Global resource depletion and climate change pressures are intensifying, endangering species and leading to broad changes in plant community composition, diversity, and distribution. Monitoring plant communities via botanical surveys is often reliable but can be time-consuming and requires taxonomic expertise. Here we explore DNA metabarcoding as a faster, scalable alternative, comparing metabarcoding with quadrat-based botanical surveys to assess species composition in experimental grassland plots under various defoliation management regimes (continuous grazing, rotational grazing, frequent cutting, conservation cutting). Botanical surveys identified 16 taxa, while DNA metabarcoding using ITS2 and rbcL markers identified 25 taxa, with both methods detecting dominant species such as Holcus lanatus and Lolium perenne. There were significant differences in plant composition between defoliation treatments, with higher species diversity in cut versus grazed plots across both methods.  Despite some discrepancies, such as DNA metabarcoding identifying more taxa at the genus level (e.g., Agrostis spp. instead of Agrostis capillaris) and potential misidentifications (e.g., Cardaminopsis helleri vs. Cardamine flexuosa), both methods provided comparable results with DNA metabarcoding offering a broader, albeit semi-quantitative, assessment of species presence. The findings support integration of DNA metabarcoding into biodiversity monitoring alongside traditional methods. Further refinement of bioinformatics tools and reference databases will enhance the utility of metabarcoding in ecological studies, contributing to more efficient monitoring of grassland biodiversity and management practices. This study demonstrates the potential of DNA metabarcoding as a complementary tool for traditional botanical surveys, offering insights into plant community composition under varying management regimes.

Plant DNA extracted with DNeasy 96 Plant Kit (Qiagen), followed by a two-step PCR to prepare indexed Illumina MiSeq libraries for ITS2 and rbcL barcode regions. Amplifications were conducted separately for each marker using specific primers and Phusion Hot Start II Master Mix. Products were verified on agarose gel, purified with AMPure XP beads, indexed, and pooled equimolarly. Libraries were sequenced on an Illumina MiSeq (2x300 bp, v3 kit) at the Centre for Genomic Research, University of Liverpool, with negative controls included throughout to check for contamination.