Application Of DNA Barcode Candidate Sequences In The Identification Of Potentilla Species

Potentilla genus is a plant that highly diverse and complicated,and the classification history of this genus is extremely complex.The complex taxonomy and limited morphological characteristics of this genus make species identification difficult.Therefore,Potentilla is an ideal group for testing plant DNA barcodes as it presents a taxonomic impediment in which many closely related species in the genus are not easy to be identified using conventional methods.In this study,we used the four DNA regions(rbcL,matK,psb A-trnH and ITS)as the barcodes for identifying 42 individuals,22 species(including below specific taxa)in the genus Potentilla.The main results are as follows:1.Among the species of Potentilla,rbcL and matK fragments were the best sequences that successfully amplified all species and the sequence assemble rate were all close to 100%.The rate of amplification and assemble of the psbA-trnH sequence was 100% and 95%.In contrast,the ITS sequence success rate is 45% and sequence quality were extremly lower than others,considering the accuracy of the data,this fragment was not used in subsequent experiments.2.After the analysis of the sequence characteristics,it was found that the higher conservative sequence of Potentilla is rbcL,and the highest variation sequence is psbaA-trnH.Comparing the GC% content of the three sequences,rbcL fragment was have the highest GC% content,about 43.3%,and the lowest was the psbA-trnH fragment,which was 25.6%.3.Among the three sequence of the Potentilla DNA barcode candidate,the highest average intra-interspecies genetic distance was the psbA-trnH sequence,followed by matK,and the smallest was the rbcL sequence.It shows that psbaA-trnH has the highest variation,rbcL is the mostconservative.4.We did not find any distinct barcoding gaps in the distribution of intra-versus interspecific variability of mat K,rbcL and psbA-trnH.5.Used TaxonDNA methods to test the the three separate DNA barcode candidate sequence and the four combinations sequence(rbcL +psbaA-trnH,mat K + psbaA-trnH,rbcL + matK,and rbcL + matK +psbaA-trnH).Found that rbcL had the highest recognition rate,the psbaA-trnH had a lowest recognition rate and about 19(28.78%)sequence at psbaA-trnH showed no matching information.The rbcL+psbA-trnH combination sequence had highest best match accuracy among the four combined sequence.The accurate identification rate of rbcL fragments was the highest.6.In this study,the NJ tree and the ML tree of three separate sequence and four combinations sequence of Potentilla are constructed.The NJ trees are more accurate than ML trees,and the species support rate of ML trees is generally higher than that of NJ trees.The ML tree of the matK fragment in the individual fragments is the best,and the identification of rbcL+matK+psbaA-trnH sequence is the best in the combination fragment.The Potentilla species in all trees distinguished well from the species of out group.All trees clustering Sect.Closterostylae,Leptostylae,Nematostylae,Potentilla,Rhopalostylae in Potentilla genus well.The phylogenetic tree of six individual fragments failed to achieve the desired effect of distinguishing all species.Individuals of the same species of Sect Conostylae and Potentilla acaulis failed to cluster on one branch,and parallel branches appeared.So combined the three segments and construct phylogenetic tree.The support rate of the ML tree is higher,but neither the single sequence nor the combined sequences distinguish the same species of Sect Conostylae and Potentilla acaulis so well.