DNA metabarcoding of biocrust lichen-forming fungi allows interpretation of ecological gradients

Biocrusts dominated by bryophytes and lichens perform important functions in dryland ecosystems but ecological gradients affecting these communities are time and cost prohibitive over broad scales using current monitoring techniques. We explored DNA metabarcoding as a potential tool for monitoring biocrust lichen communities at a site in Washington (USA) that had already been surveyed for lichen diversity and community composition. We developed a DNA reference library using the internal transcribed spacer (ITS) region from specimens collected in the study area. We also visually estimated the abundance of lichen species or groups in 22 plots spanning a range of wildfire history and invasion by the exotic annual grass B. tectorum. At these plots, we collected bulk lichen community samples for metabarcoding using two approaches: small sample dishes and combined tweezer fragments from all species observed. We used PacBio sequencing to simultaneously generate ITS barcode sequences for all lichen-forming fungi (LFF) present in the bulk samples, clustering similar sequences into operational taxonomic units (OTUs). Lichen communities detected visually and using DNA metabarcoding both captured a reduction of biocrust diversity and change in community composition related to the abundance of B. tectorum, suggesting that metabarcoding was able to identify the same dominant ecological pattern in biocrust lichens as visual sampling. The tweezer sampling approach captured on average 12.3 more OTUs than the dish approach and some taxa were more consistently detected by one approach or the other. After using the specimen library to identify species associated with LFF OTUs, we were able to determine that metabarcoding and visual sampling detected overlapping but different lichen species. Metabarcoding failed to detect common taxa that were visually observed and collected, including: Arthonia glebosa, Candelariella spp., Enchylium tenax, Lecanora muralis, Lecidella spp., Leptochidium albociliatum, Massalongia carnosa, and Trapeliopsis glaucopholis. However, metabarcoding detected OTUs not visually observed in the genera Elixia, Lecanora, Lecanoropsis, Bacidina, Pyrenodesmia, Xanthocarpia, Trapelia, Verrucaria, and Verruculopsis. Furthermore, metabarcoding allowed us to identify more OTU diversity than expected within Diploschistes muscorum, Trapeliopsis bisorediata, and Trapeliopsis steppica. While combining the two methods allowed the most comprehensive interpretation of biodiversity at the site, our results suggested that metabarcoding could be a fruitful approach for monitoring broad-scale ecological gradients affecting biocrust lichen communities.