Abstract
Background and aims: With the recent proposal of matK and rbcL as core plant DNA barcoding regions by the Consortium for the Barcoding of Life Plant Working Group, the construction of reference libraries in the botanical DNA barcoding initiative has entered a new phase. However, in a recent DNA barcoding study in the three Mexican genera of the gymnosperm order Cycadales, we found that neither matK nor rbcL allow high levels of molecular identification of previously established species.
Materials and methods: Our data analysis in that study rested on the “Characteristic Attributes Organization System” (CAOS), a character-based algorithm for the definition of “DNA diagnostics.” Here, we use CAOS to analyze a population-level molecular data set in Zamia, one of the three cycad genera occurring in Mexico, whose populations display contrasting biogeographic patterns. Our population-level study, which includes all species in the region formally known as Megamexico, is restricted to the genome region, which showed the best single-locus molecular identification performance in our previous study—namely, the noncoding intergenic chloroplast spacer psbK-I.
Results: Our comparison of single-individual vs. population-level psbK-I datasets in Zamia indicates that CAOS analyses are sensitive to slight alignment changes, which in turn derive from the different amounts of molecular variation present in each matrix type.
Conclusion: We, therefore, suggest that character-based studies that involve population-level data should contemplate this type of comparison between data matrices, before a set of DNA diagnostics in a given DNA barcoding reference library is considered definitive.
Introduction
In 2009, the Consortium for the Barcoding of Life Plant Working Group (CBOL PWG) selected the chloroplast regions matK and rbcL as a “core barcode” for the land plants. Both gene-coding loci had been widely used already in plant molecular systematics studies, some of which helped to establish the research field as such (e.g. Chase et al. Citation1993; Hilu et al. Citation2003). Therefore, these regions were included in several plant DNA barcoding projects (Chase et al. Citation2005, Citation2007; Kress et al. Citation2005; Cowan et al. Citation2006; Kress and Erickson 2007; Erickson et al. Citation2008; Fazekas et al. Citation2008, Citation2009; CitationLahaye et al. 2008a,Citationb; Ford et al. Citation2009), whose results ultimately led to the CBOL PWG consensus.
In a recent DNA barcoding (i.e. molecular identification) study in the three genera of cycads that occur in Mexico—Ceratozamia Brongn., Dioon Lindl. and Zamia L.—we tested the comparative performance of seven chloroplast coding regions and the nuclear internal transcribed spacer (ITS) (CitationNicolalde-Morejón et al. 2010). In contrast to the results presented in the CBOL PWG paper, we showed that both matK and rbcL failed to comply, in any of these cycad genera, with the criterion of “discrimination,” i.e. the third parameter according to which a candidate genome segment should be accepted as a consensus DNA barcoding region (CitationCBOL PWG 2009, p. 12794). Our findings (CitationNicolalde-Morejón et al. 2010), based on single-individual sampling in every species in three Mexican cycad genera, echoed previous indications of the absence of useful variation for DNA barcoding purposes in matK and rbcL, in practically all cycad genera (Little and Stevenson Citation2007; Sass et al. Citation2007). CitationNicolalde-Morejón et al. (2010) analyzed the data with the “Character Attributes Organization System” (CAOS), a method that defines “DNA diagnostics,” i.e. (molecular) character states shared by members of a given taxon and simultaneously absent from comparable groups. For a concise summary of the theoretical reasons behind our decision to use CAOS in a DNA barcoding context, see CitationNicolalde-Morejón et al. (2010, p. 11). For more extensive expositions of the rationale behind character-based methods in DNA barcoding and additional discussion concerning their advantages over phenetic (and other noncharacter-based) approaches, see DeSalle et al. (Citation2005), DeSalle (Citation2006, Citation2007), Rach et al. (Citation2008), Sarkar et al. (Citation2008), Bergmann et al. (Citation2009), and Lowenstein et al. (Citation2009).
Justification for the use of psbK-I as the DNA barcoding region in the present study
In our CAOS-based analysis of the Mexican cycad genera, we found that the chloroplast intergenic spacer psbK-I displayed the best overall single-region performance of all the tested loci (CitationNicolalde-Morejón et al. 2010, p. 9; see ). The measurement behind this qualitative estimation is the percentage of unique, correct species identification under the CAOS analytical regime. Whereas in Dioon, this percentage was 57 (i.e. eight out of 14 species), in the case of Zamia, one-half of the total number of species (i.e. 12 out of 24) was successfully identified. Only in Ceratozamia did the psbK-I region have low levels of discrimination between species (four out of 23, or 17%). However, this value is not very different from the best (ITS2) and the second best (atpF-atpH) performing regions in the same genus, indicating that Ceratozamia is a particularly difficult taxon to address from a DNA barcoding perspective.
Table I. DNA barcoding in species of Zamia occurring in Megamexico: comparative assessment of the number of diagnostic sites for four chloroplast coding and non-coding genome regions and one nuclear region.
CitationNicolalde-Morejón et al. (2010, p. 12) had already noted that the group of Zamia species located in the biogeographic zone known as “Megamexico”—a region that roughly covers Mexico, Guatemala, Belize, El Salvador, and part of Northern Nicaragua (Rzedowski Citation1991)—involves a set of taxa for which there is “an increasingly better understanding of taxonomy and systematics.” This assertion rests upon the contents of a recently published monograph (Nicolalde-Morejón et al. Citation2009), which includes detailed information about the sites of geographic occurrence of populations in every Megamexican Zamia species. In the present paper, we analyzed molecular data obtained from samples collected in the field, directly by our research group, for several of these Zamia populations. On the basis of the aforementioned best performance of psbK-I as a DNA barcoding region in the Mexican species of Zamia (; see also Figure 1 and Table 2 in CitationNicolalde-Morejón et al. (2010)), our analyses are focused only on this region. Although reliance on single-gene matrices might seem restrictive at first sight, the use of the same character-based DNA barcoding strategy in our Zamia population-level data set allows us to establish comparisons with character-based results from a single-individual data set, for the same genomic region (i.e. the Zamia psbK-I matrix already analyzed in CitationNicolalde-Morejón et al. (2010)). This type of comparison is useful to establish whether character-based DNA barcoding analyses are sensitive to slight alignment differences caused by the introduction of new molecular variation through the addition of samples/sequence replicas in matrices. We consider that our results might be of general interest for DNA barcoding researchers interested in character-based methods, whose projects might involve the retrieval of molecular variation from diverse populations within species.
Table II. Sequence data and geographical origin characteristics of sampled Zamia species.
Materials and methods
Sampling of biological materials
We have sampled at least one population for each of the 21 Zamia species recognized from Megamexico, which represents the entire diversity of species for this genus in Megamexico (sensu Rzedowski Citation1991; see Nicolalde-Morejón et al. Citation2009, Citation2010). In total, 63 Zamia populations are represented in our study (for a quantitative description of the distribution of these populations per Zamia species and the total number of samples processed per population, see ). All materials were obtained either from living plants at the National Cycad Collection in the Jardín Botánico “Francisco Javier Clavijero” (JBC) that is administered by the Instituto de Ecología, A.C. (Xalapa, Veracruz, Mexico) or collected in the field (for a list of the collection sites for every Zamia population included in the present study, see ). Leaf tissues from Zamia standleyi, Zamia tuerckheimii, and Zamia prasina were obtained as a gift from the Montgomery Botanical Center (MBC, Miami, FL, USA).
Leaf genomic DNA extraction, PCR amplification, and DNA sequencing
Apart from the leaf samples brought into the laboratory from the field, freshly collected materials were used in total leaf genomic DNA extractions of materials from greenhouses at the JBC. For the extractions, we used either the DNeasy Plant Mini Kit (Qiagen, Valencia, California, USA) or a modified protocol based on a widely employed CTAB DNA extraction procedure (Doyle and Doyle Citation1987). PCR amplification experiments were carried out as reported in one of the pioneer cycad DNA barcoding studies (Sass et al. Citation2007); primers specific for the selected region in the present study—i.e. the intergenic spacer psbK-I—are the same that were used by CitationNicolalde-Morejón et al. (2010, p. 5; these primers are in turn based on CitationLahaye et al. (2008a,Citationb)). Amplification products were observed and photographed after gel electrophoresis in 1% agarose gels stained with ethidium bromide. PCR products were purified directly, using the QIAquick PCR Purification Kit (Qiagen), and automated sequencing was carried out at Macrogen (Seoul, South Korea).
Sequence assembly and alignment
Electropherogram editing for assembly of psbK-I fragments into contigs was performed with the software program Sequencher 4.8 (Gene Codes Corp, Ann Arbor, Michigan, USA). Sequences were deposited on GenBank (see for accession numbers). Sequence alignment of assembled contigs was carried out in BioEdit 7.0.9 (Hall Citation1999), using the ClustalX (Thompson et al. Citation1997) multiple alignment mode function. Aligned matrices were imported into Mesquite 2.73 (CitationMaddison and Maddison 2010) and edited by hand after further visual inspection. Files were saved in Nexus format for subsequent analysis, and are available from the corresponding author upon request.
Character-based analysis of a population-level matrix of psbK-I sequences from Zamia species distributed in Megamexico
As stated in the Introduction, in the present work, we have used the software program CAOS (Sarkar et al. Citation2008) in order to define “DNA diagnostic characters” in our matrices and provided a basis for the molecular identification of samples (i.e. individuals) that potentially belong to the set of Zamia species described in the taxonomic revision by Nicolalde-Morejón et al. (Citation2009). On the basis of the instruction manual for CAOS (“CAOS Documentation and Worked Examples”; Sarkar et al. Citation2008) and the methodological protocol described in CitationNicolalde-Morejón et al. (2010, pp. 4–5), we first built a neighbor-joining phenogram (Saitou and Nei Citation1987) in PAUP 4b10 (Swofford Citation2002), in order to have a starting branching diagram, which was then converted into the definitive “guide tree” using Mesquite 2.73 (CitationMaddison and Maddison 2010). Evidently, the topology of the neighbor-joining tree had no effect on our subsequent analyses; at the same time, in this context, we emphasize that DNA barcoding studies are neither phenetic nor phylogenetic inference studies per se, but rather an identification method. Given that our data sets involve population-level molecular information, we followed the CAOS instruction manual and collapsed all nodes basal to the groups of sequences corresponding to each population; subsequently, these “clades” were nested within higher-order groups, corresponding in each case to a single unique species name. The Mesquite-generated guide tree resulting from these manipulations (see Supplementary material) was then stored in Nexus format for subsequent analysis with the programs specific to the CAOS software package.
We operated the CAOS programs following the protocol described by CitationNicolalde-Morejón et al. (2010, pp. 4–5). As stated in that protocol, determination of DNA diagnostics required the manual revision of the “CAOS-attribute file” and “CAOS-group file” archives generated by the program P-Gnome (Sarkar et al. Citation2008). After this step, characters (“attributes”) with confidence value of 1.00 were selected to construct the actual matrix of DNA diagnostics. Corroboration of attributes was achieved by visually comparing the information of the “CAOS-group file” archives with the original, Mesquite-edited matrices. The matrix of DNA diagnostics constructed after CAOS analyses of the population-level matrix of psbK-I sequences for Zamia species from Megamexico was then compared with the single-individual, DNA diagnostics matrix for the same chloroplast region, which is a subset of the multigene global matrix analyzed by CitationNicolalde-Morejón et al. (2010).
Results and discussion
Why are mitochondrial genome regions not suitable for plant DNA barcoding?
The availability of primer pairs for many different plant genome regions determined that the test for candidate land plant DNA barcodes was initially open to the inclusion of noncoding segments, as well as other coding sequences besides matK and rbcL. However, the interest that molecular biology-oriented botanists have had on the chloroplast genome as a source of DNA barcoding candidate regions did not derive only from the aforementioned record of success achieved with matK and rbcL in plant molecular systematics. An equally important reason behind that research focus has been the observation (made independently by plant molecular biologists not formally involved in plant DNA barcoding) that the cytochrome c oxidase subunit I gene—i.e. the mitochondrial region that had been already established as the “universal DNA barcode” in animals (Hebert et al. Citation2003a,Citationb)—is not suitable for DNA-based identification purposes due to specificities in the molecular evolution of plant mitochondrial genomes (Chase et al. Citation2005). We believe that this context is of interest to readers of this journal, for obvious reasons. Nevertheless, we also think that the analysis of molecular variation in the mitochondrial genome of cycads might prove useful to understand plant genome evolution dynamics in general, and perhaps also have a limited utility for molecular identification purposes. The extent of such molecular variation in the cycads is, at any rate, practically unknown to date (for a rare example in which mitochondrial DNA variation was detected among populations of a cycad species, see Huang et al. Citation2001).
How useful is psbK-I for DNA barcoding in plants?
In the context of the international initiative to find plant DNA barcoding regions, the chloroplast genome intergenic spacer trnH-psbA was proposed early as a potentially ideal plant DNA barcode on the basis of its relatively small size and ease of amplification (Kress et al. Citation2005; for the proposal of trnH-psbA and rbcL as a potential “two-locus universal DNA barcode” in plants, see CitationKress and Erickson 2008). Later on, a couple of additional chloroplast spacers—namely, atpF-atpH and psbK-I—were proposed (by Korean botanist K.-J. Kim; see Pennisi Citation2007) and their utility discussed in international meetings. Ultimately, the three noncoding chloroplast regions just mentioned were included in the CBOL PWG (Citation2009) publication, as a “supplementary set” of barcode sources.
In line with the findings of Sass et al. (Citation2007) concerning the unsuitability of matK and rbcL as DNA barcoding loci in the gymnosperm order Cycadales, CitationNicolalde-Morejón et al. (2010) further established that psbK-I has the best single-locus performance for the Mexican cycads in terms of unique molecular species identification, in comparison with several other chloroplast regions that were entertained as DNA barcoding region candidates in land plants. Interestingly, calculations carried out by the CBOL PWG for their selection of the core plant DNA barcoding regions had also indicated that psbK-I is a good single-gene DNA barcoding locus: this intergenic spacer had only a very slight deficit in terms of percentage discrimination success (i.e. the third criterion for DNA barcoding loci selection; CitationCBOL PWG 2009, p. 12794) compared with trnH-psbA, which occupied the first place (see Figure 1c in CBOL PWG (Citation2009)). Therefore, while our cycad results (CitationNicolalde-Morejón et al. 2010) in fact provide support for this CBOL PWG observation, they also suggest that the selection of matK and rbcL as core DNA barcode regions was perhaps premature. In , we show a quantitative summary of the comparative performance of psbK-I as a (potential) DNA barcoding region in Zamia species from Megamexico, according to the results previously discussed by CitationNicolalde-Morejón et al. (2010).
Comparing DNA barcoding single-individual vs. population-level matrices in Zamia: CAOS analyses can be sensitive to alignment
As shown in , our CAOS-based analysis of the population-level matrix for psbK-I in Zamia retrieves 14 DNA diagnostic sites—matching sites 1, 31, 51, 74, 83, 153, 202, 220, 253, 319, 361, 411, 622, and 664 of the corresponding alignment. The number of DNA diagnostics in this matrix represents four additional sites with respect to the total number of sites obtained with CAOS for a single-individual dataset (CitationNicolalde-Morejón et al. 2010; in Figure 1, the sites in common among the two matrices are marked with shading). Interestingly, out of the nonmatching sites between the two matrices, there is a single character that was present in the single-individual alignment (from CitationNicolalde-Morejón et al. (2010)), but disappeared in the population-level 1 (namely, character 618 in the former data set; this site contains the DNA diagnostic for Zamia lacandona). Further visual inspection of the combined matrix indicates, in addition, the presence of three dimorphic sites—two for Zamia fischeri, and one for Zamia polymorpha, which nevertheless contribute to the distinctive DNA barcode of these two species; and one site in Zamia paucijuga that distinguishes three populations (out of 13), leaving the remaining 10 with no DNA diagnostics to separate them from the most common combination of character states for the CAOS-defined DNA diagnostics.
Table III. Comparison of DNA diagnostic sites between the single-individual and the population-level psbK-psbI data sets for Zamia species from Megamexico.
In summary, our single-gene comparative evaluation of a single-individual vs. a population-level data set demonstrates that a character-based analytical regime can be sensitive to alignment discrepancies. We have observed that sequence alignments can change slightly with the addition of multiple sequence entries for the same populations and/or species; evidently, the incorporation of such information is a realistic possibility in any plant DNA barcoding project for which natural populations are available for study. Therefore, we suggest that the stipulation of a set of “definitive” DNA barcodes for reference in a DNA barcoding library in taxonomic genera, families, etc. should consider comparisons of at least two matrices of aligned sequences: on the one hand, the “minimal matrix” that only contains one sequence per species, and the matrix that contains the largest sampling at the population scale, on the other. In this way, researchers interested in using character-based methods in DNA barcoding projects can have high confidence that they will include most—if not all—naturally occurring nucleotidic variation in the regions they might have selected for the molecular identification of their study taxa.
Prospects for molecular identification and “integrative taxonomy” studies in Neotropical Zamia species
The present work constitutes only a step toward a comprehensive interpretation of the patterns of molecular variation existing in current populations in species of Zamia occurring in the Neotropics. Such interpretation could ultimately help us understand the evolution of the cycad genus Zamia in this biogeographic region. From a strictly taxonomic and systematic standpoint, it is important to keep in mind that differences in the intensity of botanical collections in the biogeographic subregions where Zamia populations occur might be biasing our estimation of such variation (Nicolalde-Morejón F, Stevenson DW, González-Astorga J, Vergara-Silva F, unpublished observations). In turn, we might be currently underestimating the utility of candidate DNA barcoding regions for the construction of reference libraries for various research and applied purposes; this is certainly an issue that should be addressed in subsequent attempts to refine DNA barcoding reference libraries in any cycad genus. In any event, restricting for the moment our scope to Zamia, and with the taxonomic data already at hand (Nicolalde-Morejón et al. Citation2009), we predict that future detailed analysis of molecular data sets for some of the species in this genus could suggest recircumscriptions and/or nomenclatural changes.
In our view, the prospects for these taxonomic advances should be addressed in the conceptual framework associated with the “taxonomic circle” inference procedure advocated by DeSalle et al. (Citation2005). In this regard, some of the best candidates for such ulterior “integrative taxonomy” studies are Zamia loddigesii, Z. paucijuga, and Z. polymorpha, three species of Zamia from Megamexico with wide ranges of morphological and karyotypic variation and ample geographic distributions (Caputo et al. Citation1996; Nicolalde-Morejón et al. Citation2009). Likewise, the Caribbean species of Zamia that are putatively sister taxa to the Mexican species Z. fischeri (Caputo et al. Citation2004) constitute another good candidate group for integrative taxonomic work. Most species of Zamia in the Caribbean clade exhibit morphological diversity between populations on any given island and from island to island. As a result there are, historically, over 35 validly published epithets for a currently recognized six to eight species (Eckenwalder Citation1980; Stevenson Citation1987; Géigel Citation2003). Currently, there are two competing and mutually exclusive hypotheses to explain this. In one scenario, all of the species on a given island have evolved on that island and thus similarities of morphologies from island to island are the result of parallel evolution to the same edaphic features. Thus, each island either has one very variable species or a set of unique species. The contrasting hypothesis is that there are several species, each with a unique distribution pattern among the islands either from vicariance or dispersal or both. As in the case of the Zamia species from Megamexico, we suggest that the key to selecting among these competing hypotheses for the Caribbean Zamias might lie in an integrative approach to inference in which the DNA barcoding information could “break out of the taxonomic circle” (sensu CitationDeSalle et al. 2005, p. 1908) already formed by the biogeographic, morphological—and possibly also the ecological—data points.
Acknowledgements
The authors thank Raúl Jiménez-Rosenberg and Martha Gual for their support during the planning stages of the project. They also acknowledge Alejandro Espinosa de los Monteros and Eduardo Ruíz for data analysis support, and Julia Hernández Villa and Janet Nolasco Soto for technical assistance in the laboratory. Finally, the authors are grateful to the staff at the JBC (Xalapa, Veracruz, Mexico) for access to living specimens in their cycad collections, and to the Montgomery Botanical Center, Miami, FL for supplying samples of some species.
This research project was supported by Mexican Comisión Nacional para el Uso y Conservación de la Biodiversidad (CONABIO) Grant GE004, and US NSF Grants BSR-8607049 and EF-0629817 (to D.W.S).
Declarations of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References
- Bergmann T, Hadrys H, Breves G, Schierwater B. 2009. Character-based DNA barcoding: A superior tool for species identification. Berl Münch Tierärztl Wochenschr. 122:446–450.
- Caputo P, Cozzolino S, Gaudio L, Moretti A, Stevenson D. 1996. Karyology and phylogeny of some Mesoamerican species of Zamia (Zamiaceae). Am J Bot. 83:1513–1520.
- Caputo P, Cozzolino S, De Luca P, Moretti A, Stevenson D. 2004. Molecular phylogeny of Zamia. In: Walters T, Osborne R. editors. Cycad classification: Concepts and recommendations. Oxford: CABI Publishing. p 149–158.
- Hollingsworth PM, Forrest LL, Spouge JL, Hajibabaei M, Ratnasingham S, van der Bank M, , CBOL PWG. 2009. A DNA barcode for land plants. Proc Natl Acad Sci USA. 106:12794–12797.
- Chase MW, Soltis DE, Olmstead RG, Morgan D, Les DH, Mishler BD, Duvall MR, Price RA, Hills HG, Qiu Y-L, Kron KA, Rettig JH, Conti E, Palmer JD, Manhart JR, Sytsma KJ, Michaels HJ, Kress WJ, Karol KG, Clark WD, Hedren M, Gaut BS, Jansen RK, Kim K-J, Wimpee CF, Smith JF, Furnier GR, Strauss SH, Qui-Yun Xiang, Plunkett GM, Soltis PS, Swensen SM, Williams SE, Gadek PA, Quinn CJ, Eguiarte LE, Golenberg E, Learn GHJr, Graham SW, Barrett SCH, Dayanandan S, Albert VA. 1993. Phylogenetics of seed plants: An analysis of nucleotide sequences from the plastid gene rbcL. Ann Mo Bot Gard. 80:528–580.
- Chase MW, Salamin N, Wilkinson M, Dunwell JM, Kesanakurth RP, Haidar N, Savolainen V. 2005. Land plants and DNA barcodes: Short-term and long-term goals. Philos Trans R Soc Lond B Biol Sci. 360:1889–1895.
- Chase MW, Cowan RS, Hollingsworth PM, van der Berg C, Madriñán S, Petersen G, Seberg O, Jørgensen T, Cameron KM, Carine M, Pedersen N, Hedderson TAJ, Conrad F, Salazar GA, Richardson JE, Hollingsworth ML, Barraclough G, Kelly L, Wilkinson M. 2007. A proposal for a standardized protocol to barcode all land plants. Taxon. 56:295–299.
- Cowan RS, Chase MW, Kress WJ, Savolainen V. 2006. 300,000 species to identity: Problems, progress, and prospects in DNA barcoding of land plants. Taxon. 55:611–616.
- DeSalle R. 2006. Species discovery versus species identification in DNA barcoding efforts: Response to Rubinoff. Conserv Biol. 20:1545–1547.
- DeSalle R. 2007. Phenetic and DNA taxonomy; a comment on Waugh. BioEssays. 29:1289–1290.
- DeSalle R, Egan MG, Siddall ME. 2005. The unholy trinity: Taxonomy, species delimitation and DNA barcoding. Philos Trans R Soc Lond B Biol Sci. 360:1905–1916.
- Doyle JJ, Doyle JL. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull. 19:11–15.
- Eckenwalder JE. 1980. Taxonomy of the West Indian cycads. J Arnold Arbor Harv Univ. 61:701–722.
- Erickson DL, Spouge J, Resch A, Weigt LA, Kress WJ. 2008. DNA barcoding in land plants: Developing standards to quantify and maximize success. Taxon. 57:1304–1316.
- Fazekas AJ, Burgess KS, Kesanakurti PR, Graham SW, Newmaster SG, Husband BC, . 2008. Multiple multilocus DNA barcodes from the plastid genome discriminate plant species equally well. PLoS ONE. 3:e2802.
- Fazekas AJ, Kesanakurti PR, Burgess KS, Percy DM, Graham SW, Barrett SCH, . 2009. Are plant species inherently harder to discriminate than animal species using DNA barcoding markers?. Mol Ecol Resour. 1 Suppl: 130–139.
- Ford CS, Ayres KL, Toomey N, Haider N, van Alphen Stahl J, Kelly LJ. 2009. Selection of candidate coding DNA barcoding regions for use on land plants. Bot J Linn Soc. 159:1–11.
- Géigel LG. 2003. Zamiaceae. Flora Rep Cuba Ser A. 8:1–21.
- Hall TA. 1999. BioEdit: A user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser. 41:95–98.
- Hebert PDN, Cywinska A, Ball SL, deWaard JR. Biological identifications through DNA barcodes. Proc R Soc London B Biol Sci. 2003a; 270:313–321.
- Hebert PDN, Ratnasingham S, de Waard JR. Barcoding animal life: Cytochrome c oxidase subunit 1 divergences among closely related species. Proc R Soc London B Biol Sci. 2003b; 270 Suppl: S96–S99.
- Hilu KW, Borsch T, Müller K, Soltis DE, Soltis PS, Savolainen V, . 2003. Angiosperm phylogeny based on matK sequence information. Am J Bot. 90:1758–1776.
- Huang S, Chiang YC, Schaal BA, Chou CH, Chiang TY. 2001. Organelle DNA phylogeography of Cycas taitungensis, a relict species in Taiwan. Mol Ecol. 10:2669–2681.
- Kress WJ, Erickson DL. 2007. A two-locus global DNA barcode for land plants: The coding rbcL gene complements the non-coding trnH-psbA spacer region. PLoS ONE. 2:e508.
- Kress WJ, Wurdack KJ, Zimmer EA, Weigt LA, Janzen DH. 2005. Use of DNA barcodes to identify flowering plants. Proc Natl Acad Sci USA. 102:8369–8374.
- Lahaye R, Savolainen V, Duthoit S, Maurin O, van der Bank M. 2008a. A test of psbK-psbI and atpF-atpH as potential plant DNA barcodes using the flora of the Kruger National Park as a model system (South Africa), Nat Prec, hdl:10101/npre.2008.1896.1 (posted May 2008, retrieved July 2010).
- Lahaye R, van der Bank M, Bogarin D, Warner J, Pupulin F, Gigot G, . DNA barcoding the floras of biodiversity hotspots. Proc Natl Acad Sci USA. 2008b; 105:2923–2928.
- Little DP, Stevenson DW. 2007. A comparison of algorithms for the identification of species using DNA barcodes: Examples for gymnosperms. Cladistics. 23:1–21.
- Lowenstein JH, Amato G, Kolokotronis SO. 2009. The real maccoyii: Identifying tuna fish with DNA barcodes—Contrasting characteristic attributes and genetic distances. PLoS ONE. 4:e7866.
- Maddison WP, Maddison DR. 2010. Mesquite: A modular system for evolutionary analysis. Version 2.73 [Online]. Available at: http://mesquiteproject.org.
- Nicolalde-Morejón F, Vovides AP, Stevenson DW. 2009. Taxonomic revision of Zamia in Mega-Mexico. Brittonia. 61:301–335.
- Nicolalde-Morejón F, Vergara-Silva F, González-Astorga J, Stevenson DW, Vovides AP, Sosa V. 2010. A character-based approach in the Mexican cycads supports diverse multigene combinations for DNA barcoding Cladistics, in press. doi:10.1111/j.1096-0031.2010.00321.x.
- Pennisi E. 2007. Wanted: A barcode for plants. Science. 318:190–191.
- Rach J, DeSalle R, Sarkar IN, Schierwater B, Hadrys H. 2008. Character-based DNA barcoding allows discrimination of genera, species and populations in Odonata. Proc R Soc London B Biol Sci. 275:237–247.
- Rzedowski J. 1991. Vegetación de México. Mexico City: Limusa. p 431.
- Saitou N, Nei M. 1987. The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol Biol Evol. 4:406–425.
- Sarkar IN, Planet PJ, DeSalle R. 2008. CAOS software for use in character-based DNA barcoding. Mol Ecol Resour. 8:1256–1259.
- Sass C, Little DP, Stevenson DW, Specht CD. 2007. DNA barcoding in the Cycadales: Testing the potential of proposed barcoding markers for species Identification of cycads. PLoS ONE. 2:e1154.
- Stevenson DW. 1987. Again, the West Indian Zamias. Fairchild Trop Gard Bull. 42:23–27.
- Swofford DL. 2002. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sunderland, MA: Sinauer Associates.
- Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. 1997. The ClustalX windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tool. Nucleic Acids Res. 25:4876–4882.