Closely-related taxa influence woody species discrimination via DNA barcoding: evidence from global forest dynamics plots

To determine how well DNA barcodes from the chloroplast region perform in forest dynamics plots (FDPs) from global CTFS-ForestGEO network, we analyzed DNA barcoding sequences of 1277 plant species from a wide phylogenetic range (3 FDPs in tropics, 5 in subtropics and 5 in temperate zone) and compared the rates of species discrimination (RSD). We quantified RSD by two DNA barcode combinations (rbcL + matK and rbcL + matK +trnH-psbA) using a monophyly-based method (GARLI). We defined two indexes of closely-related taxa (Gm/Gt and S/G ratios) and correlated these ratios with RSD. The combination of rbcL + matK averagely discriminated 88.65%, 83.84% and 72.51% at the local, regional and global scales, respectively. An additional locus trnH-psbA increased RSD by 2.87%, 1.49% and 3.58% correspondingly. RSD varied along a latitudinal gradient and were negatively correlated with ratios of closely-related taxa. Successes of species discrimination generally depend on scales in global FDPs. We suggested that the combination of rbcL + matK + trnH-psbA is currently applicable for DNA barcoding-based phylogenetic studies on forest communities.

Nancai Pei, David L. Erickson, Bufeng Chen, Xuejun Ge, Xiangcheng Mi, Nathan G. Swenson Jin-Long Zhang, Frank A. Jones, Chun-Lin Huang, Wanhui Ye, Zhanqing Hao, Chang-Fu Hsieh, Shawn Lum, Norman A. Bourg, John D. Parker, Jess K. Zimmerman, William J. McShea, Ida C. Lopez, I-Fang Sun, Stuart J. Davies, Keping Ma, & W. John Kress
Scientific Reports