Schima is an ecologically and economically important woody genus in tea family (Theaceae). Unresolved species delimitations and phylogenetic relationships within Schima limit our understanding of the genus and hinder utilization of the genus for economic purposes. In the present study, we conducted comparative analysis among the complete chloroplast (cp) genomes of 11 Schima species. Our results indicate that Schima cp genomes possess a typical quadripartite structure, with conserved genomic structure and gene order. The size of the Schima cp genome is about 157 kilo base pairs (kb). They consistently encode 114 unique genes, including 80 protein-coding genes, 30 tRNAs, and 4 rRNAs, with 17 duplicated in the inverted repeat (IR). These cp genomes are highly conserved and do not show obvious expansion or contraction of the IR region. The percent variability of the 68 coding and 93 noncoding (>150 bp) fragments is consistently less than 3%. The seven most widely touted DNA barcode regions as well as one promising barcode candidate showed low sequence divergence. Eight mutational hotspots were identified from the 11 cp genomes. These hotspots may potentially be useful as specific DNA barcodes for species identification of Schima. The 58 cpSSR loci reported here are complementary to the microsatellite markers identified from the nuclear genome, and will be leveraged for further population-level studies. Phylogenetic relationships among the 11 Schima species were resolved with strong support based on the cp genome data set, which corresponds well with the species distribution pattern. The data presented here will serve as a foundation to facilitate species identification, DNA barcoding and phylogenetic reconstructions for future exploration of Schima.
Citation: Yu X-Q, Drew BT, Yang J-B, Gao L-M, Li D-Z (2017) Comparative chloroplast genomes of eleven Schima (Theaceae) species: Insights into DNA barcoding and phylogeny. PLoS ONE 12(6): e0178026. https://doi.org/10.1371/journal.pone.0178026
Editor: Genlou Sun, Saint Mary's University, CANADA
Received: January 7, 2017; Accepted: April 11, 2017; Published: June 2, 2017
Copyright: © 2017 Yu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The GenBank accession numbers are listed in Table 1 of the paper.
Funding: This work was supported by grants from the Applied Fundamental Research Foundation of Yunnan Province (2014FB167; 2014GA003), the National Key Basic Research Program of China (2014CB954100) and the China Postdoctoral Science Foundation (2014M562352).
Competing interests: The authors have declared that no competing interests exist.
The chloroplast (cp) is a type of plastid that is critical to the growth of most plants, playing a major role in photosynthesis and fixation of CO2 . The cp genomes in angiosperms are circular DNA molecules with a highly conserved gene order and gene content, and range from 120 to 160 kb in length . These genomes typically include two copies of an inverted repeat (IR) region that is separated by a large-single-copy (LSC) region and a small-single-copy (SSC) region . Due to the rapid accumulation of genomic data gleaned from next-generation sequencing (NGS) technologies [4–6], more than 800 complete cp genomes of land plants have been sequenced (up to December 2016 from NCBI). The cp genome can provide valuable information for species identification, phylogeny and population genetic analyses [7–9]. It has also been postulated to be a potential ultra- or organelle-scale barcode for efficient plant species identification, especially for the taxonomically complex groups [10, 11].
Schima, with ca. 20 species, is an economically and ecologically important genus of the tea family (Theaceae). The genus is distributed in subtropical and tropical areas of East Asia, with 13 species (6 endemic) present in China . Species of Schima are large trees and dominant elements of the subtropical evergreen broadleaved forests (SEBLFs) in East Asia [13, 14]. Some species are used as biological fire-resistant trees, and the wood is used for building and furniture [15, 16]. Schima is distinct from other genera within Theaceae, characterized by globose to oblate fruits and small reniform seeds with a marginal membranous wing. However, the infrageneric classification of Schima is complex and controversial due to a dearth of taxonomically diagnostic characters and high morphological similarity among species. This taxonomic uncertainty may hinder our exploitation and utilization of the genus.
Since its establishment as a genus, there has been much debate regarding the number of species within Schima . Eighteen species were proposed in the second edition of the “Die Natürlichen Pflanzenfamilien” . Bloembergen  regarded the genus as monotypic and subdivided Schima wallichii into nine geographically separated subspecies and three varieties. Airy-Shaw  recognized 15 species in Schima. Keng  accepted most of Bloembergen’s subspecies and raised them to the species level, and proposed that there were 10–15 species within the genus. The most recent treatment recognized ca. 20 species in Schima . Schima is placed in tribe Gordonieae based on the results of molecular phylogenetic studies [22–24]. However, phylogenetic relationships within Schima are still unclear due to limited species sampling in previous studies, thus both species delimitations and phylogenetic reconstruction within Schima require further exploration.
Complete cp genomes have been shown to be effective in resolving interspecies phylogenetic relationships within Camellia, a genus in the sister tribe (Theeae) to Gordonieae [25, 26]. Here, we sequenced 11 cp genomes of the 13 Chinese Schima species. This study aims to: (1) investigate structural patterns of Schima cp genomes, (2) screen sequence divergence hotspots in the 11 Schima cp genomes, (3) explore simple sequence repeats (SSRs) among the 11 Schima cp genomes, (4) and reconstruct phylogenetic relationships among the 11 Schima species using the cp genome sequences. The results will provide abundant information for further studies regarding taxonomy, phylogeny, and population genetics of Schima, and will also assist in the exploration and utilization of the resources within the genus.
Materials and methods
In this study, we follow the classification of Schima from Min and Bartholomew . Healthy and fresh leaves from 11 species of Schima were sampled from various localities across southern China (Table 1). Voucher specimens of each species were collected and deposited in the Herbarium of Kunming Institute of Botany, Chinese Academy of Sciences (KUN). Gordonia lasianthus and Franklinia alatamaha were used as outgroups in the phylogenetic analyses, and the cp genomes of these two species were obtained from our previous work (Yu et al., unpublished work).
DNA extraction, sequencing, chloroplast genome assembly
Total genomic DNA was isolated from fresh leaves (~100 mg) using the modified CTAB method (Doyle and Doyle 1987). Subsequently, the cp genomes were amplified using long-range PCR with fifteen primers . The PCR products were fragmented for constructing short-insert (500 bp) libraries following the Illumina Nextera XT DNA library preparation instructions. Paired-end sequencing (250 bp) was performed on the Illumina MiSeq 2000 at the Laboratory of Molecular Biology of Germplasm Bank of Wild Species in Southwest China. Quality control of the raw sequence reads was performed using the NGS QC Tool Kit , with a cut-off value for percentage of read length and PHRED quality score as 80 and 30 following Yang et al. . High-quality reads were assembled into contigs using the de novo assembler in CLC Genomics Workbench v6.5 (CLC Bio), using a k-mer of 64 and a minimum contig length of 500 base pairs (bp). The de novo contigs were assembled into complete chloroplast genomes followed the procedure of Yang et al. .
Chloroplast genome annotation and comparisons
The complete cp genomes were annotated with the identification of introns and exons using DOGMA . The positions of start and stop codons and boundaries between introns and exons were investigated according to the published cp genome of Camellia taliensis (NC022264). The annotated GenBank files were used to draw the circular chloroplast genome maps using OrganellarGenomeDRAW . The mVISTA program  was employed in the LAGAN mode to detect the variation of the chloroplast genomes. The cp genome of Schima sinensis was used as a reference. Microsatellites (mono-, di-, tri-, tetra-, penta- and hexanucleotide repeats) were detected using Phobos v3.311 , with the parameters set to ten repeat units (≥ 10) for mononucleotide SSRs, six repeat units (≥ 6) for dinucleotide, four repeat units (≥ 4) for trinucleotide, four repeat units (≥ 4) for tetranucleotide, and three repeat units (≥ 3) for pentanucleotide and hexanucleotide SSRs. The percent variability for all protein-coding and noncoding (intergenic spacers and introns) regions of the cp genomes with an aligned length larger than 150 bp among the 11 Schima species was estimated in Geneious .
The cp genomes were aligned using MAFFT v7.221  under default settings (FFT-NS-2 strategy). One of the IRs was removed from the data set for the phylogenetic analysis. Poorly aligned regions (mainly introns and spacers) of the data set were realigned using the G-INS-i (accurate strategy) to improve the quality of the alignment. We used jModelTest v0.11  to select the best-fitting nucleotide substitution models for maximum-likelihood (ML) according to the Akaike information criterion (AIC; Akaike, 1974). ML analysis was implemented in RAxML v8.20 . We conducted a rapid bootstrap analysis (1000 replicates) and searched for the best-scoring ML tree simultaneously (the “-f a” option). Numbers of variable and informative sites were calculated in DnaSP v5.10 .
Chloroplast genome features
Illumina paired-end sequencing of long-range PCR amplified cp DNA generated 28,866–10,152,425 clean reads for the 11 sampled Schima species, with mean coverage from 71.7 to 1882.5. The genome size ranged from 157,227 bp in Schima brevipedicellata to 157,302 bp in Schima sericans (Table 1). All of the 11 cp genomes showed typical quadripartite structure consisting of a pair of IR (25,954–25,985 bp) separated by the LSC (87,202–87,272 bp) and SSC (18,089–18,122 bp) regions (Table 1). The cp genome map of Schima superba is presented as a representative (Fig 1). Excluding the duplicated IR region, the 11 Schima cp genomes identically encoded 114 different genes that were arranged in the same order, including 80 protein-coding genes, 30 tRNAs and 4 rRNAs. Seventeen genes were duplicated in the IRs, with six protein-coding genes, four rRNA and seven tRNA genes. Twelve of the protein-coding genes and six of the tRNA genes contained introns. Fifteen out of those eighteen genes contained a single intron, while the other three (clpP, rps12 and ycf3) had two introns. The 11 Schima cp genomes exhibited high similarity at the LSC/IR/SSC boundaries (Fig 2). The rps19 gene crossed the LSC/IRB (JLB) region with no variation of sequence length within the two parts. The SSC/IRB (JSB) junction occurred between the ycf1_like (incompletely duplicated in IRB) and the 3’ end of ndhF gene, with the sequence length of ycf1_like gene within IRB as 1388 or 1394. The ycf1 gene crossed the SSC/IRA (JSA) region, with 1388 or 1394 bp of ycf1 within IRA. The ycf1 related length changes were the only variation detected in these junctions. The LSC/IRA (JLA) junction was located at the 3’ end of the rps19_like (6 bp; incompletely duplicated in IRA), with a 14 bp noncoding sequence between JLA and trnH gene. In addition, we identified unusual start codons for four genes, ACG for ndhD, ATC for psbI, ATT for psbT and GTG for rps19.
Chloroplast genome comparisons and divergence hotspots
Sequence identity plots of the 11 Schima cp genomes, generated using mVISTA, are shown in Fig 3. The plots illustrate the high sequence similarity across the Schima cp genomes, with a sequence identity of 99.1%. Two (ccsA and rps15) of the 49 variable protein-coding (>150 bp) genes had a percentage of variation above 1.00% (Table 2), while 19 (>150 bp) had no variation. Both of the two core DNA barcodes (rbcL and matK)  showed extremely low sequence divergence (0.21% and 0.33%, respectively). Furthermore, the variation of ycf1, the proposed “most promising chloroplast DNA barcode” of land plants , was only 0.67%. Among the 79 noncoding (>150 bp) regions, the percentage of variation ranged from 0.11% to 2.85% (Fig 4 and Table 3). Fourteen fragments (atpI-rps2, trnS (UGA)-psbZ, rps4-trnT (UGU), trnL (UAA)-trnF (GAA), petB-petD, rpl2 intron, rpl23-trnI (CAU), ycf15-trnL (CAA), trnL (CAA)-ndhB, ndhB intron, trnV (GAC)-rrn16, rrn16-trnI (GAU), trnA (UGC) intron, trnN (GUU)-ndhF) did not show any sequence variation. Eight potential mutational hotspots (trnW (CCA)-trnP (UGG), trnT (UGU)-trnL (UAA), trnG (UCC)-trnfM (CAU), petD-rpoA, psbB-psbT, ndhE-ndhG, ndhC-trnV (UAC), rpl32-trnL (UAG)) were identified, with the variation percentage exceeding 2.0% among the 11 sampled species (Fig 4 and Table 3). These eight highly variable hotspots may have the potential to be used as special DNA barcodes for identifying Schima species.
These regions are oriented according to their locations in the genome.
The aligned length of the complete cp genome (with one of the IR removed) among the 11 Schima species was 130,508 bp, with the total number of variable and parsimony informative (PI) sites being 1,121 bp and 261 bp, respectively. This data set contained 131 indels with a total length of 586 bp, and the percent variability was 0.51%. These results indicate that the global variation of the cp genome within Schima is extremely low. A similar pattern was reported in other long-lived plants [40–42]. The ability to identify species within the genus using cp genome data needs to be assessed by sampling multiple individuals per species, even though the phylogenetic analyses have most of the species separated from each other (see below).
In total, 58 cpSSRs, including 55 mononucleotide (A, T), 1 dinucleotide (AT) and 2 trinucleotide (ATT, TTA) repeats were detected within the 11 Schima cp genomes. No tetranucleotide, pentanucleotide or hexanucleotide repeats were observed. The mononucleotide repeat (A, T) was found to be the most abundant, with repeat numbers of 10, 11 and 12 (Table 4). The proportion of A and T repeats in mononucleotide repeat unit was 43.64% and 56.36%, respectively. Only one SSR locus with a different repeat unit (C) was detected in the trnG (UCC)-trnfM (CAU) intergenic spacer region. Within the 11 Schima cp genomes, SSR loci were primarily located in the LSC region (89.09%), followed by the SSC portion (14.55%), with only one present in the IR region (rrn5-trnR (ACG)) (Table 4). One SSR locus was detected in the protein-coding gene psbI, with all others located in gene spacers and introns. No SSRs were found in the tRNAs and rRNAs. The mononucleotide repeat (A) in trnH-psbA was the most variable SSR, with the size ranging from 12 to 42 bp. The cpSSRs of the 11 Schima species represented here showed abundant variation, and could be useful for research at the population level. They will provide complementary data to the SSR markers of Schima identified from the nuclear genome .
The data matrix we used for phylogenetic estimation consisted of an alignment containing entire cp genomes with one of the IRs removed. This data set was comprised of 131,113 nucleotide positions, with 2,508 variable sites (1.91%) and 427 PI sites (0.33%). ML analysis resulted in a well-resolved tree, with eight of the 10 nodes supported by 100% bootstrap values (BS). All Schima species grouped into a strongly supported clade (BS = 100%, Fig 5), indicating Schima is monophyletic. Two main clades were recovered, with Schima sericans being sister to those two lineages. Five species (S. argentea, S. brevipedicellata, S. khasiana, S. noronhae, S. wallichii) formed clade I (BS = 100%, Fig 5). The remaining five species (S. sinensis, S. superba, S. remotiserrata, S. multibracteata and S. crenata) grouped in clade II (BS = 100%, Fig 5). The branch leading to S. superba and three closely related species is extremely short, and bootstrap support values for two internal nodes within this clade are less than 80%.
Chloroplast genome features and comparison within Theaceae
Prior to this study, Camellia was the only genus within Theaceae to have its cp genome sequenced [25, 26]. In the present study, we sequenced cp genomes of 11 species from Schima. The cp genomes all displayed typical quadripartite structure (Fig 1), which is consistent amongst most lineages of angiosperms . The expansion and contraction of the IR region is considered to be the primary mechanism affecting length variation of angiosperm cp genomes, as demonstrated in Trochodendraceae  and Apiales . However, only minor variation was detected at the SSC/IRA boundary of all of the 11 Schima cp genomes (Fig 2). Although the genes located at the IR junctions are identical in cp genomes of Schima and Camellia, the overall cp genome sequences of Schima are more homogenous as compared to Camellia, which was suggested to show more differences at the junction regions . The cp genomes of Schima encode the same set of protein-coding genes as previously reported Camellia species, with the exception of Orf 42 and Orf 188 which were reported in Camellia , but not in other Ericales members such as Actinidia (Actinidiaceae) and Ardisia (Primulaceae) [46, 47]. For the whole cp genomes of Schima, 37 tRNA genes were annotated, which is consistent with Huang et al. . However, 38 tRNA genes were found in Yang et al. , due to a redundant annotation of trnP (UGG) in their study. As compared with sequences of Camellia, no significant structural rearrangements such as inversions or changes of gene locations were found in the 11 Schima cp genomes. The high sequence similarity across the Schima cp genomes (Fig 3) may be associated with long generation time and recent radiation.
Potentially specific DNA barcodes for Schima
Since the concept of DNA barcoding was proposed over a decade ago , substantial efforts have been made to develop DNA barcodes possessing both high universality and efficiency. Kress et al.  suggested that the internal transcribed spacer (ITS) and trnH-psbA spacer region had potential as useful DNA barcode regions for flowering plants. Hollingsworth et al.  later advocated matK and rbcL as a two-locus core barcode for land plants after comparing seven leading candidate loci, subsequently the nrDNA ITS was recommended to be incorporated into core barcode based on a large-scale sapmpling of seed plants . Dong et al.  proposed that ycf1 was the most variable loci of the cp genome, which might be a promising DNA barcode performing better than existing plastid candidate barcodes of land plants. However, all of the five candidate protein-coding DNA barcodes (matK, rbcL, rpoB, rpoC1 and ycf1) showed extremely low sequence variation (<1.00%), and the other three fragments are also not among the most variable spacers. The eight potential mutational hotspots (trnW (CCA)-trnP (UGG), trnT (UGU)-trnL (UAA), trnG (UCC)-trnfM (CAU), petD-rpoA, psbB-psbT, ndhE-ndhG, ndhC-trnV (UAC), rpl32-trnL (UAG)) (Fig 4 and Table 3) identified in this study could be suitable barcodes for Schima. Recently, using the cp genome as a possible ultra- or organelle-scale barcode for efficient plant species identification was discussed [10, 11]. The high phylogenetic resolution among closely related species of Schima (Fig 5) suggests that the cp genome may indeed be useful as an organelle-scale barcode for species identification of Schima. Further studies based on sampling at the population scale are needed to evaluate the efficiency of the barcodes mentioned above and also the cp genome as an organelle-scale barcode.
Phylogenetic relationships among species of Schima
The cp genome has been suggested to be useful for phylogenetic reconstructions at low taxonomic levels [7, 8, 10, 51]. Interspecies phylogenetic relationships within Camellia (Theaceae) were well-resolved using cp genome data [25, 26]. In the present study, based on a recent classification of the genus , 11 out of 13 Schima species occurring in China were represented. The phylogenetic relationships within Schima were well resolved with strong support based on cp genome sequences (Fig 5). Therefore, our study indicates that the complete cp genome has significant potential to resolve the low level phylogenetic relationships. Schima sericans, the first diverging lineage among sampled species, is distributed in southeastern Xizang and northwestern Yunnan in China. Schima sericans was sister to the remaining taxa, which formed two clades. Clade I includes five species (S. argentea, S. brevipedicellata, S. khasiana, S. noronhae and S. wallichii) that are primarily distributed in southwestern China and Indochina. Clade II comprises five species (S. sinensis, S. superba, S. remotiserrata, S. multibracteata and S. crenata) that mainly occur within central and eastern China (Fig 5). The phylogenetic relationships within Schima found here correspond well with the geographic distribution pattern, but do not match well with morphology. Schima was classified into two groups based on the shape of the leaf margin (entire or serrate) . However, all of the species within clade I possess a serrate leaf margin except S. khasiana. Likewise, S. multibracteata is the only species with an entire leaf margin in clade II (Fig 5). Our results suggest that the taxonomic value of leaf margin shape should be reassessed for classification of Schima. Additionally, the branch length of the clade including S. superba and three closely related species is extremely short, indicating that these four species have recently diversified, or perhaps illustrating past hybridization within the group. These results indicate that the current treatment of the genus needs to be reevaluated by integrating more types of evidence.
We thank Liang Fang, Jie Cai, Ting Zhang and Ji-Dong Ya for their help of sample collection. We are grateful to Shi-Xiong Yang for assistance with the species identification, to Jing Yang and Ji-Xiong Yang for assisting with the laboratory work, and to Chao-Nan Fu for help with data analysis. We also thank the staff at KUN for providing access to study specimens.
- Conceptualization: DZL LMG.
- Formal analysis: XQY JBY.
- Funding acquisition: DZL XQY.
- Investigation: XQY.
- Supervision: LMG DZL.
- Writing – original draft: XQY.
- Writing – review & editing: BTD LMG DZL.
- 1. Neuhaus H, Emes M. Nonphotosynthetic metabolism in plastids. Annu Rev Plant Biol. 2000;51(1):111–40.
- 2. Wicke S, Schneeweiss GM, Müller KF, Quandt D. The evolution of the plastid chromosome in land plants: gene content, gene order, gene function. Plant Mol Biol. 2011;76(3–5):273–97. pmid:21424877
- 3. Palmer JD. Comparative organization of chloroplast genomes. Annual Review of Genetics. 1985;19(1):325–54.
- 4. Zeng LP, Zhang Q, Sun RR, Kong HZ, Zhang N, Ma H. Resolution of deep Angiosperm phylogeny using conserved nuclear genes and estimates of early divergence times. Nat Commun. 2014;5:4956. pmid:25249442
- 5. Yang JB, Li DZ, Li HT. Highly effective sequencing whole chloroplast genomes of Angiosperms by nine novel universal primer pairs. Mol Ecol Resour. 2014;14(5):1024–31. pmid:24620934
- 6. Barrett CF, Baker WJ, Comer JR, Conran JG, Lahmeyer SC, Leebens-Mack JH, et al. Plastid genomes reveal support for deep phylogenetic relationships and extensive rate variation among palms and other commelinid monocots. New Phytol. 2016;209:855–70. pmid:26350789
- 7. Zhang YJ, Du LW, Liu A, Chen JJ, Wu L, Hu WM, et al. The complete chloroplast genome sequences of five Epimedium species: lights into phylogenetic and taxonomic analyses. Front Plant Sci. 2016;7:306. pmid:27014326
- 8. Zhang YJ, Ma PF, Li DZ. High-throughput sequencing of six bamboo chloroplast genomes: phylogenetic implications for temperate woody bamboos (Poaceae: Bambusoideae). PLoS ONE. 2011;6(5):e20596. pmid:21655229
- 9. Drew BT, Ruhfel BR, Smith SA, Moore MJ, Briggs BG, Gitzendanner MA, et al. Another look at the root of the angiosperms reveals a familiar tale. Syst Biol. 2014;63(3):368–82. pmid:24391149
- 10. Yang JB, Tang M, Li HT, Zhang ZR, Li DZ. Complete chloroplast genome of the genus Cymbidium: lights into the species identification, phylogenetic implications and population genetic analyses. BMC Evol Biol. 2013;13:84. pmid:23597078
- 11. Kane N, Sveinsson S, Dempewolf H, Yang JY, Zhang D, Engels JM, et al. Ultra-barcoding in cacao (Theobroma spp.; Malvaceae) using whole chloroplast genomes and nuclear ribosomal DNA. Am J Bot. 2012;99(2):320–9. pmid:22301895
- 12. Min TL, Bartholomew B. Theaceae. In: Wu CY, Raven PH, editors. Flora of China. 12. Beijing, China and Saint Louis, Missouri, USA: Science Press and Missouri Botanical Garden Press; 2007. p. 366–478.
- 13. Fang JY, Yoda K. Climate and vegetation in China II. Distribution of main vegetation types and thermal climate. Ecol Res. 1989;4(1):71–83.
- 14. Tang CQ. Evergreen Broad-Leaved Forests. In: Tang CQ, editor. The Subtropical Vegetation of Southwestern China: Plant Distribution, Diversity and Ecology Utrecht, The Netherlands: Springer Netherlands; 2015. p. 49–112.
- 15. Chang HD, Ren SX. Theaceae (1). In: Wu CY, editor. Flora Reipublicae Popularis Sinicae. 49. Beijing: Science Press; 1998.
- 16. Tian XR, Shu LF, He QT. Selection of fire-resistant tree species for southwestern China. Forestry Studies in China. 2001;3(2):32–8.
- 17. Blume CL. Bijdragen tot de flora van Nederlandsch Indië. Batavia [Jakarta], Indonesia: Ter Lands Drukkerij; 1826.
- 18. Melchior H. Theaceae. In: Engle A, Prantl E, editors. Die naturlichen Pflanzenfamilien, 2nd ed. 21. Leipzig, Germany: Wilhelm Engelmann; 1925. p. 109–54.
- 19. Bloembergen S. A critical study in the complex-polymorphous genus Schima (Theaceae). Reinwardtia. 1952;2:133–83.
- 20. Airy Shaw HK. A dictionary of the flowering plants and ferns (Willis J.C.) Revised 8th edition. Cambridge, UK: Cambridge University Press; 1985.
- 21. Keng H. Flora Malesianae precursors LVIII, part 4: the genus Schima (Theaceae) in Malesia. Gardens' Bulletin (Singapore) 1994;46(1):77–87.
- 22. Prince LM, Parks CR. Phylogenetic relationships of Theaceae inferred from chloroplast DNA sequence data. Am J Bot. 2001;88(12):2309–20. pmid:21669662
- 23. Yang SX, Yang JB, Lei LG, Li DZ, Yoshino H, Ikeda T. Reassessing the relationships between Gordonia and Polyspora (Theaceae) based on the combined analyses of molecular data from the nuclear, plastid and mitochondrial genomes. Plant Syst Evol. 2004;248(1–4):45–55.
- 24. Li MM, Li JH, Del Tredici P, Corajod J, Fu CX. Phylogenetics and biogeography of Theaceae based on sequences of plastid genes. Journal of Systematics and Evolution. 2013;51(4):396–404.
- 25. Yang JB, Yang SX, Li HT, Yang J, Li DZ. Comparative Chloroplast Genomes of Camellia Species. PLoS ONE. 2013;8(8):e73053. pmid:24009730
- 26. Huang H, Shi C, Liu Y, Mao SY, Gao LZ. Thirteen Camellia chloroplast genome sequences determined by high-throughput sequencing: genome structure and phylogenetic relationships. BMC Evol Biol. 2014;14:151. pmid:25001059
- 27. Zhang T, Zeng CX, Yang JB, Li HT, Li DZ. Fifteen novel universal primer pairs for sequencing whole chloroplast genomes and a primer pair for nuclear ribosomal DNAs. Journal of Systematics and Evolution. 2016;54(3):219–27.
- 28. Patel RK, Jain M. NGS QC Toolkit: A Toolkit for Quality Control of Next Generation Sequencing Data. PLoS ONE. 2012;7(2):e30619. pmid:22312429
- 29. Wyman SK, Jansen RK, Boore JL. Automatic annotation of organellar genomes with DOGMA. Bioinformatics. 2004;20(17):3252–5. pmid:15180927
- 30. Lohse M, Drechsel O, Kahlau S, Bock R. OrganellarGenomeDRAW—a suite of tools for generating physical maps of plastid and mitochondrial genomes and visualizing expression data sets. Nucleic Acids Res. 2013;41((Web Server issue)):W575–81. pmid:23609545
- 31. Frazer KA, Pachter L, Poliakov A, Rubin EM, Dubchak I. VISTA: computational tools for comparative genomics. Nucleic Acids Res. 2004;32(suppl 2):W273–W9.
- 32. Mayer C. Phobos 3.3.11. Available from: http://www.rub.de/spezzoo/cm/cm_-phobos.htm. 2010.
- 33. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, et al. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012;28(12):1647–9. pmid:22543367
- 34. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80. pmid:23329690
- 35. Posada D. jModelTest: phylogenetic model averaging. Mol Biol Evol. 2008;25(7):1253–6. pmid:18397919
- 36. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics. 2014;30(9):1312–3. pmid:24451623
- 37. Rozas J, Sánchez-DelBarrio JC, Messeguer X, Rozas R. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics. 2003;19(18):2496–7. pmid:14668244
- 38. Hollingsworth PM, Forrest LL, Spouge JL, et al. A DNA barcode for land plants. Proc Natl Acad Sci USA. 2009;106(31):12794–7. pmid:19666622
- 39. Dong WP, Xu C, Li CH, Sun JS, Zuo YJ, Shi S, et al. ycf1, the most promising plastid DNA barcode of land plants. Sci Rep. 2015;5:8348. pmid:25672218
- 40. Cai J, Ma PF, Li HT, Li DZ. Complete plastid genome sequencing of four Tilia species (Malvaceae): a comparative analysis and phylogenetic implications. PloS ONE. 2015;10(11):e0142705. pmid:26566230
- 41. Song Y, Dong WP, Liu B, Xu C, Yao X, Gao J, et al. Comparative analysis of complete chloroplast genome sequences of two tropical trees Machilus yunnanensis and Machilus balansae in the family Lauraceae. Front Plant Sci. 2015;6:662. pmid:26379689
- 42. Turner B, Paun O, Munzinger J, Chase MW, Samuel R. Sequencing of whole plastid genomes and nuclear ribosomal DNA of Diospyros species (Ebenaceae) endemic to New Caledonia: many species, little divergence. Ann Bot. 2016;117(7):1175–85. pmid:27098088
- 43. Niu HY, Li XY, Ye WH, Wang ZF, Cao HL, Wang ZM. Isolation and characterization of 36 polymorphic microsatellite markers in Schima superba (Theaceae). Am J Bot. 2012;99(3):E123–E6. pmid:22371858
- 44. Sun YX, Moore MJ, Meng AP, Soltis PS, Soltis DE, Li JQ, et al. Complete plastid genome sequencing of Trochodendraceae reveals a significant expansion of the inverted repeat and suggests a Paleogene divergence between the two extant species. PLoS ONE. 2013;8(4):e60429. pmid:23577110
- 45. Downie SR, Jansen RK. A comparative analysis of whole plastid genomes from the Apiales:expansion and contraction of the Inverted Repeat, Mitochondrial to plastid transfer of DNA, and Identification of highly divergent noncoding regions. Syst Bot. 2015;40(1):336–51.
- 46. Ku C, Hu JM, Kuo CH. Complete plastid genome sequence of the basal asterid Ardisia polysticta Miq. and comparative analyses of asterid plastid genomes. PLoS ONE. 2013;8(4):e62548. pmid:23638113
- 47. Yao XH, Tang P, Li ZZ, Li DW, Liu YF, Huang HW. The First Complete Chloroplast Genome Sequences in Actinidiaceae: Genome Structure and Comparative Analysis. PLoS ONE. 2015;10(6):e0129347. pmid:26046631
- 48. Hebert PDN, Cywinska A, Ball SL, DeWaard JR. Biological identifications through DNA barcodes. Phil Trans R Soc B. 2003;270(1512):313–21.
- 49. Kress WJ, Wurdack KJ, Zimmer EA, Weigt LA, Janzen DH. Use of DNA barcodes to identify flowering plants. Proc Natl Acad Sci USA. 2005;102(23):8369–74. pmid:15928076
- 50. Li DZ, Gao LM, Li HT, Wang H, Ge XJ, Liu JQ, et al. Comparative analysis of a large dataset indicates that internal transcribed spacer (ITS) should be incorporated into the core barcode for seed plants. Proc Natl Acad Sci USA. 2011;108(49):19641–6. pmid:22100737
- 51. Ma PF, Zhang YX, Zeng CX, Guo ZH, Li DZ. Chloroplast Phylogenomic Analyses Resolve Deep-Level Relationships of an Intractable Bamboo Tribe Arundinarieae (Poaceae). Syst Biol. 2014;63(6):933–50. pmid:25092479