The Southern Ocean plays a pivotal role in global carbon cycling and climate regulation. As a fundamental component of Antarctic ecosystems, phytoplankton provide the main nutritional support for Antarctic krill (Euphausia superba Dana), a vital fishery resource in the Southern Ocean. Variations in phytoplankton community composition and abundance directly affect the growth, reproduction, and distribution of Antarctic krill, subsequently affecting the entire Antarctic food web and biological carbon pump processes. Understanding phytoplankton community dynamics is essential for assessing ecosystem responses to climate change and managing Southern Ocean fishery resources. In the context of global climate change, the physical environment and ecosystems of the Southern Ocean are undergoing rapid transformation. Since the beginning of the 21st century, environmental DNA (eDNA) metabarcoding has emerged as a powerful tool for studying phytoplankton diversity by enabling the rapid molecular-level assessment of distribution patterns across marine environments. This technique utilizes universal primers to amplify targeted DNA fragments for high-throughput sequencing, enabling the simultaneous analysis of multiple species in environmental samples, while effectively capturing microscopic and cryptic species. For eukaryotic phytoplankton, the 18S rDNA gene is most commonly used because of its universality, ease of amplification, and ability to provide phylogenetic information across taxonomic levels. Among the nine hypervariable regions (V1–V9), the V4 region is particularly valuable for marine phytoplankton studies. As the longest variable region, it shows superior species discrimination for diatoms and dinoflagellates, making it widely used in marine phytoplankton diversity research. However, the accuracy of diversity assessments depends heavily on bioinformatics methods for delineating taxonomic units, primarily operational taxonomic units (OTUs) and amplicon sequence variants (ASVs). Although both methods are widely used, their comparative performance in eukaryotic phytoplankton communities, particularly in Polar Regions, remains underexplored.
This study compared two widely used bioinformatics pipelines for processing Antarctic phytoplankton eDNA metabarcoding data: One (VSEARCH-UPARSE) that employs a clustering method that generates OTUs at 97% similarity and the other (USEARCH-UNOISE3) based on a denoising (error-correcting) algorithm that generates ASVs. Specifically, we assessed differences in their taxonomic resolution, α and β diversity indices, and their implications for identifying ecologically significant taxa, providing insights into methodological choices for polar biodiversity studies. Seawater samples were collected from surface waters at seven stations in the Antarctic Peninsula, Cosmonaut Sea, and Prydz Bay during the austral summer (January–March 2022) aboard the icebreaker Xuelong2 as part of China’s 38th Antarctic Expedition. Samples were immediately filtered through 0.22 μm mixed cellulose ester membranes, which were then wrapped in aluminum foil and stored at –80°C until DNA extraction. Total eDNA was extracted using the EZNA Soil DNA Kit (Omega Bio-Tek). The V4 hypervariable region of the 18S rDNA was amplified using universal primers 573F (5ʹ-CGCGGTAATTCCAGCTCCA-3ʹ) and 951R (5ʹ-TTGGYRAATGCTTTCGC-3ʹ). The PCR products were purified, quantified, and paired-end sequenced on an Illumina MiSeq platform, with blank filters used as negative controls.
The results showed that phytoplankton communities derived from both pipelines were dominated by three key phyla: Dinoflagellata (dinoflagellates), Bacillariophyta (diatoms), and Haptophyta (mainly Phaeocystis antarctica). In α diversity assessments, the Simpson and Shannon diversity indices showed no significant differences between OTU and ASV datasets (P>0.05), with ecologically reasonable values ranging from 0.54–0.91 (Simpson) and 1.80–3.13 (Shannon), indicating that both pipelines can effectively characterize the overall structure of phytoplankton communities in the study areas. However, the OTU approach generated significantly higher richness indices than the ASVs (P<0.05), whereas the ASVs exhibited greater Pielou’s evenness indices. This discrepancy likely stems from artifacts in OTU clustering, where sequence errors or intragenomic variants may inflate diversity estimates by generating low-abundance false-positive OTUs. At the species level, both pipelines identified common dominant taxa in Antarctic waters, including P. antarctica, Corethron inerme, Chaetoceros dichaeta, Gymnodinium sp., and Prorocentrum sp., which is consistent with known Antarctic assemblages. Among these species, P. antarctica has frequently been underestimated in traditional microscopic examinations owing to its small cell size (typically 5–10 μm). Crucially, ASVs provided more precise taxonomic information, identifying dominant species such as Fragilariopsis cylindrus and Fragilariopsis kerguelensis, which were underestimated or exhibited less accurate annotation in the OTU database. The ASV pipeline also detected the potentially invasive dinoflagellate Ptychodiscus noctiluca, which was absent in the OTU results. For β diversity analysis, weighted distance indices (Bray-Curtis and Weighted Unifrac) showed consistent patterns between pipelines (P>0.1), whereas unweighted indices (Jaccard and Unweighted Unifrac) exhibited significant discrepancies (P<0.05). Therefore, we suggest prioritizing abundance-weighted approaches to describe β diversities for more robust community comparisons.
In conclusion, both OTU and ASV pipelines can effectively characterize the Antarctic phytoplankton community characteristics from eDNA metabarcoding based on high-throughput sequencing data, whereas the ASV pipeline appears to show greater potential for future applications in achieving finer taxonomic resolution or inter-study comparisons. A comparative analysis of different pipelines for minimum taxonomic unit divisions can provide a theoretical basis for the selection of methods for future phytoplankton diversity research. With the advances in data processing methods, the influence of different algorithms and genetic markers on metabarcoding-based phytoplankton community diversity analysis requires further exploration.
1 材料与方法
2 结果
3 讨论




