Harmful algal blooms driven by toxigenic Pseudo-nitzschia species pose a critical threat to coastal ecosystems, necessitating precise tracking of marine biotoxins domoic acid. To assess the prevalence and ecological drivers of these potential toxin-producing assemblages in the Taiwan Strait, this study investigated their molecular diversity and spatial-temporal dynamics. An integrated molecular approach was employed combining high-resolution LSU and ITS rDNA metabarcoding with species-specific quantitative PCR (qPCR) assays across nine cruise expeditions conducted between November 2018 and April 2022. Processing of the amplicon dataset retained 11,295 eukaryotic amplicon sequence variants (ASVs) for compositional analysis, of which 1,188 were classified as diatoms (Bacillariophyceae). Pseudo-nitzschia exhibited a high regional prevalence, occupying 69% (231 of 334) of the surveyed samples. Rigorous taxonomic assignment and phylogenetic inference further resolved 64 Pseudo-nitzschia ASVs into 30 distinct species, highlighting distinct seasonal successions among the planktonic assemblages. Deploying a multi-marker analysis utilizing both LSU and ITS1 resolved cryptic species limitations, confirming P. delicatissima as the dominant species, differentiating it from P. allochrona, and unveiling previously undetected intraspecific divergence within the P. delicatissima complex. By synthesizing metabarcoding read counts with targeted qPCR data for dominant taxa (P. delicatissima, P. galaxiae, and P. multistriata), this study mapped the biogeographic patterns of key clades in relation to biotic and abiotic environmental drivers. These findings deepen the current understanding of microalgal diversity and the genetic heterogeneity underlying toxic bloom events, providing a robust scientific foundation for the ongoing management and mitigation of harmful algal blooms in marine environments.