Benthic dinoflagellates have gained significant research attention due to the increased occurrence of human poisoning linked to Benthic Harmful Algal Blooms (BHABs). Leveraging metabarcoding with high throughput sequencing enable investigation of the diversity and dynamics of benthic dinoflagellate community assemblages. Importantly, this approach complements previous studies that predominantly emphasized morphological aspects. To examine the diversity of benthic harmful dinoflagellates in a tropical reef, this study performed a multiple metabarcode approach by sequenced the small subunit (SSU) and large subunit (LSU) ribosomal RNA genes simultaneously across 35 natural and artificial substrate samples collected at Perhentian Islands, Malaysia. Our results yielded a total of 2,611,086 reads (12,777 ASVs) and 2,366,464 reads (6,621ASVs) for SSU and LSU datasets, respectively. Only 31.2% of the ASVs annotated to protists, and more than 20% of these protist species remained unclassified. Benthic dinoflagellates constituted a minor fraction of the community (9.5% of the total protists), with approximately 6% annotated to species level. The compositions of benthic harmful dinoflagellates were extremely low, with Amphidinium the highest, constituting 15.5% of the dinoflagellate composition, followed by Gambierdiscus (8%), Ostreopsis (7.3%), Prorocentrum (2.9%), and Coolia (0.1%). Despite the substantial cell abundances observed morphologically in the samples, both Prorocentrum and Coolia exhibited comparatively lower detection frequencies. Notably, several potentially harmful dinoflagellates, including species of Alexandrium, Azadinium, Amphidoma, Blastodinium, Gonyaulax, Gyrodinium, Heterocapsa, Margalefidinium, andPfiesteria were detected for the first time in the study sites. Our study provides a comprehensive assessment of benthic dinoflagellate diversity within a tropical reef ecosystem. Despite the possible underrepresentation attributed to constraints in current reference databases, this taxonomic inventory serves as a basis for future monitoring efforts in the study sites.  

harmful algal blooms (HABs), benthic harmful algal blooms (BHABs), eDNA, metabarcoding, artificial substrate