• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of transbhomepageaboutsubmitalertseditorial board
Philos Trans R Soc Lond B Biol Sci. Oct 29, 2005; 360(1462): 1925–1933.
Published online Sep 12, 2005. doi:  10.1098/rstb.2005.1724
PMCID: PMC1609231

DNA-based species delineation in tropical beetles using mitochondrial and nuclear markers


DNA barcoding has been successfully implemented in the identification of previously described species, and in the process has revealed several cryptic species. It has been noted that such methods could also greatly assist in the discovery and delineation of undescribed species in poorly studied groups, although to date the feasibility of such an approach has not been examined explicitly. Here, we investigate the possibility of using short mitochondrial and nuclear DNA sequences to delimit putative species in groups lacking an existing taxonomic framework. We focussed on poorly known tropical water beetles (Coleoptera: Dytiscidae, Hydrophilidae) from Madagascar and dung beetles (Scarabaeidae) in the genus Canthon from the Neotropics. Mitochondrial DNA sequence variation proved to be highly structured, with >95% of the observed variation existing between discrete sets of very closely related genotypes. Sequence variation in nuclear 28S rRNA among the same individuals was lower by at least an order of magnitude, but 16 different genotypes were found in water beetles and 12 genotypes in Canthon, differing from each other by a minimum of two base pairs. The distribution of these 28S rRNA genotypes in individuals exactly matched the distribution of mtDNA clusters, suggesting that mtDNA patterns were not misleading because of introgression. Moreover, in a few cases where sequence information was available in GenBank for morphologically defined species of Canthon, these matched some of the DNA-based clusters. These findings demonstrate that clusters of close relatives can be identified readily in the sequence variation obtained in field collected samples, and that these clusters are likely to correspond to either previously described or unknown species. The results suggest that DNA-assisted taxonomy will not require more than a short fragment of mtDNA to provide a largely accurate picture of species boundaries in these groups. Applied on a large scale, this DNA-based approach could greatly improve the rate of species discovery in the large assemblages of insects that remain undescribed.

Keywords: taxonomy, large subunit ribosomal RNA, Madagascar, DNA barcoding, cox1, COI

1. Introduction

A reliable and accessible classification of species is fundamental to research in ecology, evolutionary biology, biodiversity and conservation biology. While ca 1.5 million species have been described to date, this represents only a fraction of the actual diversity on Earth (Tudge 2000; Wilson 2003). Owing to the constant threat of biodiversity loss, there is an increasingly urgent need to accelerate the pace of species discovery and taxonomic databasing (Godfray 2002). Even the routine identification of known species can be difficult, often requiring highly specialized knowledge and representing a limiting factor in ecological studies and biodiversity inventories. In response, recent proposals have called for a more prominent role of efficient DNA-based methods in the delineation and identification of species (Blaxter 2004; Floyd et al. 2002; Hebert et al. 2003a; Tautz et al. 2003). Reactions to such proposals have ranged widely, from strongly supportive (Janzen 2004; Proudlove & Wood 2003; Stoeckle 2003) to vigorously opposed (Lipscomb et al. 2003; Seberg et al. 2003; Wheeler 2004; Will & Rubinoff 2004). The use of DNA-based methods for the delineation and discovery of new species, and thus their broader role in taxonomy, represents an especially contentious issue in this regard. Unfortunately, much of this debate has remained rhetorical, with limited empirical assessment of the benefits and limitations of a DNA-assisted programme of species discovery.

The objective of any method of species delineation, including DNA-based approaches, is to identify reproductively isolated groups of organisms that warrant classification as distinct species. It is widely acknowledged, and reflected in the Linnaean taxonomic system, that living organisms fall into largely discrete groupings recognizable by differences in morphology or other traits. It is then the role of taxonomy to define and name these groupings. However, their recognition may be difficult because diagnostic traits are lacking, or species divergences are very small. Hence, even for cases of biologically distinct species, their delineation will depend on the thoroughness of study and the interpretation of complex, sometimes variable traits. Further, the accuracy of species delineation depends on the degree of sampling, as local variation may affect the conclusions about population separation (Davis & Nixon 1992).

To date, DNA barcoding studies have focussed on the identification of pre-defined species (e.g. Hebert et al. 2003a, 2004b; Hogg & Hebert 2004; Vences et al. 2005; Smith et al. 2005, etc.), and have yet to address the issue of species delineation per se. However, where more than a single individual per species has been sequenced, a minimum threshold of approximately one-tenth of the average p-distance found between well-established species in a lineage has been interpreted as intra-specific variation, while greater divergences are thought to indicate misidentifications of specimens or overlooked cryptic species (Hebert et al. 2004a). These cut-off values roughly correspond to maximum intra-specific divergences in mtDNA of 1–2%, and at the upper bound of this range may include several geographically defined ‘phylogroups’ (Avise & Walker 1999).

The discovery of new, cryptic species from existing, morphologically indiscriminate groups using DNA is neither controversial nor novel (Knowlton 1993), and potential taxonomic revisions inspired by DNA barcoding results have been typically left to experts to resolve on the basis of morphology, behaviour and other features (Hebert et al. 2004b). It is to be expected that a successful global DNA barcoding program would provide a comprehensive barcoding inventory for a majority of described taxa in the foreseeable future, facilitating the systematic discovery of cryptic species. However, with 85% or more of species still unknown to science, a much greater challenge lies in the potential application of DNA-based methods to the discovery and delineation of new species in poorly characterized taxa.

To explore the utility of DNA-based approaches to species recognition in poorly known groups, we evaluated patterns of variation in both mitochondrial and nuclear genes in a broad sample of water beetles collected from Madagascar. These samples exhibited an unknown level of species diversity within the families Dytiscidae and Hydrophilidae. Using the same methods, we also examined specimens from a single genus of Neotropical dung beetles (Canthon). Specimens were collected from various localities in the Neotropics, comprising an unknown number of species in a group that is acknowledged to present difficulties for morphological discrimination. The analysis shows that sequences cluster into cohesive, well-differentiated groups, and identical groups are recovered by both nuclear and mitochondrial markers. Based on multiple lines of evidence, these DNA-based clusters are taken to represent putative species boundaries and could assist with the assembly of a framework for the taxonomy of poorly studied lineages.

2. Methods

(a) Field sampling, selection of specimens and DNA sequencing

Water beetles (Dytiscidae and Hydrophilidae) were collected at five sites in the North and central parts of Madagascar as part of a survey of insect biodiversity in 2004 (Monaghan et al., unpublished). Specimens were collected by sieving through small stream pools, ponds, and packs of leaf litter, and were sorted under a dissecting microscope (10×) into externally distinct morphological groups. Between two and five individuals from each group and each locality were selected for DNA analysis, as a representative sample of the variation of this group. This initial morphological treatment was superficial and was intended to maximize the disparity included in the subset of samples used for sequencing. Specimens of Canthon were obtained using baited pitfall traps from locations in Belize, French Guyana, Ecuador and Costa Rica between 1997 and 2001 (Inward 2003). Additional samples were collected from Belize in 2004 (L. Powell, MSc, Imperial College London, 2004). They were assigned to the genus Canthon based on a phylogenetic analysis combining them with unpublished sequences for most major groups of Scarabaeinae that also included five species of Canthon: C. doesburgi, C. indigaceus, C luteicollis, C. smaragdulus and C. viridis, plus the closely related Scybalocanthon pygidialis (GenBank accessions: AY131633-7 for 28S, AY131814-7 for cox1, plus AY131673 and AY131849 for Scybalocanthon).

Genomic DNA extraction was performed using Wizard SV 96-well plates (Promega, UK). For both groups, a ca 700 bp fragment of 28S rRNA was amplified using primers FF and DD (Inward 2003). Fragments of cox1 were amplified with primers Pat and Jerry (Simon et al. 1994) for Canthon (800 bp) or with LCO1490 and HCO2198 (Folmer et al. 1994) for water beetles (660 bp). Sequencing was performed in both directions using a BigDye v. 2.1 terminator reaction with the same primers used for PCR. Sequences were analysed on an ABI3730 automated sequencer and forward and reverse strands were assembled in Sequencher software. The 28S fragment was length-variable and was aligned separately for the two datasets using BlastAlign (Belshaw & Katzourakis 2005). Cox1 was not length-variable for either group.

(b) Tree construction

Parsimony trees were obtained with PAUP v. 4b10 (Swofford 2002), with gap characters treated as a ‘fifth character state’, and branch length optimized under accelerated transformation. Heuristic searches were performed using TBR branch swapping and 100 replicates. We performed 1000 random addition replicates saving only a single tree in each case. Because the dataset contained many identical or very similar haplotypes, a large number of trees were found, one of which was selected arbitrarily for further analysis. To calculate Bremer Support (Bremer 1994), constraint files for parsimony searches enforcing the absence of the focal nodes were produced with TreeRot v. 2.0a. Bremer Support values of 0 indicate unresolved nodes which would be collapsed in a strict consensus of all shortest trees. Trees were rooted with sequences from related Carabidae taken from GenBank in the case of water beetles, and using a sequence from the Canthon dataset generated here for rooting the 28S tree. The single species of Scybalocanthon was used as the outgroup to root the tree of Canthon.

(c) Variation in cox1

We examined cox1 variation within and among clusters of sequences (see §3) using analysis of molecular variance (AMOVA) (Excoffier et al. 1992) of pairwise differences as implemented in Arlequin v.2.000 (Schneider et al. 2000). We used a two-level hierarchical analysis to partition total cox1 variation into within-cluster and among-cluster covariance components. Individuals were included into a single cox1 cluster if they exhibited identical sequences in the 28S gene (below) or, if no 28S sequence was available for a specimen, the cox1 sequence grouped within these clusters. The fixation index calculated among groups (analogous to a population genetics FST) was tested for significance using Arlequin v. 2.000. Groups with only one representative cox1 sequence (e.g. WB-6, WB-11, see figure 1) were excluded from analysis because within-group variation could not be measured.

Figure 1
Parsimony trees for water beetles using 28S (a) and cox1 (b) sequences. Cluster names are given to the right of groups. Italics in 28S denote groups for which no cox1 data were available, and vice versa. Bremer Support values (see table 2) are reported ...

3. Results

A total of 75 and 71 individuals were included in the analysis of water beetles and Canthon, respectively. Sequencing of 28S rRNA was successful for 63 and 62 specimens, and aligned matrices contained 699 and 746 characters in the respective groups. In the water beetle dataset, we detected 16 different 28S genotypes. Four genotypes were present in only a single individual, and the remaining occurred in groups ranging in size from two to nine individuals (figure 1a). Sequences differed from one another by a minimum of two nucleotides, e.g. an AC insertion separated WB-4 and WB-5 (figure 1a). DNA sequencing for cox1 revealed 42 haplotypes. Parsimony tree searches uncovered 14 clusters of similar cox1 sequences, plus five isolated sequences without close relatives (‘singletons’) (figure 1b). Results for Canthon were very similar. There were 12 different 28S genotypes and all but two were represented by >1 individual (figure 2a). The 46 cox1 haplotypes grouped into 12 clusters, with two singletons (figure 2b).

Figure 2
Parsimony trees for Canthon, as in figure 1.

The clustering of cox1 sequences in the parsimony trees showed complete congruence with the 28S genotypes for both the water beetle and Canthon datasets. Closely related cox1 haplotypes all exhibited the same 28S genotype and none of the groups defined by 28S genotypes were polyphyletic in the cox1 tree (figures 1 and and2).2). It was not possible to judge incongruence in the 12 water beetles for which the 28S sequencing had failed (figure 1b). Equally, the Canthon dataset included missing sequences in both the 28S and cox1 datasets, although there was perfect congruence for the 50 individuals sequenced for both genes (called ‘core terminals’, below).

Sequences making up the cox1 clusters were very similar to each other, but very different from other clusters. Based on pairwise differences in AMOVA, within-group variation accounted for only 4.1% of the total variation in the dataset in Canthon, and only 2.5% in water beetles (table 1). Absolute divergence (uncorrected p-distance) of sequences within clusters ranged from 0 to 2%, whereas the mean divergence between clusters was 10 and 19% for Canthon and water beetles, respectively (table 2). These patterns of divergence are similar to those reported for intra- versus inter-species comparisons in other barcoding studies (Hebert et al. 2003b). The findings also appear to indicate that clusters in Canthon were more closely related to one another than were the clusters in the more diverse sample of water beetles.

Table 1
Cox1 variation among and within clusters (figures 1 and and2)2) measured with AMOVA (Excoffier et al. 1994).
Table 2
Mean uncorrected p-distances for cox1 within and among clusters.

The monophyly of the cox1 clusters was highly supported. When Bremer Support was considered separately for three categories of node levels (tip nodes within a cluster, nodes immediately subtending a cluster, and nodes defining basal relationships between the clusters), the majority of total tree support was derived from nodes immediately below (i.e. defining) the clusters (88% of total Bremer Support for water beetles, 80% for Canthon; table 3). Tip nodes within clusters, and basal nodes had low support (figures 1 and and2,2, table 3). To test for congruence in phylogenetic signal, a simultaneous analysis of cox1 and 28S datasets was conducted for Canthon. Because some individuals were successfully sequenced for only one fragment, we either combined all terminals in a single ‘supermatrix’ (cox1+28S all terminals, n=71), or removed all terminals which were not complete for either one of the two gene partitions (core terminals, n=50; table 3). Total tree support was much higher in the analysis of core terminals as compared to when all individuals were included in the combined analysis (figure 3, table 3). However, the incongruence length difference was minimal, indicating that the drop is not due to conflict between both markers but the reduced discriminatory power of the dataset once a large number of missing entries is included in the data matrix. In all cases, total tree support was higher for cox1 than 28S for all analyses, regardless of whether all Canthon or only core individuals were used in the calculation (table 3), presumably due to the larger number of character changes in the former.

Figure 3
Tree from the combined data matrix (28S-cox1 all terminals; table 2) for Canthon. Bremer Support values are reported above branches for the tree pictured. Values in parentheses are from the analysis of only the core terminals (50 individuals; table 2 ...
Table 3
Parsimony analysis on the four datasets produced trees with minimal length and homoplasy values as indicated.

4. Discussion

(a) The partitioning of genetic variation

The most striking result of the DNA analysis was the strong clustering of the sequence variation, with comparably large distances between groups of closely related sequences. In addition, these clusters showed remarkably high levels of nodal support for their monophyly according to cox1 and 28S genes. Support in the combined analysis was even higher and is essentially the sum of the individual partitions, showing a high degree of congruence for the two markers. Nodes defining the clusters included >80% of the total Bremer Support provided by the datasets, although they specify only one-third or less of the total number of nodes in the tree. Variation in the 28S nuclear gene, while showing overall fewer character changes, was also strongly clustered. Genotypes were shared by many individuals and were separated by a minimum of a single base pair (bp) in Canthon and a minimum 2 bp insertion differentiating two water beetle genotypes (e.g. WB-4 and WB-5). For the cox1 variation, >95% of variation occurred among these different clusters, with only ca 2.5–4.5% of the variation within these groups. Remarkably, the observed pattern of clustering was very similar in the two groups of beetles, even though they are composed of members of two different suborders Adephaga and Polyphaga, and obtained from different parts of the world (Madagascar and the Neotropics).

A further key finding of this study is that the nuclear and mitochondrial gene data were completely congruent for both samples of beetles. Individuals were grouped into clusters in the exact same way whether based on the cox1 or 28S genotypes. Cox1 sequences were more variable than 28S, but multiple cox1 haplotypes in a cluster were monophyletic with respect to a single 28S genotype. For the water beetles this may be biased by the fact that these were field-samples from a given locality, raising uncertainty as to whether sister or even closely related taxa co-occur and were collected. For Canthon, by contrast, a single lineage was deliberately chosen from a larger sample of dung beetle communities (unpublished), and a wider sampling range covered, in order to increase the probability of sampling sister taxa. Notably, even in the case of the very closely related Can-1 and Can-2 clusters, where only a 2 bp insertion segregated 28S genotypes, the cox1 phylogenetic analysis was completely congruent with separation into two distinct groups.

(b) What is the nature of the clusters?

Several lines of evidence suggest that the clusters identified in this study represent distinct species, rather than any other level of hierarchical organization. Phenetic sequence divergence in mtDNA within these groups never exceeded 2% and usually was much lower, whereas divergence between the clusters was often greater by more than an order of magnitude. This is in general agreement with empirical levels of divergence found between species in phylogeographic analyses (Avise & Walker 1999) and barcoding studies (Hebert et al. 2003b). For Canthon, the existing molecular phylogenetic and taxonomic framework also supports this conclusion, as clusters from our study necessarily represent subgroups below the genus level, and GenBank database entries of various species of Canthon correspond to different clusters in our analysis.

Beyond simple comparisons of phenetic divergence, inspection of the phylogenetic trees revealed a striking shift in branch length, long branches leading to subtending nodes and short branches within tip clusters, as seen in other studies of closely related species (Barraclough unpublished; Pons et al. unpublished). In addition, the strong support for the sub-cluster nodes (and no other node level) also confirms the unique status of this particular level of hierarchy in the trees. Whether or not this represents the species boundary remains to be investigated further. Ultimately, additional information, such as field studies of the sampled populations and broader genetic surveys including sister species, is required to confirm that the groups defined by these nodes are defining the species. However, the species category does take up a special place in the taxonomic hierarchy as the only ‘natural’ level of organization of the classificatory system, in contrast to the higher levels, such as genera and families (Cracraft 1983). It is our hypothesis that the transition in branching patterns, and the shift from strong to negligible branch support, represents a genetic signature of this unique level of organization.

(c) Methodological issues of species delineation from sequence data

Existing approaches to species delineation from sequence variation alone have been applied mainly to very small organisms, such as prokaryotes or soil nematodes, in which morphological discrimination is difficult or impossible (Floyd et al. 2002; Gregory & DeSalle 2005). In the case of nematodes, Molecular Operational Taxonomic Units (MOTUs) have been assigned based solely on sequence divergence (Blaxter 2004). While there may be no better way to classify these organisms to date, it remains unclear how these MOTUs correspond to evolutionarily differentiated groups, and how meaningful they are with respect to species cohesion. While the observation of large inter- and low intra-species variation promises easy identification of described species and the discovery of many cryptic species (Hebert et al. 2003b; Hebert et al. 2004b), there is concern regarding variability in the threshold values both between individual sister species pairs and among major lineages (DeSalle et al. 2005; Moritz & Cicero 2004).

In part, the problem of quantitative species delimitation could be overcome by searching for diagnostic character variation (Cracraft 1983), or complex character combinations (DeSalle et al. 2005) to define the species limits based on quantitative methods (Sites & Marshall 2003). These tests, which are rooted in the phylogenetic species concept, establish whether a priori populations can be ‘aggregated’ into a single species based on the distribution of characters or tree topology. These methods may not be practical when applied to large-scale species discovery and barcoding studies, where the cohesion of populations is unknown and broad sampling across species' geographic ranges may not be possible.

A possible alternative to aggregation methods is to interpret branch length itself as being suggestive of species boundaries, assuming that the long branches defining the clusters could only have arisen if populations diverged longer than around Ne (effective population size) generations ago (Hudson & Coyne 2002). Appropriate methods for estimating these shifts include Templeton's statistical parsimony analysis that partitions the variation into homoplastic (i.e. long branches) and non-homoplastic (short branches) variation (Templeton 2001). Similarly, it may be possible to statistically differentiate the shifts of lineage branching from interspecific, long branches to intraspecific, short branches using maximum likelihood methods (Barraclough, unpublished; Pons et al. submitted). A further approach could be based on population genetics analyses. It is possible to interpret the AMOVA results used to calculate intra- versus inter-cluster variation in a way analogous to F-statistics (Wright 1978). In this scenario, FST>0.95 for both water beetles and Canthon datasets, meaning that >95% of the total genetic variation in the dataset arises from differences among groups. A threshold of <5% within-group variation seems a reasonable means of minimizing the chance of overlooking distinct taxa. As an example, combining Can-1 and Can-2 into a single group (‘Can1-2’) and recalculating AMOVA statistics results in a 15.6% value for within-group variation, as opposed to the 2.5% value when these two groups are considered as separate entities (table 1). This demonstrates the stringent clustering of the data, and provides a simple procedure to identify groups that have been grouped incorrectly.

(d) Conclusions and prospects

The aim of the present study was to investigate the efficacy of short sequence fragments for use in the discovery, delineation and routine identification of species. The analysis neither strictly constitutes a test of whether DNA can delimit pre-defined species, nor was it an analysis of Type I or Type II errors of species assignment (Quicke 2004). Instead, we used parsimony analysis to simultaneously examine a large number of sequences to assess patterns of variation in nature, and enquired whether this conforms to expectations of clustering at the species level of the biological hierarchy. The results were striking: sequence variation clustered very strongly for both nuclear and mitochondrial markers, and nodes defining these clusters were well supported whereas tip nodes, connecting closely related individuals, were not. Whether all the clusters we identified correspond to pre-existing, named species remains to be tested, and would require the input of specialists experienced in these taxonomic groups. Notably, comparisons with GenBank sequences indicated that at least three of the clusters identified here do indeed correspond to named species of Canthon.

The analytical approach will allow these clusters to be delineated objectively and repeatedly by anyone using the sequence data matrix. As a result, DNA data can form the basis of testable taxonomic hypotheses that could be examined with additional types of data in the future. The benefits of this approach are manifold: it provides a rapid division into probable groups of reproductively isolated individuals and generates more direct links to their evolutionary past; it will facilitate the determination of distinctive morphological features (i.e. through a focussed comparison of pre-delineated groups); it would allow the study of these putative species to proceed even while formal description is pending; it would link individuals from the same species collected in different localities or in different studies in a way that arbitrary designations (e.g. ‘Canthon sp.1’) do not; and it would immediately provide the data needed for future DNA barcoding identification.

The results of the present study are based on a relatively small number of species, but nonetheless demonstrate the general feasibility of using DNA-based methods in the large-scale delineation and discovery of new species, even in poorly characterized groups. While more research is needed to establish the best approach for species delineation using DNA (e.g. through phylogenetic or coalescent methods, with phenetic barcode ‘thresholds’, or some combination thereof), it is becoming evident that DNA methods present a promising new means of assessing and identifying biological diversity in some of the most species rich taxa and environments on Earth. There is reason for optimism that, if fully developed and implemented on a broad scale, DNA-based tools such as those examined here may provide the first opportunity for creating a comprehensive inventory of life.


We are grateful to Daegan Inward, Richard Davies and Liz Powell for collecting Canthon dung beetles; to David Lees, Ravomiarana Ranaivosolo, Pierre Razafindraire, Roger Andriamparany and Doug Ottke for assistance with water beetle collection; to Ruth Wild and Miranda Elliot for laboratory analysis; and to Silvia Fabrizi for mounting specimens. In Madagascar, we thank York Pareik at King de la Piste and Madame Liva and Benjamin Andriamihaja at MICET (Madagascar Institut pour la Conservation des Ecosystèmes Tropicaux). Tim Barraclough and an anonymous referee provided helpful comments on the manuscript.


One contribution of 18 to a Theme Issue ‘DNA barcoding of life’.

Present address: Zoologische Staatssammlung, Muenchhausenstrasse 21, 81247 Munich, Germany.


  • Avise J.C, Walker D.E. Species realities and numbers in sexual vertebrates: perspectives from an asexually transmitted genome. Proc. Natl Acad. Sci. 1999;96:992–995. 10.1073/pnas.96.3.992 [PMC free article] [PubMed]
  • Belshaw R, Katzourakis A. BlastAlign: a program that uses blast to align problematic nucleotide sequences. Bioinformatics. 2005;21:122–123. 10.1093/bioinformatics/bth459 [PubMed]
  • Blaxter M.L. The promise of a DNA taxonomy. Phil. Trans. R. Soc. B. 2004;359:669–679. 10.1098/rstb.2003.1447 [PMC free article] [PubMed]
  • Bremer K. Branch support and tree stability. Cladistics. 1994;10:295–304. 10.1111/j.1096-0031.1994.tb00179.x
  • Cracraft J. Species concept and speciation analysis. Curr. Ornithol. 1983;1:159–187.
  • Davis J.I, Nixon K.C. Populations, genetic variation, and the delimitation of phylogenetic species. Syst. Biol. 1992;41:421–435.
  • DeSalle R, Egan M.G, Siddall M. The unholy trinity: taxonomy, species delimitation and DNA barcoding. Phil. Trans. R. Soc. B. 2005;360:1905–1916. 10.1098/rstb.2005.1722 [PMC free article] [PubMed]
  • Excoffier L, Smouse P, Quattro J. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics. 1992;131:479–491. [PMC free article] [PubMed]
  • Floyd R, Abebe E, Papert A, Blaxter M. Molecular barcodes for soil nematode identification. Mol. Ecol. 2002;11:839–850. 10.1046/j.1365-294X.2002.01485.x [PubMed]
  • Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 1994;3:294–299. [PubMed]
  • Godfray H.C.J. Challenges for taxonomy—the discipline will have to reinvent itself if it is to survive and flourish. Nature. 2002;417:17–19. 10.1038/417017a [PubMed]
  • Gregory T.R, DeSalle R. Comparative genomics in prokaryotes. In: Gregory T.R, editor. The evolution of the genome. Elsevier; San Diego: 2005. pp. 585–675.
  • Hebert P.D.N, Cywinska A, Ball S.L, DeWaard J.R. Biological identifications through DNA barcodes. Proc. R. Soc. B. 2003a;270:313–321. 10.1098/rspb.2002.2218 [PMC free article] [PubMed]
  • Hebert P.D.N, Ratnasingham S, DeWaard J.R. Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc. R. Soc. B. 2003;270(Suppl. 1):S96–S99. 10.1098/rspb.2002.2218 [PMC free article] [PubMed]
  • Hebert P.D.N, Penton E.H, Burns J.M, Janzen D.H, Hallwachs W. Ten species in one: DNA barcoding reveals cryptic species in the Neotropical skipper butterfly Astraptes fulgerator. Proc. Natl Acad. Sci. 2004a;101:14 812–14 817. 10.1073/pnas.0406166101 [PMC free article] [PubMed]
  • Hebert P.D.N, Stoeckle M.Y, Zemlak T.S, Francis C.M. Identification of birds through DNA barcodes. PLOS Biol. 2004b;2:1657–1663. 10.1371/journal.pbio.0020312 [PMC free article] [PubMed]
  • Hogg I.D, Hebert P.D.N. Biological identification of springtails (Hexapoda: Collembola) from the Canadian Arctic, using mitochondrial DNA barcodes. Can. J. Zool. 2004;82:749–754. 10.1139/z04-041
  • Hudson R.R, Coyne J.A. Mathematical consequences of the genealogical species concept. Evolution. 2002;56:1557–1565. [PubMed]
  • Inward, D. G. 2003 The evolution of dung beetle assemblages. Ph.D. thesis, Imperial College, London.
  • Janzen D.H. Now is the time. Phil. Trans. R. Soc. B. 2004;359:731–732. 10.1098/rstb.2003.1444 [PMC free article] [PubMed]
  • Knowlton N. Sibling species in the sea. Trends Ecol. Evol. 1993;24:189–216.
  • Lipscomb D, Platnick N, Wheeler Q. The intellectual content of taxonomy: a comment on DNA taxonomy. Trends Ecol. Evol. 2003;18:65–68. 10.1016/S0169-5347(02)00060-5
  • Moritz C, Cicero C. DNA barcoding: promise and pitfalls. PLOS Biol. 2004;2:1529–1531. 10.1371/journal.pbio.0020354 [PMC free article] [PubMed]
  • Proudlove G, Wood P.J. The blind leading the blind: cryptic subterranean species and DNA taxonomy. Trends Ecol. Evol. 2003;18:272–273. 10.1016/S0169-5347(03)00095-8
  • Quicke D.J.L. The world of DNA barcoding and morphology—collision or synergism and what of the future. Systematist. 2004;23:8–12.
  • Schneider S.D, Roessli D, Excoffier L. Genetics and Biomery Laboratory, University of Geneva; Switzerland: 2000. ARLEQUIN version 2.000: a software for population genetics data analysis.
  • Seberg O, Humphries C.J, Knapp S, Stevenson D.W, Petersen G, Scharff N, Andersen N.M. Shortcuts in systematics? A commentary on DNA-based taxonomy. Trends Ecol. Evol. 2003;18:63–65. 10.1016/S0169-5347(02)00059-9
  • Simon C, Frati F, Beckenbach A, Crespi B, Liu H, Flook P. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am. 1994;87:651–701.
  • Sites J.W, Marshall J.C. Delimiting species: a Renaissance issue in systematic biology. Trends Ecol. Evol. 2003;18:462–470. 10.1016/S0169-5347(03)00184-8
  • Smith M.A, Fisher B.L, Hebert P.D.N. DNA barcoding for effective biodiversity assessment of a hyperdiverse arthropod group: the ants of Madagascar. Phil. Trans. R. Soc. B. 2005;360:1825–1834. 10.1098/rstb.2005.1714 [PMC free article] [PubMed]
  • Stoeckle M. Taxonomy, DNA, and the barcode of life. Bioscience. 2003;53:796–797.
  • Swofford D.L. Sinauer Associates; Sunderland, MA: 2002. PAUP*: Phylogenetic analysis using parsimony. Version 4.0b.
  • Tautz D, Arctander P, Minelli A, Thomas R.H, Vogler A.P. A plea for DNA taxonomy. Trends Ecol. Evol. 2003;18:70–74. 10.1016/S0169-5347(02)00041-1
  • Templeton A.R. Using phylogeographic analyses of gene trees to test species status and processes. Mol. Ecol. 2001;10:779–791. 10.1046/j.1365-294x.2001.01199.x [PubMed]
  • Tudge C. Oxford University Press; Oxford: 2000. The variety of life.
  • Vences M, Thomas M, Bonett R.M, Vieites D.R. Deciphering amphibian diversity through DNA barcoding: chances and challenges. Phil. Trans. R. Soc. B. 2005;360 10.1098/rstb.2005.1717 [PMC free article] [PubMed]
  • Wheeler Q.D. Taxonomic triage and the poverty of phylogeny. Phil. Trans. R. Soc. B. 2004;359:571–583. 10.1098/rstb.2003.1452 [PMC free article] [PubMed]
  • Will K.W, Rubinoff D. Myth of the molecule: DNA barcodes for species cannot replace morphology for identification and classification. Cladistics. 2004;20:47–55. 10.1111/j.1096-0031.2003.00008.x
  • Wilson E.O. The encyclopedia of life. Trends Ecol. Evol. 2003;18:77–80. 10.1016/S0169-5347(02)00040-X
  • Wright S. University of Chicago; Chicago: 1978. Evolution and the genetics of populations.

Articles from Philosophical Transactions of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...