Abstract: Genome skimming has dramatically extended DNA barcoding from short DNA fragments to next generation barcodes in plants. However, conserved DNA barcoding markers, including complete plastid genome and nuclear ribosomal DNA (nrDNA) sequences, are inadequate for accurate species identification. Skmer, a recently proposed approach that estimates genetic distances among species based on unassembled genome skims, has been proposed to effectively improve species discrimination rate. In this study, we used Skmer to identify species based on genomic skims of 47 individuals representing 10 out of 13 species of Schima (Theaceae) from China. The unassembled reads identified six species, with a species identification rate of 60%, twice as high as previous efforts that used plastid genomes (27.27%). In addition, Skmer was able to identify Schima species with only 0.5×sequencing depth, as six species were well-supported with unassembled data sizes as small as 0.5 Gb. These findings demonstrate the potential for Skmer approach in species identification, where nuclear genomic data plays a crucial role. For taxonomically difficult taxa such as Schima, which have diverged recently and have low levels of genetic variation, Skmer is a promising alternative to next generation barcodes.

Key words: Schima, Genome skimming, Species discrimination, Skmer