Seasonal variation of fish biodiversity in Xisha Islands based on dual-marker eDNA metabarcoding.

Environmental DNA (eDNA) is increasingly used for assessing marine biodiversity, yet its potential for monitoring seasonal dynamics in tropical coral reef ecosystems requires further systematic investigation. This study applied dual mitochondrial marker (Cytb and 12S rRNA) eDNA metabarcoding to systematically analyze fish biodiversity and community dynamics in spring (March 2021) and winter (January 2022) at Xisha Islands, a typical coral reef ecosystem in the South China Sea. A total of 432 fish species from 89 families were identified, with a stable core assemblage accounting for 64.81% of all species across both seasons. Community structure differed between seasons (ANOSIM, R = 0.375, p < 0.001). eDNA metabarcoding effectively detected cryptobenthic, deep-sea, and pelagic species that may be underdetected by established survey methods. Based on relative abundance, fish composition at Xisha Islands is dominated by Scaridae (parrotfishes) and Acanthuridae (surgeonfishes), consistent with previously documented local trophic transitions (from carnivore-dominated to herbivory-prevalent ecosystem) in this area. The dual-marker eDNA metabarcoding enhanced the efficiency of species detection, increasing taxonomic coverage by 54.38% (spring) and 62.50% (winter) compared to single markers, while mitigating primer bias and database limitations. Our results confirm that eDNA metabarcoding is an effective tool for monitoring fish biodiversity and community dynamics in remote and complex marine ecosystems. This study establishes critical baseline data for Xisha Islands coral reef ecosystem and provides a validated dual-marker framework aimed at supporting the effectiveness assessment and adaptive management of marine protected areas.