Rapid urbanization and its associated human activities have facilitated the colonization and spread of non-native species, rendering urban ecosystems, particularly in megacities such as Beijing, highly susceptible to biological invasions. This study employed environmental DNA (eDNA) metabarcoding to evaluate the biodiversity and geographical distribution of non-native fish, as well as their interactions with native fish species, across three river basins in Beijing pertaining to the Daqing River, the North Canal, and the Ji Canal. Across all the 67 sampling sites, we identified 60 fish taxa, representing 11 orders, 23 families, and 40 genera, with an average of 33.0 taxa per site. Of these, 40 taxa were native, accounting for only 47.1% of the historically recorded native fish species. Additionally, we detected 20 non-native fish taxa, spanning 11 orders, 13 families, and 17 genera. Native fish exhibited geographical homogenization across the basins, while non-native taxa displayed varied geographical distributions. Non-metric multidimensional scaling (NMDS) and analysis of similarities (ANOSIM) revealed no significant variation in the non-native communities across the river basins. Although most of the non-native taxa were widespread, some were restricted to specific sites or basins. The North Canal exhibited significantly lower non-native biodiversity compared with the Ji Canal across all alpha diversity indices. Simple linear regression analyses indicated positive correlations between the number of taxa and species richness for both native and non-native taxa. Interestingly, species co-occurrence analyses revealed predominantly positive interactions among both native and non-native species pairs, with only two negative relationships involving one native and two non-native taxa. This study provides insights into the biodiversity and geographical distribution of non-native fish in Beijing and establishes a baseline for future biomonitoring and conservation efforts. The findings underscore the need for further investigation into the mechanisms and dynamics of biological invasions within urban environments in Beijing.
Environmental RNA (eRNA)‐based techniques have rapidly emerged as powerful tools for biodiversity assessment across a range of theoretical and applied disciplines. Reverse transcription, the process of converting eRNA into complementary DNA (cDNA), is a crucial step in eRNA‐based assessments. However, primer length optimization for reverse transcription is often overlooked. Here we evaluated how varying primer lengths affected two key parameters—biodiversity recovery and reproducibility—across the two best‐performing reverse transcription strategies: random priming and oligo(dT) priming for eukaryotic eRNA‐metabarcoding. Using fish eRNA collected from a coastal zone, we conducted technical replicates to assess the performance of primer lengths ranging from 4‐mer to 30‐mer for both strategies. Our results showed that primer length significantly affected biodiversity recovery, reproducibility, and community structure of replicates. Specifically, shorter primers, particularly N 4 and N 5 , in random priming and medium‐length primers (T 15 ‐ 24 ), particularly T 20 , in oligo(dT) priming demonstrated superior performance, especially in detecting biodiversity in the rare biosphere. A comparative analysis of random and oligo(dT) primers highlighted the advantage of shorter random primers in improving biodiversity detection sensitivity and consistency. To explicitly reduce false negatives arising from reliance on a single reverse transcription strategy, we recommend the combined use of short random primers (e.g., N 4 and N 5 ) alongside medium‐length primers (e.g., T 20 ) to enhance comprehensive biodiversity recovery. Our findings provide valuable insights for improving the accuracy and reliability of eRNA‐metabarcoding in biodiversity assessments, thus facilitating the development of standardized methods to enable cross‐laboratory comparisons and ensure consistent, robust outcomes in large‐scale meta‐analyses.
Environmental RNA (eRNA) metabarcoding has rapidly emerged as a powerful tool for assessing contemporary biodiversity patterns across diverse ecosystems. However, the potential for false positive detections caused by co‐extracted environmental DNA (eDNA) remains unquantified. Distinguishing true signals from false positives caused by residual eDNA is a technical challenge in eRNA‐based metabarcoding. To address this issue, we employed a freshwater river receiving treated effluent from a wastewater treatment plant as a model system. In such settings, eDNA in the treated effluent can lead to the detection of non‐local species (e.g., marine taxa). Treated effluent typically contains minimal or no eRNA, making it well‐suited for evaluating the influence of eDNA carryover. By comparing DNase‐treated and untreated eRNA samples, we assessed the impact of residual eDNA on fish species richness and community composition. Our results showed that omitting DNase treatment significantly inflated taxonomic richness, with untreated samples detecting a conservative estimate of over 25% more taxa per site. Fold‐change analysis revealed that residual eDNA inflated taxon abundances in both high‐ and low‐abundance taxa, with some showing over 10‐fold increases. Community composition analyses revealed clear clustering between treated and untreated samples, highlighting substantial shifts driven by residual eDNA. These findings demonstrate that co‐extracted eDNA can severely distort eRNA‐based biodiversity estimates, leading to false positives and misrepresented contemporary community profiles. We recommend further evaluation of DNase treatment parameters, including enzyme concentration, incubation time and treatment times, and the adoption of optimised protocols to standardise and improve the accuracy of eRNA‐based biodiversity monitoring.