Interactions between phytoplankton and bacterioplankton communities in Caohai plateau lake, revealed by environmental DNA metagenomics.

Phytoplankton-bacterioplankton interactions critically influence aquatic ecosystem stability, yet their dynamics remain poorly understood in eutrophic plateau lakes. This study employed environmental DNA (eDNA) metagenomics to investigate these cross-kingdom relationships in Caohai Plateau Lake, a vulnerable wetland undergoing macrophyte-to-algae regime shifts, integrating correlation analysis, niche overlap, redundancy analysis (RDA), co-occurrence networks, and neutral community model (NCM). A total of 331 phytoplankton species across 10 phyla were characterized in the phytoplankton community, dominated by Cyanophyta with Microcystis as the representative genus, while bacterioplankton communities were primarily structured by Proteobacteria and Sphingomonas. The assembly of phytoplankton community was primarily driven by stochastic processes (R²>0.90). Co-occurrence network analysis showed phytoplankton interactions were dominated by positive effects (84.25%), whereas bacterioplankton networks exhibited balanced positive and negative effects. Metabolic specialization emerged through LEfSe analysis: phytoplankton specialized in photosynthesis and carbon storage, while bacterioplankton dominated anaerobic respiration (propanoate metabolism). The positive interactions were more prevalent than negative ones; combined with the metabolic complementarity of phytoplankton and bacterioplankton, this suggests that mutualism is more dominant than competition in cross-kingdom interactions. High niche overlap under sufficient nutrients (TP) facilitated species coexistence of phytoplankton and bacterioplankton by minimizing resource competition, thereby promoting stochastic community assembly, while keystone taxa (Cyanothece, Sphingobium) mediated ecosystem stability. This work demonstrates that nutrient enrichment promote stochastic assembly in eutrophic plateau lakes.