eDNA metabarcoding dataset of fish community composition in Qiandao Lake, China, for assessing niche dynamics of invasive Coptodon zillii

1. Background and study objective
This dataset supports the research article titled "Multi-dimensional niche partitioning at the invasion front: an eDNA-based study of invasive Coptodon zillii in Qiandao Lake, China". The study aimed to investigate the multi-dimensional spatial niche dynamics of the invasive redbelly tilapia (Coptodon zillii) at its northern invasion front in Qiandao Lake, a large subtropical deep-water reservoir in China, using environmental DNA (eDNA) metabarcoding. Specifically, we quantified niche breadth and niche overlap between C. zillii and co-occurring native fish species across vertical (surface, middle, bottom), horizontal (littoral vs. pelagic), and regional (five lake basins: central, southeastern, southwestern, northeastern, northwestern) dimensions.
2. Sampling design
In July 2025 (the breeding season of C. zillii), we established 24 sampling sites across five lake regions, with a 1:1 ratio between littoral and pelagic zones. At each site, water samples were synchronously collected from three layers: surface (0.5 m below surface), middle (50% water depth), and bottom (0.5 m above bottom). Three replicate 2 L water samples were collected per layer (total 6 L per layer), homogenized, and filtered on-site using mixed cellulose acetate membranes (47 mm diameter, 0.45 µm pore size; Whatman, UK). An additional 6 L of sterilized ultrapure water was filtered as an on-site negative control. All membranes were flash-frozen in liquid nitrogen and stored at −80°C until DNA extraction.
3. eDNA metabarcoding and sequencing
eDNA was extracted using the Fast DNA Spin Kit (MP Bio, USA). PCR amplification targeted a ~170 bp fragment of the 12S mitochondrial rRNA gene using the MiFish-U primers (Miya et al., 2015): forward 5′-GTCGGTAAAACTCGTGCCAGC-3′ and reverse 5′-CATAGTGGGGTATCTAATCCCAGTTTG-3′, appended with Illumina adapters. Eight PCR replicates per sample were performed to reduce stochasticity. Illumina paired-end sequencing (250 bp, v2 kit) was performed on an Illumina MiSeq platform with a target depth of 10,000 reads per sample. Raw sequencing data were processed using a standard bioinformatics pipeline including Cutadapt (v3.4), DADA2 (v1.20), and taxonomic assignment against MitoFish and GenBank databases.
4. Data usage
These raw data can be used for fish community composition analysis, biodiversity assessment, detection of invasive species, and validation of eDNA metabarcoding protocols in large deep-water reservoirs. Researchers may also reuse these data for meta-analyses of freshwater fish eDNA studies or for comparative studies with traditional capture-based fish surveys.
5. Contact information
For any questions regarding this dataset, please contact Shoujie Tang ([e-mail]) or Prof. Jinliang Zhao ([e-mail]), Shanghai Ocean University, Shanghai, China.