DNA metabarcoding provides a scalable alternative to traditional botanical surveys, which are often time-consuming and reliant on taxonomic expertise. Here, we compare DNA metabarcoding with quadrat-based botanical surveys to assess plant species composition in experimental grassland plots under four defoliation regimes (continuous grazing, rotational grazing, frequent cutting and conservation cutting). Botanical surveys identified 16 taxa, while metabarcoding detected 25 taxa, including the dominant species Holcus lanatus and Lolium perenne. Despite detecting more taxa, there were some discrepancies in identification, with the sequence data only able to resolve some taxa at the genus level (e.g., Agrostis spp. instead of Agrostis capillaris) and potential species misidentifications (e.g., Cardaminopsis helleri vs. Cardamine flexuosa). However, both methods provided comparable results and revealed statistically significant differences in species composition between treatments, with higher diversity in cut versus grazed plots. The semi-quantitative nature of metabarcoding limits its capacity to accurately reflect species abundance, posing challenges for ecological interpretations where precise quantification is required. However, it provides a broader view of biodiversity and can complement traditional methods, offering new opportunities for efficient biodiversity monitoring. The findings support the integration of DNA metabarcoding into biodiversity assessments, particularly when used alongside traditional techniques. Further refinement of bioinformatics tools and reference databases will enhance their accuracy and reliability, enabling more effective monitoring of grassland biodiversity and sustainable management practices. This study highlights DNA metabarcoding as a valuable tool for understanding plant community responses to management interventions.

• Klíčová slova
• DNA metabarcoding, botanical survey, ecological monitoring, grassland biodiversity, species composition,
• Publikační typ
• časopisecké články MeSH