Find in Library

Accurate species identification is key to conservation and phylogenetic inference. Living plant collections from botanical gardens/arboretum are important resources for the purpose of scientific research, but the proportion of cultivated plant misidentification are un-tested using DNA barcodes. Here, we assembled the next-generation barcode (complete plastid genome and complete nrDNA cistron) and mitochondrial genes from genome skimming data of <i>Torreya</i> species with multiple accessions for each species to test the species discrimination and the misidentification proportion of cultivated plants used in <i>Torreya</i> studies. A total of 38 accessions were included for analyses, representing all nine recognized species of genus <i>Torreya</i>. The plastid phylogeny showed that all 21 wild samples formed species-specific clades, except <i>T. jiulongshanensis</i>. Disregarding this putative hybrid, seven recognized species sampled here were successfully discriminated by the plastid genome. Only the <i>T. nucifera</i> accessions grouped into two grades. The species identification rate of the nrDNA cistron was 62.5%. The Skmer analysis based on nuclear reads from genome skims showed promise for species identification with seven species discriminated. The proportion of misidentified cultivated plants from arboreta/botanical gardens was relatively high with four accessions (23.5%) representing three species. Interspecific relationships within <i>Torreya</i> were fully resolved with maximum support by plastomes, where <i>Torreya jackii</i> was on the earliest diverging branch, though sister to <i>T. grandis</i> in the nrDNA cistron tree, suggesting that this is likely a hybrid species between <i>T</i>. <i>grandis</i> and an extinct <i>Torreya</i> ancestor lineage. The findings here provide quantitative insights into the usage of cultivated samples for phylogenetic study.