向无模拟气候状态的转变是一致的，其特征是流域尺度气候过程的变率模式明显不同。提出的假设将用统计生态模型和数值海洋模型相结合的方法加以检验。统计方法（阈值广义加性模型）将使用时间序列数据（1965年至今）对来自GOA的17个鲑鱼、底栖鱼和甲壳类种群的海温变率与群落和种群水平上的生态特征之间的非平稳关系进行正式检验。此外，在盆地和区域尺度上，大气和水文过程之间的非平稳关系将用海洋资料同化和后播模式产生的时间序列进行检验。最后，这些确定的非平稳关系将用于参数化生物变异的统计模型，该模型解释了系统中的非类似状态。

This research has the potential to transform our understanding of how climate affects marine ecosystems and improve efforts toward ecosystem-based fisheries management. Investigators will analyze existing data to determine how shifts in climate properties over time may have affected commercially important fishes in the Gulf of Alaska (GOA). For several decades, significant relationships between sea surface temperature (SST) and abundance of a broad group of marine organisms in the GOA offered some promise for incorporating environmental data into fisheries management strategies. However, many of these statistical connections deteriorated abruptly in the late 1980s, while at the same time interactions between GOA SST and other aspects of climate conditions changed as well. This study will test the hypothesis that a switch in large-scale climate variability in the 1980s led to reorganization of relationships among GOA atmospheric and oceanographic properties, which in turn, produced a change in connection between temperature and ocean biology. These types of "no-analogue climate states" are well recognized in paleoecology and have been anticipated as a potential outcome of climate change, but few studies of ecological response to such switches are available. The project will support a postdoctoral scientist, as well as graduate and undergraduate student researchers. It will facilitate cooperation among scientists at three public universities, a non-profit research lab, and a federal management agency in order to combine the range of expertise that will be required to carry out this research.This research has the potential to transform our understanding of how climate affects marine ecosystems. It is based on preliminary analyses showing that the statistical relationships between SST and community state in the Gulf of Alaska appear to be nonstationary, with driver-response relationships that differ markedly before and after 1988/89. Preliminary analyses also show that correlations between SST and a number of other GOA climate parameters are significantly different before and after 1988/89. Additionally, leading modes in North Pacific SST anomalies for 1950-present changed their relative importance before and after 1988/89, with the second mode (North Pacific Gyre Oscillation, or NPGO) explaining more variability than the first mode (Pacific Decadal Oscillation, or PDO) during the past three decades. This change is consistent with a switch to a no-analogue climate state, characterized by markedly different patterns of variability among basin-scale climate processes. The proposed hypothesis will be tested with a combination of statistical ecological models and numerical ocean models. The statistical approach (threshold generalized additive models) will provide formal tests for nonstationary relationships between SST variability and ecological characteristics at the community and population level using time series data (1965-present) for 17 salmon, groundfish and crustacean populations from the GOA. In addition, nonstationary relationships among atmospheric and hydrographic processes at basin and regional scales will be tested with time series generated by ocean data-assimilation and hindcast models. Finally, those nonstationary relationships identified will be used to parameterize statistical models of biological variability that account for non-analogous states in the system.