Modeling the biogeography of pelagic diatoms of the Southern Ocean, links to model results and images

Species distribution models (SDM) are a widely used and well established method for biogeographical research on terrestrial organisms. Though already used for decades, experience with marine species is scarce especially for protists. More and more observation data, sometimes even aggregated over centuries, become available also for the marine world, which together with high quality environmental data form a promising base for marine SDMs. In contrast to these SDMs, typical biogeographical studies of diatoms only considered observation data from a few transects.
Species distribution methods were evaluated for marine pelagic diatoms in the Southern Ocean at the example of F. kerguelensis. Based on the experience with these models, SDMs for further species are built to study biogeographical patterns. The anthropogenic impact of climate change on these species is assessed by model projections on future scenarios for the end of this century.
Besides observation data from public data repositories such as GBIF, data from the Hustedt diatom collection was used. The models presented here rely on so called presence only observation data. For this simple data type Maxent has been proven to be a good modeling method.
SDM seems a suitable modeling method to study biogeography of marine pelagic diatoms in the Southern Ocean. Models of decent quality could be build, despite partly poor data. Future projections indicate a moderate decrease of the suitable areas towards the end of the century for most of the species.

物种分布模型（Species Distribution Models，SDM）是陆地生物生物地理学研究中应用广泛且成熟的经典方法。尽管该方法已应用数十年，但针对海洋物种，尤其是原生生物的应用经验仍较为匮乏。如今海洋领域也涌现出越来越多的观测数据，部分数据集甚至是数百年间累积的记录；这些数据结合高质量的环境数据，为海洋物种分布模型的构建提供了极具潜力的基础。与这类海洋物种分布模型不同，典型的硅藻生物地理学研究仅依托少量断面的观测数据开展。
本研究以凯尔盖朗舟形藻（F. kerguelensis）为实例，针对南大洋海洋浮游硅藻评估了物种分布建模方法。基于上述模型的构建经验，本研究进一步为其他物种搭建了物种分布模型，以探究其生物地理分布格局。本研究还通过基于本世纪末未来气候情景的模型预测，评估了人为气候变化对这些物种的影响。
除了取自全球生物多样性信息设施（Global Biodiversity Information Facility，GBIF）等公共数据仓库的观测数据外，本研究还使用了胡斯特特硅藻馆藏的数据。本研究所构建的模型均基于仅出现记录观测数据（presence-only observation data），针对这类简易数据类型，最大熵模型（Maxent）已被证实为优质的建模方法。
物种分布模型是研究南大洋海洋浮游硅藻生物地理学的适宜建模方法。尽管部分数据集质量欠佳，但仍可构建出质量尚可的模型。未来情景预测显示，至本世纪末，多数物种的适宜生境面积将出现小幅缩减。