Research
| Title： | Environmental drivers and microbial interactions in harmful dinoflagellate blooms: Insights from metagenomics and machine learning |
|---|---|
| First author： | Zhang, Kuidong; Xi, Mansong; Wu, Guimei; Lu, Feimiao; Wu, Guichun; Zhou, Jun; Zhang, Jie; Wang, Xin; Li, Yanzhao; Xu, Cile; Yang, Mengxuan; Wang, Hongxia; Wu, Mingcan; Ma, Mingyang |
| Journal： | PROCESS SAFETY AND ENVIRONMENTAL PROTECTION |
| Years： | 2025 |
| DOI： | 10.1016/j.psep.2025.107205 |
| Abstract： | Freshwater dinoflagellate blooms have become a global environmental challenge, particularly in highland lakes. This study investigated dinoflagellate blooms in Lake Erhai, China, using microscopy, 18S rRNA sequencing, and metagenomics. The dominant species was identified as a variant of Peridinium gatunense. While no significant diurnal vertical distribution was observed (p > 0.05), geographic distribution varied significantly (p < 0.05). Metagenomic analysis revealed that specific microbial communities in dinoflagellate-dense areas promote their growth through enhanced metabolic pathways, particularly in nitrogen cycling and organic matter decomposition. Areas with high dinoflagellate density showed significantly reduced antibiotic resistance genes. Among various machine learning models tested, the gradient boosting model achieved optimal performance (R-2 = 0.88), identifying transparency, pH, permanganate index, ammonia nitrogen, and Ceratium spp. as key environmental drivers. These findings suggest that dinoflagellate blooms in Lake Erhai result from complex interactions between environmental factors and microbial communities, providing valuable insights for monitoring and controlling harmful algal blooms in highland lakes. |
