Expanding and testing fluorescent amplified fragment length polymorphisms for identifying roots of boreal forest plant species

PREMISE OF THE STUDY: Identifying roots to species is challenging, but is a common problem in ecology. Fluorescent amplified fragment length polymorphisms (FAFLPs) can distinguish species within a mixed sample, are high throughput, and are inexpensive. To broaden the use of this tool across ecosystems, unique size profiles must be established for species, and its limits identified. METHODS: Fragments of three noncoding cpDNA regions were used to create size profiles for 193 species common to the western Canadian boreal forest. We compared detection success among congeners using FAFLPs and Sanger sequencing of the trnL intron. We also simulated and experimentally created communities to test the influence of species richness, cpDNA regions used, and extraction/amplification biases on detection success. RESULTS: Of the 193 species, 54% had unique size profiles. This value decreased when fewer cpDNA regions were used. In simulated communities, ambiguous species identifications were positively related to the species richness of the community. In mock communities, some species evaded detection owing to poor extraction or amplification. Sequencing did not increase detection success compared to FAFLPs for a subset of 24 species across nine genera. DISCUSSION: We recommend FAFLPs are best suited to confirm rather than discover species occurring belowground.