Skip to main content

The DNA barcode identification of Dalbergia odorifera T. Chen and Dalbergia tonkinensis Prain

Abstract

Background

Dalbergia odorifera is a precious tree species with unique economic and medicinal values, which is difficult to distinguish from Dalbergia tonkinensis by traditional identification methods such as morphological characteristics and wood structure characteristics. It has been demonstrated that the identification of tree species can be effectively achieved using DNA barcoding, but there is a lack of study of the combined sequences used as DNA barcodes in the two tree species. In this study, 10 single sequences and 4 combined sequences were selected for analysis, and the identification effect of each sequence was evaluated by the distance-based method, BLAST-based search, character-based method, and tree-based method.

Results

Among the single sequences and the combined sequences, the interspecies distance of trnH-psbA and ITS2 + trnH-psbA was greater than the intraspecies distance, and there was no overlap in their frequency distribution plots. The results of the Wilcoxon signed-rank test for the interspecies distance of each sequence showed that the interspecies differences of the single sequences except trnL-trnF, trnH-psbA, and ycf3 were significantly smaller than those of the combined sequences. The results of BLAST analysis showed that trnH-psbA could accurately identify D. odorifera and D. tonkinensis at the species level. In the character-based method, single sequences of trnL-trnF, trnH-psbA with all the combined sequences can be used for the identification of D. odorifera and D. tonkinensis. In addition, the neighbor-joining (NJ) trees constructed based on trnH-psbA and ITS2 + trnH-psbA were able to cluster D. odorifera and D. tonkinensis on two clades.

Conclusions

The results showed that the character-based method with the BLOG algorithm was the most effective among all the evaluation methods, and the combined sequences can improve the ability to identify tree species compared with single sequences. Finally, the trnH-psbA and ITS2 + trnH-psbA were proposed as DNA barcodes to identify D. odorifera and D. tonkinensis.

Peer Review reports

Introduction

Dalbergia odorifera T. Chen is a semi-deciduous tree of the Leguminosae [1], which is endemic to Hainan Province, China, and has been introduced to Guangxi, Guangdong, and Yunnan [2]. Its heartwood has a beautiful color, excellent quality, and a unique aroma, which is often used in the production of valuable furniture [3]. In addition, D. odorifera also has various medicinal values, and modern pharmacological studies have shown that its ingredients have anti-inflammatory [4, 5], antioxidant [6, 7], blood-activating [8], anti-tumor [9, 10], anti-bacterial [11] and cardiovascular disease treatment effects [12, 13], etc. Due to its unique economic and medicinal value, the wild resources of D. odorifera have been severely damaged in recent years, and it was listed on the Red List of Threatened Species by the International Union for Conservation of Nature (IUCN) in 1998 [14].

In recent years, researchers have found that Dalbergia tonkinensis Prain from Vietnam is very similar to D. odorifera in terms of wood color, density, and even structural characteristics, but the two types of wood are widely disparate in price in the market [15, 16]. And unlike D. odorifera, the heartwood extract of D. tonkinensis has been shown to have anti-bacterial [17], anti-diabetic [18] and anti-hypertensive [19] properties. Due to the disparity in their economic and medicinal values, finding methods that can effectively identify them becomes especially important. It has been shown that they are difficult to distinguish by traditional methods such as morphological characteristics of the trees and structural characteristics of the wood [20, 21], and the chemical analysis results also show no significant differentiation in their chemical composition [22]. Therefore, it is difficult to distinguish D. odorifera from D. tonkinensis by the three methods mentioned above, and all of these methods require the extensive expertise of the evaluators.

DNA barcoding technology has shown powerful advantages in species identification, which can identify different species quickly and accurately, effectively solving the problem of relying on traditional methods that require a high level of evaluators. Currently, chloroplast genomic regions (such as rbcL, matK, trnH-psbA), as well as the internal transcribed spacer region (ITS) of nuclear ribosomal DNA, have become candidates for plant DNA barcoding [23, 24]. Some researchers used the sequences of Dalbergia spp. downloaded from the National Center of Biotechnology Information (NCBI) database for analysis and selected the ITS + matK + rbcL region as a DNA barcode to identify the species of this genus [25]; In addition, researchers have proposed different single sequences or combined sequences as DNA barcodes for the identification of Dalbergia species from different regions, including ITS2 + trnH-psbA [26], matK and matK + rbcL [27], ITS [28]. Highly variable regions in the chloroplast genome have also been shown to be useful for the identification of the trees of this genus [29,30,31]. In the DNA barcoding studies of D. odorifera and D. tonkinensis, some researchers suggested the trnH-psbA region as the DNA barcode to distinguish them [15], but other researchers have argued that the variable loci in this region are located in the palindromic region, which is not conserved among individuals making it unusable as a DNA barcode, thus they recommended the eight regions selected from the chloroplast genome as DNA barcodes for the two tree species [32]. The above analysis shows that the results of DNA barcoding studies on Dalbergia species, especially D. odorifera and D. tonkinensis, are still not consistent, and the combined sequences as DNA barcodes have not attracted much attention in the studies of these two species.

In this study, 10 single sequences (ITS2, rpoB, rpoC1, trnH-psbA, trnL-trnF, matK, ycf3, trnL intron, trnS-psbC, rbcL) and 4 combined sequences were analyzed by various analysis methods. The purpose was to compare different analysis methods and to investigate whether the combination of sequences could improve the identification abilities of D. odorifera and D. tonkinensis, and finally to select the optimal DNA barcode region for the identification of the two species.

Materials and methods

Plant materials

A total of 38 plant leaf materials were collected from China and Vietnam. Among them, all 16 D. odorifera materials were collected from southern China, and 16 of the D. tonkinensis materials were collected from Vietnam and 6 from China. The sampling information is shown in Table 1.

Among the leaf materials of D. odorifera, H1-5 were obtained from the Experimental Center of Tropical Forestry Chinese Academy of Forestry in Pingxiang, Guangxi, H6-10 from Nanning Arboretum in Nanning, Guangxi, H11-13 from Fengmu Forestry Farm in Fengmu Town, Tunchang County, Hainan, and H14-16 from Haikou People’s Park in Haikou, Hainan. Among the materials of D. tonkinensis, Y1 was sourced from the Qinzhou Port Wood Storage Plant in Qinzhou, Guangxi, Y2-7 from the Hanoi Botanical Garden in Hanoi, Vietnam, Y8-12 from street trees in Nanshan Town, Pingxiang, Guangxi, Y13-18 from street trees in Tongji Village, Bac Ninh, Vietnam, and Y19-22 from the experimental forestry farm of Vietnam Forestry University in Quang Ninh, Vietnam. All plant materials have been formally identified by Professor Xu Feng (Guangxi University Forest Products Quality Testing Center). The voucher specimens are preserved in the Wood Herbarium of the College of Forestry, Guangxi University, and the specimen information is shown in Table S1. All the collected materials require no specific permission or license.

Table 1 Information of the plant samples

DNA extraction, amplification, and sequencing

DNA was extracted using the DNeasy Plant Mini Kit (Qiagen), and the extracted DNA was stored at -20℃. Ten DNA sequences (ITS2, rpoB, rpoC1, trnH-psbA, trnL-trnF, matK, ycf3, trnL intron, trnS-psbC, rbcL) were selected as candidate barcodes for primer design, and the sequences and primer information are shown in Table S2. PCR amplification was performed in a 25 µL system, and the amplified products were first detected by 1% agarose gel electrophoresis, then purified by Biospin PCR product purification kit, and the purified products were sequenced by Beijing Liuhe Huada Gene Technology Co., Ltd (China). The GeneBank accession numbers of each sequence are shown in Table S3.

Data analysis

Multiple comparisons of sequences were performed using MEGA 11 software, and the basic information of each sequence was counted after manual adjustment, after which the sequences with differences were selected for further analysis. In addition, four combinations of trnL-trnF + ITS2, trnL-trnF + trnH-psbA, ITS2 + trnH-psbA, and trnL-trnF + ITS2 + trnH-psbA, which were selected based on the discrimination ability of single sequences in this study, were used as candidate DNA barcodes together for subsequent identification analysis.

The ideal DNA barcode should have low intraspecific and high interspecific variations [33]. The Kimura two-parameter (K2P) model in MEGA 11 software was used to calculate interspecific and intraspecific genetic distances. Interspecific genetic variation was assessed by the average distance, and intraspecific variation was assessed by three parameters: average intraspecific distance, theta (θ), and average coalescence depth. Subsequently, the barcoding gap was assessed based on the genetic distance distribution plots of inter- and intraspecies distances. The sequences that can be used as DNA barcodes should exhibit independent, non-overlapping distributions of genetic variation in intra- and interspecific samples [34]. At last, the Wilcoxon signed-rank test was performed by IBM SPSS Statistics 27 to confirm whether there were significant interspecies differences among the different sequences.

The identification ability of each sequence was further evaluated by the BLAST-based method, character-based method, and tree-based method. All sequences were aligned in the NCBI database by BLAST sequence similarity search, and the best match was selected as the identification results with E value < cut-off value (E value < 10− 6) according to Ross et al. [35]. Barcoding with LOGic is an analysis method based on sequence characteristics, and in this study, the BLOG 2.0 software was used to evaluate the discrimination rate of different sequences and to obtain logical rules for the discrimination of each sequence. The MEGA 11 software was used to construct the neighbor-joining (NJ) trees, the bootstrap supporting option was set to 1000 random addition replicates to determine statistical support for the clades. When all individuals of the same species can be clustered in a single clade, the species is considered to be successfully identified.

Results

Sequence features

Multiple alignment was performed for all sequences and the information of each sequence is shown in Table 2. The length of single sequences ranged from 256 bp (trnH-psbA) to 866 bp (matK), and there were no variable sites in rpoB, rpoC1, and rbcL, indicating that these three sequences did not apply to the identification of D. odorifera and D. tonkinensis. Among the other sequences, trnL-trnF had the highest proportion of informative sites (66/409 bp), followed by trnH-psbA (7/256 bp), ycf3 (18/690 bp), ITS2 (11/558), trnS-psbC (4/347), and trnL intron (2/519), and the lowest is matK (1/866 bp). The combined sequences ranged from 665 bp (trnL-trnF + trnH-psbA) to 1223 bp (trnL-trnF + ITS2 + trnH-psbA) in length, and the order of the proportion of informative sites for each sequence was trnL-trnF + trnH-psbA (70/665), trnL-trnF + ITS2 (69/967), trnL-trnF + ITS2 + trnH-psbA (73/1223), and ITS2 + trnH-psbA (18/814).

Table 2 Statistics of sequence characteristics

Distance analysis and barcoding gap assessment

The statistics of the K2P distances of single sequences and combined sequences are shown in Table 3. The interspecific distances ranged from 0.0005 to 0.0828 and intraspecific distances ranged from 0.0000 to 0.0652 in the single sequences, and the inter- and intraspecific distances can be used to assess the genetic variation of the sequences. Among them, trnL-trnF had the highest interspecific and intraspecific genetic variation, trnL intron had the lowest interspecific genetic variation, and trnH-psbA had the lowest intraspecific genetic variation. In addition, the average interspecific distances of all sequences except ycf3 and trnL intron were higher than the average intraspecific distances. The interspecies distances of the combined sequences ranged from 0.0137 to 0.0626 and intraspecies distances from 0.0036 to 0.0387. Among them, trnL-trnF + trnH-psbA had the highest inter- and intraspecific genetic variation, and ITS2 + trnH-psbA had the lowest inter- and intraspecific variation.

The frequency distributions of inter- and intraspecific distances for single and combined sequences are shown in Fig. 1. The results indicated that the interspecific variation of trnH-psbA was greater than the intraspecific variation, forming a clear gap (Fig. 1-D), while the distribution of inter- and intraspecific distances of the rest of the single sequences showed some overlaps. Among the combined sequences, the distribution of inter- and intraspecific distances did not overlap in ITS2 + trnH-psbA (Fig. 1-J) and overlapped in the rest of the sequences.

Table 3 Interspecific and intraspecific distance analysis of sequences based on the Kimura two-parameter model
Fig. 1
figure 1

Distribution of interspecific and intraspecific distances of sequences (A: ITS2; B: matK; C: trnS-psbC; D: trnH-psbA; E: trnL-trnF; F: trnL intron; G: ycf3; H: trnL-trnF + ITS2; I: trnL-trnF + trnH-psbA; J: ITS2 + trnH-psbA; K: trnL-trnF + ITS2 + trnH-psbA)

Wilcoxon signed-rank test analysis

The interspecies distances of single and combined sequences were analyzed with the Wilcoxon signed-rank test and the results are shown in Tables 4 and 5. The interspecific distances of all sequences in the single sequences were significantly different, in the order of trnL-trnF > trnH-psbA > ycf3 > ITS2 > trnS-psbC > matK > trnL intron. The test results for the combined sequences were trnL-trnF + trnH-psbA > trnL-trnF + ITS2, trnL-trnF + ITS2 + trnH-psbA > ITS2 + trnH-psbA. And trnL-trnF + ITS2 and trnL-trnF + ITS2 + trnH-psbA could not be compared as there was no significant difference (p = 0.9000) between them. In addition, the interspecies distances of single and combined sequences were also tested in this study (Table S4). The results showed that among the single sequences, the interspecies differences of trnL-trnF were higher than all the combined sequences, trnH-psbA > ITS2 + trnH-psbA, and no significant differences existed between ycf3 and ITS2 + trnH-psbA (p = 0.295) thus cannot be compared. Except for these, the interspecies differences of all the single sequences were significantly lower than the combined sequences.

Table 4 Wilcoxon signed-rank test for the interspecies distances of the single sequences
Table 5 Wilcoxon signed-rank test for the interspecies distances of the combined sequences

BLAST search-based analysis

The identification rates of different sequences were compared based on the BLAST search. For single sequences, the identification rates of ITS2, matK, trnH-psbA, and trnL intron were in the range of 99-100%. Besides, among the trnS-psbC sequences, 3 sequences were identified at less than 99%, and the other 26 sequences were identified at 99-100%; among the sequences of trnL-trnF, 5 sequences were identified at 90-95%, 2 sequences were identified at less than 90%, and the rest 7 sequences were identified at 100%; and among the sequences of ycf3, 9 sequences were identified at less than 99% and 6 sequences were identified at 99-100%. The identification rates were 99-100% for all of the combined sequences.

The analysis of the sequence identification results is shown in Table 6. The identification results at the genus level showed that the identification success rate of matK was 0%, the identification success rate of trnL-trnF was 42.9%, and the identification error rate was 57.1%, all the remaining sequences were 100% in the identification success rate. At the species level, the identification success rate of trnH-psbA was 100%, which means that this sequence can achieve accurate identification at the species level. Among the other single sequences, the identification success rate of ITS2 was 67.6%, the error rate was 24.3% and the ambiguous identification rate was 8.1%; the identification success rate of ycf3 was 53.3%, the error rate was 26.7% and the ambiguous identification rate was 20%, while the identification success rates of matK, trnS-psbC, trnL-trnF, and trnL intron were all 0%. Among the combined sequences, the highest identification success rate was trnL-trnF + ITS2 + trnH-psbA with 76.9%, and the others were followed by trnL-trnF + ITS2 (71.4%), ITS2 + trnH-psbA (69.7%), and trnL-trnF + trnH-psbA (30.8%). The above results show that among all sequences, trnH-psbA can accurately distinguish D. odorifera and D. tonkinensis by the BLAST-based method.

Table 6 BLAST analysis of sequences at the genus and species level

Character-based analysis of sequences

The identification rate and logic formulae of each sequence based on the BLOG algorithm are shown in Table 7. The correct classification rate of trnL-trnF, trnH-psbA, and all the combined sequences for D. odorifera and D. tonkinensis was 100%. Based on trnL-trnF, if locus 243 was A and locus 375 was T, then the sequence belonged to D. odorifera, and if locus 322 was C and locus 373 was C, then the sequence belonged to D. tonkinensis. Based on trnH-psbA, if locus 196 was A, it belonged to D. odorifera, and if the locus was C, it belonged to D. tonkinensis. The logic formula for the identification of trnL-trnF + ITS2 in the combined sequence was the summary of these two single sequences, when the sequence with T at locus 375 and A at locus 412, it belonged to D. odorifera, and with C at locus 322 and C at locus 373, it belonged to D. tonkinensis. The logic formulae for the identification of the other three combined sequences were consistent with trnH-psbA. This result shows that trnL-trnF, trnH-psbA, and all the combined sequences can accurately identify D. odorifera and D. tonkinensis by the character-based method.

Table 7 Character-based approach for species identification

Tree-based analysis

The NJ trees were constructed separately with single and combined sequences. The results showed that the NJ tree constructed based on trnH-psbA was able to separate D. odorifera and D. tonkinensis (Fig. 2-A). The samples of these two tree species were clustered on two clades with a support value of 100%. Among the other sequences, the construction of NJ trees based on trnL-trnF with the greatest interspecific variation revealed that Y1, Y4, and Y5 in D. tonkinensis were able to cluster on the same clade with 88% support, but Y6 and Y7 clustered in a clade with the samples of D. odorifera (Fig. S1). It showed that the NJ tree constructed by trnL-trnF could not identify D. odorifera and D. tonkinensis.

Among the combined sequences, ITS2 + trnH-psbA was able to be used for the identification of D. odorifera and D. tonkinensis (Fig. 2-B). In the NJ tree constructed based on this sequence, the samples of D. odorifera and D. tonkinensis clustered in two clades, and the clade clustered by the samples of D. odorifera had 100% support. The NJ tree constructed based on trnL-trnF + ITS2 showed similar results to that of trnL-trnF, with Y1, Y4, and Y5 in D. tonkinensis able to cluster in the same clade with 83% support, while Y6 and Y7 clustered with the samples of D. odorifera (Fig. S2). In the NJ trees constructed based on trnL-trnF + trnH-psbA and trnL-trnF + trnH-psbA + ITS2, Y6 and Y7 in D. tonkinensis were also clustered in the same clade as the samples of D. odorifera, but all the samples of D. odorifera were able to be clustered together with more than 90% support (Fig. S3, S4). Therefore, none of these three combined sequences could be used for the identification of D. odorifera and D. tonkinensis.

The above results showed that trnH-psbA in single sequences and ITS2 + trnH-psbA in combined sequences could identify D. odorifera and D. tonkinensis by the method of constructing NJ trees.

Fig. 2
figure 2

NJ trees constructed based on trnH-psbA and ITS2 + trnH-psbA (A: trnH-psbA; B: ITS2 + trnH-psbA)

Discussions

Comparison of DNA barcode analysis methods

In the current study, the main DNA barcode analysis methods include distance-based method, BLAST-based method, character-based method, and tree-based method. In addition, there are other methods that use DNA barcoding software, such as best match (BM) and best close match (BCM) analysis in TAXONDNA, to evaluate the identification effect [36]. All of the above analysis methods have been applied in the study of Dalbergia species, but the character-based method has not been reported in the studies of D. odorifera and D. tonkinensis. However, these analysis methods have not been evaluated either in Dalbergia species or in studies of these two tree species, making it difficult to find the most convenient and effective identification method for practical application. Therefore, in this study, the character-based method via BLOG was added to the methods of the distance-based method, BLAST-based method, and tree-based method, and the role of these four analysis methods in the selection of DNA barcodes in D. odorifera and D. tonkinensis was also evaluated.

The distance-based method is commonly used to assess interspecific and intraspecific variation in sequences and use their distribution to determine the barcoding gap. And some researchers will compare the interspecies variation of different sequences by the Wilcoxon signed-rank test for further analysis [27]. Combining the genetic distance distribution of each sequence with the results of the Wilcoxon signed-rank test, trnH-psbA and ITS2 + trnH-psbA were recommended as DNA barcodes of D. odorifera and D. tonkinensis. However, this method could not evaluate the identification effect of barcodes.

BLAST is an analysis method based on the similarity comparison of sequences. Ross et al. used four different algorithms to evaluate the effect of species identification based on the BLAST method, with the BLAST1 algorithm having the highest species identification rate [35]. This study also used the algorithm of BLAST1 to evaluate the identification effect of DNA barcodes. Among all sequences, all 9 sequences except matK and trnL-trnF could be accurately compared to the species of Dalbergia, but only trnH-psbA could identify D. odorifera and D. tonkinensis. It shows that the BLAST-based method has a high identification rate at the genus level, but it is difficult to achieve accurate identification at the species level, and this conclusion is consistent with the results of previous studies [36, 37].

The character-based method is used to identify species by the presence or absence of discrete nucleotide substitutions (character states) in a sequence [38]. This analysis method can be implemented by the Characteristic Attributes Organization System (CAOS) algorithm [38] or the BLOG algorithm [39]. This method was used for DNA barcode selection in the study of Dalbergia species [40]. In this study, the DNA barcodes of D. odorifera and D. tonkinensis were analyzed by the BLOG algorithm, in which the accurate identification rate of ITS2 was over 80%, while trnL-trnF, trnH-psbA, and four combined sequences were able to accurately identify these two species.

Methods based on the reconstruction of phylogenetic relationships usually use methods such as NJ, MP, or UPGMA to construct phylogenetic trees, and the clustering of each species in the tree is used as the identification basis for the evaluation of DNA barcodes. However, some researchers consider that the evolutionary relationships of species based on phylogenetic trees constructed with DNA barcodes are unreliable because the sequences used as DNA barcodes are usually short [41]. In addition, some researchers have argued that using this method in the classification of paraphyletic groups or incomplete lineages may produce incorrect or ambiguous identification results [42]. The results of this study showed that trnH-psbA, ITS2 + trnH-psbA were able to cluster samples of D. odorifera and D. tonkinensis in two clades by constructing NJ trees, indicating that this method can be used for DNA barcode evaluation in the identification of these two tree species.

In summary, the distance-based method cannot be used to directly assess the identification effect of DNA barcodes, but it can be used as one of the bases for selection. The BLAST-based method is suitable for identification at the genus level but cannot be used for identification at the species level. Both character-based and tree-based methods can evaluate the identification effect of DNA barcodes. In this study, the sequences showed the best identification effect using the character-based method with the BLOG algorithm.

Comparison of the discriminatory ability of single sequences and combined sequences

Based on current studies of genome structure, nucleotide substitution rates, and variation in rate distribution, it is difficult to find a DNA barcode that is suitable for all plants [43]. Since Kress proposed the idea of sequence combination, a growing number of studies have demonstrated that combined sequences have higher discriminatory power in species than single sequences [23]. In the study of Dalbergia species, Yu et al. analyzed 8 barcode sequences and their combinations of 9 species of this genus and found that the combined sequences had higher identification power than any single sequence [26]. Similarly, other researchers also proposed the combined sequence as the DNA barcodes for this genus [25, 40, 44]. However, the combination of sequences as DNA barcodes did not attract much attention in the studies of D. odorifera and D. tonkinensis, and the researchers tested only the identification abilities of single sequences in the studies of the two species [15]. In this study, in addition to evaluating the single sequences, three sequences, trnL-trnF, ITS2, and trnH-psbA, were selected for combination based on the discrimination ability of the single sequences, and the four combined sequences were evaluated for their discrimination abilities in D. odorifera and D. tonkinensis.

Seven single sequences and four combined sequences of D. odorifera and D. tonkinensis were analyzed in this study. The results of the analysis based on the distance method showed that the mean interspecies distances of the combined sequences were higher than those of the other single sequences except for trnL-trnF and trnH-psbA, and the results of the Wilcoxon signed-rank test also supported this conclusion. However, the results of the genetic distance distribution showed that only the inter- and intraspecific distances of ITS2 + trnH-psbA in the combined sequences did not show overlapping regions. The results of BLAST-based analysis showed that the combined sequences could improve the identification success rate to a certain extent compared with single sequences, but the method could not be used for identification at the species level. The analysis of the character-based method showed that only two of the single sequences were able to accurately identify D. odorifera and D. tonkinensis, while all combinations of sequences were able to achieve this goal. According to the logical formulas given by the BLOG algorithm, the logical formulas of the three combined sequences were consistent with trnH-psbA in the single sequences, indicating that the sequence trnH-psbA may still play a major role in these combined sequences. But this still indicates a significant improvement in the identification ability of the combined sequences compared to the single sequences. However, this conclusion was not supported by the results of the tree-based analysis. In addition to the trnH-psbA sequence in the single sequence, only the NJ tree constructed based on ITS2 + trnH-psbA was observed to cluster the samples of D. odorifera and D. tonkinensis on two clades among all combined sequences, while the others could not accurately identify the two species.

Therefore, the results of this study indicate that the combined sequences have higher identification abilities for D. odorifera and D. tonkinensis compared to single sequences. However, not all of the combined sequences were able to accurately identify the two species, which was also related to the selection of the analysis methods.

Selection of DNA barcodes for  Dalbergia odorifera  and  D. tonkinensis

In previous studies, there was some controversy among different researchers about whether trnH-psbA could be used as the DNA barcode of D. odorifera and D. tonkinensis. Although Qin et al. suggested that the variable loci in trnH-psbA are located in the palindromic region which is not conserved in different individuals, they did not perform experiments to verify this inference [15, 32]. In this study, four analysis methods, including distance-based method, BLAST-based method, character-based method, and tree-based method, were used to evaluate whether the sequence trnH-psbA could be used as a DNA barcode for D. odorifera and D. tonkinensis. The chloroplast genome sequences such as rpoB, trnL intron, trnL-trnF, ycf3, and trnS-psbC, which have not been used in the study of D. odorifera and D. tonkinensis were added to this study, the aim was to select new DNA barcodes from these sequences that are suitable for distinguishing these two tree species.

In this study, based on the results of genetic distance distribution, the interspecific variation of trnH-psbA was greater than the intraspecific variation and formed a clear gap. The BLAST-based analysis showed that the sequence was able to accurately identify the two tree species at the species level. The results of the analysis based on sequence characteristics also showed that the accurate identification of these two species by this sequence was 100%. The samples of D. odorifera and D. tonkinensis were able to cluster in two different clades in the NJ tree constructed based on trnH-psbA. The experimental results obtained from multiple evaluation methods demonstrated the strong identification ability of trnH-psbA, providing powerful support for the conclusion that it can be used as a DNA barcode for these two species. Except for trnH-psbA, ITS2 + trnH-psbA was found to be able to distinguish D. odorifera and D. tonkinensis in all analyses other than those of the BLAST-based method in the study of combined sequences. Moreover, it was demonstrated in previous studies that this combined sequence can be used for the identification of Dalbergia species, which is consistent with the results of this study [26]. In summary, trnH-psbA and ITS2 + trnH-psbA are recommended as DNA barcodes for the identification of D. odorifera and D. tonkinensis.

In addition, in the studies of Dalbergia species, different researchers have proposed different sequences as DNA barcodes to distinguish the species of this genus, such as matK [27], ITS [28] in single sequences, and matK + rbcL [27], ITS + matK + rbcL [25] in combined sequences. In this study, the identification of some of the above sequences in D. odorifera and D. tonkinensis was verified. Among them, the sequencing results of rbcL showed that there were no variable sites in this sequence of D. odorifera and D. tonkinensis, and the lowest proportion of informative sites in matK, both of which and the related combined sequences could not be used as DNA barcodes for these two species, which was confirmed by the further analysis results of matK. The above results are consistent with the previous conclusion that these sequences cannot identify closely related tree species [32]. As for other candidate regions suitable for plants or wood identification, such as ycf3 and trnS-psbC, the analysis results of this study proved that they could not be used for the identification of D. odorifera and D. tonkinensis.

Conclusions

In this study, four methods were used to assess the identification abilities of 10 single sequences and 4 combined sequences for D. odorifera and D. tonkinensis. Among all methods, sequences showed the best identification effect using the character-based method with the BLOG algorithm. Moreover, compared with single sequences, the combined sequences can improve the identification abilities of tree species to a certain extent. In addition, the experimental results demonstrated that trnH-psbA in the single sequences were able to accurately distinguish D. odorifera and D. tonkinensis by the above four evaluation methods, while in the analysis of the combined sequences, ITS2 + trnH-psbA showed good identification ability. Both sequences were able to be used as DNA barcodes for these two species.

Data Availability

The original contributions presented in the study are included in the article/supplementary material, and the GeneBank accession numbers of each sequence are shown in Table S3. Further inquiries can be directed to the corresponding authors.

References

  1. Editorial committee of Chinese flora of Chinese Academy of Science. Flora of China. Beijing: Science Press; 1994. p. 114.

    Google Scholar 

  2. Zhao X, Wang C, Meng H, Yu Z, Yang M, Wei J. Dalbergia odorifera: a review of its traditional uses, phytochemistry, pharmacology, and quality control. J Ethnopharmacol. 2020;248:112328.

    Article  CAS  PubMed  Google Scholar 

  3. Wariss HM, Yi T, Wang H, Zhang R. Characterization of the complete chloroplast genome of Dalbergia odorifera (Leguminosae), a rare and critically endangered legume endemic to China. Conserv Genet Resour. 2018;10(3):527–30.

    Article  Google Scholar 

  4. Lee SH, Kim J, Seo GS, Kim YC, Sohn DH. Isoliquiritigenin, from Dalbergia odorifera, up-regulates anti-inflammatory heme oxygenase-1 expression in RAW264.7 macrophages. Inflamm Res. 2009;58:257–62.

    Article  CAS  PubMed  Google Scholar 

  5. Lee D, Li B, Im N, Kim Y, Jeong G. 4,2′,5′-Trihydroxy-4′-methoxychalcone from Dalbergia odorifera exhibits anti-inflammatory properties by inducing heme oxygenase-1 in murine macrophages. Int Immunopharmacol. 2013;16(1):114–21.

    Article  CAS  PubMed  Google Scholar 

  6. Cheng Z, Kuo S, Chan S, Ko F, Teng C. Antioxidant properties of butein isolated from Dalbergia odorifera. Biochim Biophys Acta. 1998;1392(2):291–9.

    Article  CAS  PubMed  Google Scholar 

  7. Ham SA, Hwang JS, Kang ES, Yoo T, Lim HH, Lee WJ, Paek KS, Seo HG. Ethanol extract of Dalbergia odorifera protects skin keratinocytes against ultraviolet B-induced photoaging by suppressing production of reactive oxygen species. Biosci Biotech Bioch. 2015;79(5):760–6.

    Article  CAS  Google Scholar 

  8. Tao Y, Wang Y. Bioactive sesquiterpenes isolated from the essential oil of Dalbergia odorifera T. Chen. Fitoterapia. 2010;81(5):393–6.

    Article  CAS  PubMed  Google Scholar 

  9. Park JD, Lee YH, Baek NI, Kim SI, Ahn BZ. Isolation of antitumor agent from the heartwood of Dalbergia odorifera. Kor J Pharmacogn. 1995;26(4):323–6.

    CAS  Google Scholar 

  10. Park K, Yun H, Quang T, Oh H, Lee D, Auh Q, Kim E. 4-Methoxydalbergione suppresses growth and induces apoptosis in human osteosarcoma cells in vitro and in vivo xenograft model through down-regulation of the JAK2/STAT3 pathway. Oncotarget. 2016;7(6):6960–71.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Zhao X, Mei W, Gong M, Zuo W, Bai H, Dai H. Antibacterial activity of the flavonoids from Dalbergia odorifera on Ralstonia solanacearum. Molecules. 2011;16(12):9775–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Yu S, Kuo S. Vasorelaxant effect of isoliquiritigenin, a novel soluble guanylate cyclase activator, in rat aorta. Brit J Pharmacol. 1995;114(8):1587–94.

    Article  CAS  Google Scholar 

  13. Yang Z, Mei C, He X, Sun X. Advance in studies on chemical constitutions, pharmacological mechanism and pharmacokinetic profile of Dalbergiae odoriferae lignum. China J Chin Mater Med. 2013;38(11):1679–83.

    CAS  Google Scholar 

  14. Wang F, Huang A, Yin X, Wang W, Chen J. Multilevel profiling and identification of Dalbergia odorifera and Dalbergia stevensonii by FTIR, NMR and GC/MS. Chin Chem Lett. 2018;29(9):1395–8.

    Article  Google Scholar 

  15. Yu M, Liu K, Zhou L, Zhao L, Liu S. Testing three proposed DNA barcodes for the wood identification of Dalbergia odorifera T. Chen and Dalbergia tonkinensis Prain. Holzforschung. 2016;70(2):127–36.

    Article  CAS  Google Scholar 

  16. Luo Z, Zhang X, Pan B, Lu B, Ruan Z, Wu B. Anatomy structure and physical and mechanical properties of Dalbergia tonkinensis. J Anhui Agric Univ. 2012;39(04):493–6.

    Google Scholar 

  17. Son NT, Oda M, Hayashi N, Yamaguchi D, Kawagishi Y, Takahashi F, Harada K, Cuong NM, Fukuyama Y. Antimicrobial activity of the constituents of Dalbergia tonkinensis and structural-bioactive bighlights. Nat Prod Commun. 2018;13(2):1934578X–801300212.

    Google Scholar 

  18. Nguyen VB, Wang S, Nhan NT, Nguyen TH, Nguyen NPD, Nghi DH, Cuong NM. New records of potent in-vitro antidiabetic properties of Dalbergia tonkinensis heartwood and the bioactivity-guided isolation of active compounds. Molecules. 2018;23(7):E1589.

    Article  Google Scholar 

  19. Cuong N, Ninh The S, Nhan N, Khanh P, Thu Hường T, Nguyen T, Sgaragli G, Ahmed A, Trezza A, Spiga O, et al. Vasorelaxing activity of R-(–)-3′-Hydroxy-2,4,5-trimethoxydalbergiquinol from Dalbergia tonkinensis: involvement of smooth muscle CaV1.2 channels. Planta Med. 2020;86(4):284–93.

    Article  PubMed  Google Scholar 

  20. Liu H, Shi F, Li W, Wei Z, Ma R, Liu Z, Li Y, Fu Y. Distinguishability of Dalbergia odorifera and Dalbergia tonkinensis in wood anatomic properties and extraction chemical compositions. Sci Silvae Sin. 2021;57(02):103–14.

    CAS  Google Scholar 

  21. Li G, Xu F, Luo J, Li Y, Lan F. Comparative anatomical research on the wood structure between Dalbergia odorifera T. Chen and D. rimosa Roxb. J Guangxi Agric Biol Sci. 2008;27:154–7.

    CAS  Google Scholar 

  22. Yang L, Fang C, Zhang J, Yu H, Wang Z, Zhu J, Yang W, Chai Z. The identification of Dalbergia odorifera and D. tonkinensis by gas chromatography mass spectrometry. J Nanjing for Univ. 2016;40(01):97–103.

    Google Scholar 

  23. Kress WJ, Wurdack KJ, Zimmer EA, Weigt LA, Janzen DH. Use of DNA barcodes to identify flowering plants. Proc Natl Acad Sci U S A. 2005;102(23):8369–74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Gonzalez M, Baraloto C, Engel J, Mori S, Pétronelli P, Riera B, Roger A, Thébaud C, Chave J. Identification of amazonian trees with DNA barcodes. PLoS ONE. 2009;4:e7483.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Li Q, Wu J, Wang Y, Lian X, Wu F, Zhou L, Huang Z, Zhu S. The phylogenetic analysis of Dalbergia (Fabaceae: Papilionaceae) based on different DNA barcodes. Holzforschung. 2017;71(12):939–49.

    Article  CAS  Google Scholar 

  26. Yu M, Jiao L, Guo J, Wiedenhoeft AC, He T, Jiang X, Yin Y. DNA barcoding of vouchered xylarium wood specimens of nine endangered Dalbergia species. Planta. 2017;246(6):1165–76.

    Article  CAS  PubMed  Google Scholar 

  27. Bhagwat R, Dholakia B, Kadoo N, Mangalashery B, Gupta V. Two new potential barcodes to discriminate Dalbergia Species. PLoS ONE. 2015;10:e142965.

    Article  Google Scholar 

  28. Phong D, Duong T, Hien V, Nguyen DT, Nong V. Nucleotide diversity of a nuclear and four chloroplast DNA regions in rare tropical wood species of Dalbergia in Vietnam: a DNA barcode identifying utility. Asian J Appl Sci. 2014;2(02):116–25.

    Google Scholar 

  29. Song Y, Zhang Y, Xu J, Li W, Li M. Characterization of the complete chloroplast genome sequence of Dalbergia species and its phylogenetic implications. Sci Rep-UK. 2019;9(1):20401.

    Article  CAS  Google Scholar 

  30. Win P, Li X, Chen L, Tan Y, Yu W. Complete plastid genome of two Dalbergia species (Fabaceae), and their significance in conservation and phylogeny. Mitochondrial DNA B. 2020;5:1967–9.

    Article  Google Scholar 

  31. Hong Z, He W, Liu X, Tembrock L, Wu Z, Xu D, Liao X. Comparative analyses of 35 complete chloroplast genomes from the genus Dalbergia (Fabaceae) and the identification of DNA barcodes for tracking illegal logging and counterfeit rosewood. Forests. 2022;13:626.

    Article  Google Scholar 

  32. Qin M, Zhu C, Yang J, Vatanparast M, Schley R, Lai Q, Zhang D, Tu T, Klitgård BB, Li S, et al. Comparative analysis of complete plastid genome reveals powerful barcode regions for identifying wood of Dalbergia odorifera and D. tonkinensis (Leguminosae). J Syst Evol. 2022;60(1):73–84.

    Article  Google Scholar 

  33. Zhang Z, Song M, Guan Y, Li H, Niu Y, Zhang L, Ma X. DNA barcoding in medicinal plants: testing the potential of a proposed barcoding marker for identification of Uncaria species from China. Biochem Syst Ecol. 2015;60:8–14.

    Article  CAS  Google Scholar 

  34. Chen R, Jiang L, Liu L, Liu QH, Wen J, Zhang R, Li X, Wang YG, Lei F, Qiao G. The gnd gene of Buchnera as a new, effective DNA barcode for aphid identification. Syst Entomol. 2013;38:615–25.

    Article  Google Scholar 

  35. Ross H, Murugan S, Li WLS. Testing the reliability of genetic methods of species identification via simulation. Syst Biol. 2008;57(2):216–30.

    Article  PubMed  Google Scholar 

  36. GU J, SU JX, LIN RZ, LI RQ, Xiao P. Testing four proposed barcoding markers for the identification of species within Ligustrum L. (Oleaceae). J Syst Evol. 2011;49:213–24.

    Article  Google Scholar 

  37. Little DP, Stevenson DW. A comparison of algorithms for the identification of specimens using DNA barcodes: examples from gymnosperms. Cladistics. 2007;23(1):1–21.

    Article  PubMed  Google Scholar 

  38. Rach J, DeSalle R, Sarkar IN, Schierwater B, Hadrys H. Character-based DNA barcoding allows discrimination of genera, species and populations in Odonata. Proc Royal Soc B. 2007;275:237–47.

    Article  Google Scholar 

  39. Weitschek E, Van Velzen R, Felici G, Bertolazzi P. BLOG 2.0: a software system for character-based species classification with DNA barcode sequences. What it does, how to use it. Mol Ecol Resour. 2013;13(6):1043–6.

    Article  PubMed  Google Scholar 

  40. He T, Jiao L, Yu M, Guo J, Jiang X, Yin Y. DNA barcoding authentication for the wood of eight endangered Dalbergia timber species using machine learning approaches. Holzforschung. 2019;73(3):277–85.

    Article  CAS  Google Scholar 

  41. Will KW, Rubinoff D. Myth of the molecule: DNA barcodes for species cannot replace morphology for identification and classification. Cladistics. 2004;20(1):47–55.

    Article  PubMed  Google Scholar 

  42. Yassin A, Markow TA, Narechania A, O Grady PM, DeSalle R. The genus Drosophila as a model for testing tree- and character-based methods of species identification using DNA barcoding. Mol Phylogenet Evol. 2010;57(2):509–17.

    Article  CAS  PubMed  Google Scholar 

  43. Newmaster SG, Fazekas AJ, Ragupathy S. DNA barcoding in land plants: evaluation of rbcL in a multigene tiered approach. Can J Bot. 2006;84(3):335–41.

    Article  CAS  Google Scholar 

  44. Rakotonirina TJ, Viljoen E, Rakotonirina AH, Leong Pock Tsy JM, Radanielina T. A DNA barcode reference library for CITES listed Malagasy Dalbergia species. Ecol Evol. 2023;13(3):e9887.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The authors gratefully acknowledge the financial supports by National Natural Science Foundation of China under Grant numbers 31870540.

Funding

This research was funded by the National Natural Science Foundation of China, grant number 31870540.

Author information

Authors and Affiliations

Authors

Contributions

M.J.Q, B.X.C and W.J.W conceived and designed the experiments. B.X.C and W.J.W performed the experiments and analyzed the data. W.J.W wrote the manuscript. R.K.M, M.J.Q and Y.L.F edited and revised the manuscript. All authors read and agreed at the last version of the manuscript.

Corresponding authors

Correspondence to Mengji Qiao or Yunlin Fu.

Ethics declarations

Ethics approval and consent to participate

No specific permits were required, and all local, national, or international guidelines and legislation were adhered to in the production of this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, W., Chen, B., Ma, R. et al. The DNA barcode identification of Dalbergia odorifera T. Chen and Dalbergia tonkinensis Prain. BMC Plant Biol 23, 546 (2023). https://0-doi-org.brum.beds.ac.uk/10.1186/s12870-023-04513-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s12870-023-04513-3

Keywords