Contrasting summer and winter phytoplankton community states in the regulated Xiaolangdi Reservoir of the Yellow River revealed by shotgun metagenomics

Freshwater phytoplankton communities can differ strongly between sampling periods, but the relative importance of seasonal and spatial variation in regulated reservoirs is not fully resolved. Shotgun metagenomics characterized phytoplankton in Xiaolangdi Reservoir, China, at two contrasting endpoints: winter (31 December 2023) and early summer (6 June 2024). Four reservoir sites were investigated, with three spatially separated field replicates per site and period, yielding 24 samples. Genus-level alpha diversity, Bray–Curtis PERMANOVA and PCoA, CLR/Aitchison sensitivity analysis, multivariate dispersion, and site-stratified indicator analysis were used. Rarefied genus richness was higher in summer than winter (61.74 ± 6.82 vs. 41.22 ± 8.88), whereas Shannon diversity (1.814 ± 0.093 vs. 2.402 ± 0.195) and Pielou evenness (0.441 ± 0.028 vs. 0.650 ± 0.045) were lower in summer. Bray–Curtis PERMANOVA showed that sampled season/state accounted for 91.6% of genus-level compositional variation (pseudo-F(1, 16) = 466.16, R2 = 0.916, P = 0.0001), while CLR/Aitchison analysis gave a more conservative estimate of 52.9% (R2 = 0.529, P = 0.0001). Site and Season × Site effects were much smaller. Summer communities showed high relative sequence representation of Microcystis, Synechococcus, and Cyanobium, whereas winter communities were more evenly distributed across eukaryotic lineages. High-confidence summer indicators were Cyanobium, Synechococcus, Microcystis, and Ceratium; winter indicators were Chrysochromulina, Cryptomonas, Mychonastes, Discostella, Choricystis, and Cyclotella. Several abundant winter assignments, especially Teleaulax-affiliated cryptophyte, Thalassiosira, and Skeletonema, require reference-genome validation before species-level interpretation.