A high-throughput belowground plant diversity assay using next-generation sequencing of the trnL intron

Aims We refine and test a next-generation sequencing assay for the molecular identification and quantification of plant roots from mixed-species samples. Methods We modified primers targeting the trnL intron to provide greater taxonomic resolution and developed an improved bioinformatics pipeline that can identify roots based on global, site-, and plot-specific taxon lists. We tested our primers and pipeline on mock plant communities of known composition andmixed-species samples of roots collected from temperate grassland and high Arctic tundra communities. Results We retrieved a high correlation (0.72) between observed and expected community dissimilarities. We found positive linear relationships between above-ground and belowground species richness in the grassland community, with higher correlations for plot-specific reference databases (R = 0.70) than a global reference database (R = 0.48). This highlights the importance of local reference databases within the bioinformatics pipeline. Lower root than aboveground richness suggests that typical root lateral spread in this grassland is less than 25 cm. We observed lateral rooting extents of up to 40 cm and rooting depths of up to 30 cm for six high Arctic species. Conclusions Testing on mock community, grassland, and tundra samples demonstrated that next-generation sequencing using our improved trnL primers and pipeline can successfully recover plant root community structure.