文献类型：期刊文献

英文题名：Unveiling the dynamic photosynthetic strategies of Heterosigma akashiwo: An interpretable machine learning approach to light and phosphorus stress

作者：Zheng, Qiwen[1,2,3,4];Zhao, Hui[1,2,3,4];Duan, Meina[1,3]

机构：[1]Guangdong Ocean Univ, Coll Chem & Environm Sci, Zhanjiang, Guangdong, Peoples R China;[2]Southern Marine Sci & Engn Guangdong Lab Zhuhai, Zhuhai, Peoples R China;[3]Guangdong Ocean Univ, Res Ctr Coastal Environm Protect & Ecol Resilience, Zhanjiang, Peoples R China;[4]Guangdong Ocean Univ, Cooperat Res Ctr Nearshore Marine Environm Change, Zhanjiang, Peoples R China

年份：2026

卷号：213

外文期刊名：MARINE ENVIRONMENTAL RESEARCH

收录：SCI-EXPANDED(收录号：WOS:001620892700001)、、WOS

基金：This work was supported by Doctoral research launch project of School of chemistry and environment, Guangdong Ocean University (R20032) ; National Natural Science Foundation of China (No. 42076162) ; Innovation Group Project of Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai) (No. 311020004) ; Guangdong Basic and Applied Basic Research Fund Project (2022A1515010460) . The authors would like to thank Xiaotong He for her assistance with the experiments.

语种：英文

外文关键词：Heterosigma akashiwo; Harmful algal blooms; Ecological management

外文摘要：The marine eukaryotic alga Heterosigma akashiwo (H. akashiwo), a notorious species that forms fish toxic blooms, threatens coastal and freshwater ecosystems globally through harmful algal blooms (HABs). This study innovatively integrates algal cultivation experiments, chlorophyll fluorescence dynamics, and interpretable machine learning techniques to elucidate the regulatory mechanisms of light and phosphorus on the growth and photosynthesis of H. akashiwo. We found that light intensity significantly influenced the growth of H. akashiwo, with the highest cell density observed at 4000 lx. Elevated phosphorus concentrations enhanced biomass, and a high supply may partially alleviate photoinhibition under high light. The regulation of photosynthetic activity showed significant independent and interactive effects, with light intensity as the dominant factor and phosphorus as an important regulatory factor. Optimal photosynthetic efficiency and PSII activity were observed around 4000 lx and 49 mu mol L- 1. Machine learning identified light absorption per excited cross section (ABS/CSO), energy dissipation per excited cross section (DIO/CSO), and trapped energy per excited cross section (TRO/CSO) as key growth predictors. The importance of photosynthetic parameters varied across growth stages: cells shift their regulatory strategy from multiparameter coordination to light utilization efficiency, eventually prioritizing both enhanced light utilization and energy dissipation. This work is the first to integrate OJIP chlorophyll fluorescence kinetics with explainable machine learning models to unravel how light and phosphorus jointly regulate the photosynthetic strategies and bloom potential of H. akashiwo. Our interpretable framework offers both high predictive accuracy and mechanistic insights, providing a novel foundation for future HABs monitoring and ecological management.