Diatom community composition has a critical influence on global ocean health and ecological processes. Developing accurate and efficient methods for diatom identification under dynamic environmental conditions is essential to understanding the implications of diatom community changes. Two developing methods for identifying and enumerating phytoplankton, cell imaging and molecular sequencing, are experiencing rapid advancements. This study aims to compare diatom taxonomic composition results within natural assemblages derived from rapidly advancing methods, FlowCam imaging and metabarcoding of the V4 region of the 18S rRNA gene, with traditional light microscopy cell counting techniques. All three methods were implemented in tandem to analyze changes in dynamic diatom assemblages within simulated upwelling experiments conducted in the California upwelling zone. The results of this study indicate that, summed across all samples, DNA sequencing detected four times as many genera as morphology-based methods, thus supporting previous findings that DNA sequencing is the most powerful method for analyzing species richness. Results indicate that all three methods returned comparable relative abundance for the most abundant genera. However, the three methods did not return comparable absolute abundance, primarily due to barriers in deriving quantities in equal units. Overall, this study indicates that at the semi-quantitative level of relative abundance measurements, FlowCam imaging and metabarcoding of the V4 region of the 18S rRNA gene yield comparable results with light microscopy but at the qualitative and quantitative levels, enumeration metrics diverge, and thus method selection and cross-method comparison should be performed with caution.

中文翻译：

---

动态沿海群落中硅藻识别先进方法的比较

硅藻群落组成对全球海洋健康和生态过程具有至关重要的影响。开发动态环境条件下准确有效的硅藻识别方法对于了解硅藻群落变化的影响至关重要。细胞成像和分子测序这两种用于识别和计数浮游植物的开发方法正在快速发展。本研究旨在将来自快速发展的方法、FlowCam 成像和 18S rRNA 基因 V4 区域的元条形码的自然组合中的硅藻分类组成结果与传统的光学显微镜细胞计数技术进行比较。所有三种方法同时实施，以分析在加利福尼亚上升流区进行的模拟上升流实验中动态硅藻组合的变化。这项研究的结果表明，对所有样本进行总结，DNA 测序检测到的属数是基于形态学的方法的四倍，从而支持了之前的发现，即 DNA 测序是分析物种丰富度的最强大方法。结果表明，所有三种方法都返回了最丰富属的相对丰度。然而，这三种方法没有返回可比较的绝对丰度，主要是由于在导出相同单位的数量方面存在障碍。总体而言，本研究表明，在相对丰度测量的半定量水平上，18S rRNA 基因 V4 区域的 FlowCam 成像和元条形码可产生与光学显微镜相当的结果，但在定性和定量水平上，计数指标存在差异，因此应谨慎进行方法选择和交叉方法比较。