机器学习预测珊瑚三角区的连通性、生物多样性和复原力
Machine Learning Prediction of Connectivity, Biodiversity and Resilience in the Coral Triangle
作者：Lyuba Novi, Annalisa Bracco
期刊：Communications Biology
Even optimistic climate scenarios predict catastrophic consequences for coral reef ecosystems by 2100. Understanding how reef connectivity, biodiversity and resilience are shaped by climate variability would improve chances to establish sustainable management practices. In this regard, ecoregionalization and connectivity are pivotal to designating effective marine protected areas. Here, machine learning algorithms and physical intuition are applied to sea surface temperature anomaly data over a twenty-four-year period to extract ecoregions and assess connectivity and bleaching recovery potential in the Coral Triangle and surrounding oceans. Furthermore, the impacts of the El Niño Southern Oscillation (ENSO) on biodiversity and resilience are quantified. We find that resilience is higher for reefs north of the Equator and that the extraordinary biodiversity of the Coral Triangle is dynamic in time and space, and benefits from ENSO. The large-scale exchange of genetic material is enhanced between the Indian Ocean and the Coral Triangle during La Niña years, and between the Coral Triangle and the central Pacific in neutral conditions. Through machine learning the outstanding biodiversity of the Coral Triangle, its evolution and the increase of species richness are contextualized through geological times, while offering new hope for monitoring its future.
Figure 1. Conceptual framework and workflow
Schematic representation of the proposed ecoregionalization, connectivity and resilience framework.
到2100年，珊瑚礁生态系统将面临灾难性的后果，这样的预测甚至出现在了那些对未来持乐观态度的气候情景中。了解珊瑚礁的连通性、生物多样性和复原力如何受到气候变化的影响，将有助于提高建立可持续管理做法的机会。在这方面，生态区域化和连通性对于指定有效的海洋保护区是至关重要的。本研究将机器学习算法和物理直觉用于研究长达24年的海面温度异常数据，以选取生态区域，评估珊瑚三角区和周边海洋的连接性和白化恢复潜力。此外，该研究对厄尔尼诺-南方振荡现象（ENSO）对生物多样性和复原力的影响进行了量化分析。我们发现，赤道以北的珊瑚礁复原力更强，珊瑚三角区非凡的生物多样性在时间和空间上是动态的，并受益于厄尔尼诺-南方振荡现象。在拉尼娜年，印度洋和珊瑚三角区之间的大规模遗传物质交换得到加强，而在中性条件下，珊瑚三角区和太平洋中部之间的交换也得到加强。通过机器学习，珊瑚三角区杰出的生物多样性、其演变和物种丰富度的增加在地质时代被情景化的同时，也为监测其未来提供了新的希望。
