Content uploaded by Gianni Barcaccia
Author content
All content in this area was uploaded by Gianni Barcaccia on Jan 03, 2014
Content may be subject to copyright.
Biodiversity studies in Phaseolus species by DNA
barcoding
Silvia Nicolè, David L. Erickson, Daria Ambrosi, Elisa Bellucci, Margherita Lucchin,
Roberto Papa, W. John Kress, and Gianni Barcaccia
Abstract: The potential of DNA barcoding was tested as a system for studying genetic diversity and genetic traceability in
bean germplasm. This technique was applied to several pure lines of Phaseolus vulgaris L. belonging to wild, domesticated,
and cultivated common beans, along with some accessions of Phaseolus coccineus L., Phaseolus lunatus L., and Vigna un-
guiculata (L.) Walp. A multilocus approach was exploited using three chloroplast genic regions (rbcL,trnL, and matK),
four intergenic spacers (rpoB-trnC,atpBrbcL,trnT-trnL, and psbA-trnH), and nuclear ITS1 and ITS2 rDNA sequences. Our
main goals were to identify the markers and SNPs that show the best discriminant power at the variety level in common
bean germplasm, to examine two methods (tree based versus character based) for biodiversity analysis and traceability as-
says, and to evaluate the overall utility of chloroplast DNA barcodes for reconstructing the origins of modern Italian vari-
eties. Our results indicate that the neighbor-joining method is a powerful approach for comparing genetic diversity within
plant species, but it is relatively uninformative for the genetic traceability of plant varieties. In contrast, the character-based
method was able to identify several distinct haplotypes over all target regions corresponding to Mesoamerican or Andean ac-
cessions; Italian accessions originated from both gene pools. On the whole, our findings raise some concerns about the use
of DNA barcoding for intraspecific genetic diversity studies in common beans and highlights its limitations for resolving ge-
netic relationships between landraces and varieties.
Key words: Phaseolus spp., plastid DNA, internal transcribed spacers, DNA barcoding, varietal groups, single-nucleotide
polymorphisms.
Résumé : Les auteurs ont exploré le potentiel des codes barres génétiques pour étudier la diversité et la traçabilité généti-
ques au sein du germoplasme du haricot. Cette technique a été employée sur plusieurs lignées pures sauvages, domestiquées
et cultivées du Phaseolus vulgaris, ainsi qu’à quelques accessions du P. coccineus,duP. lunatus et du Vigna unguiculata.
Une approche multilocus a été exploitée au moyen de trois régions géniques chloroplastiques (rbcL,trnL et matK), de quatre
espaceurs intergéniques (rpoB-trnC,atpB-rbcL,trnT-trnL et psbA-trnH), et les séquences nucléaires ITS1 et ITS2 de
l’ADNr. Les buts principaux étaient d’identifier les marqueurs et SNP qui offraient le pouvoir discriminant le plus grand en-
tre les variétés chez le haricot, de comparer deux méthodes (fondée sur les arbres ou les caractères) pour l’analyse de la bio-
diversité et pour des essais de traçabilité, et d’évaluer l’utilité globale des codes barres d’ADN chloroplastique pour retracer
l’origine des variétés italiennes modernes. Les résultats obtenus montrent que la méthode NJ constitue une approche puis-
sante pour comparer la diversité génétique au sein des espèces, mais qu’elle s’avère relativement peu informative pour ce
qui est de la traçabilité génétique des cultivars. Au contraire, la méthode basée sur l’examen des caractères a permis d’iden-
tifier plusieurs haplotypes distincts pour toutes les régions étudiées au sein des accessions mésoaméricaines ou andéennes,
ces deux pools génétiques étant la source des accessions italiennes. Globalement, ces observations soulèvent des interroga-
tions sur l’emploi des codes barres génétiques pour des études de diversité génétique intraspécifique chez le haricot et souli-
gnent les limites de cet outil pour la résolution des relations génétiques entre variétés de pays et cultivars.
Mots‐clés : Phaseolus spp., ADN plastidique, espaceurs internes transcrits, codes barres génétiques, groupes variétaux, poly-
morphisme mononucléotidique.
[Traduit par la Rédaction]
Received 28 December 2010. Accepted 19 February 2011. Published at www.nrcresearchpress.com/gen on 21 July 2011.
Paper handled by Associate Editor Paolo Donini.
S. Nicolè, D. Ambrosi, M. Lucchin, and G. Barcaccia. Department of Environmental Agronomy and Crop Science, Università degli
Studi di Padova, Via dell'Università 16 –Campus of Agripolis, 35020 Legnaro, Padova, Italy.
D.L. Erickson and W.J. Kress. Department of Botany and Laboratory of Analytical Biology, National Museum of Natural History,
Smithsonian Institution, P.O. Box 37012, Washington, DC 20013-7012, USA.
E. Bellucci. Department of Environmental Sciences and Crop Production, Università Politecnica delle Marche, Ancona, Via Brecce
Bianche, 60131 Ancona, Italy.
R. Papa. Department of Environmental Sciences and Crop Production, Università Politecnica delle Marche, Ancona, Via Brecce Bianche,
60131 Ancona, Italy; Cereal Research Centre, Agricultural Research Council, S.S. 16, Km 675, 71122 Foggia, Italy.
Corresponding author: Gianni Barcaccia (e-mail: gianni.barcaccia@unipd.it).
529
Genome 54: 529–545 (2011) doi:10.1139/G11-018 Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Introduction
The genomic advances of the last decade have provided
the technological tools for the development of a universal,
DNA-enhanced system of taxonomy suitable for addressing
the current “biodiversity crisis”that requires innovative and
informative technologies (Tautz et al. 2003). DNA barcoding
has been proposed as a cost-effective technology (Hebert et
al. 2003) able to contribute to the study of biodiversity,
which, until recently, relied primarily on morphology in the
Linnaean classification system. DNA-based methods are fast
and not limited by taxonomic impediments such as missing
morphological features of a particular life stage (e.g., eggs
and juvenile forms) (Velzen et al. 2007), missing body parts
(Wong and Hanner 2008), or homoplasy of some characters
(Vences et al. 2005). Although the application of DNA fin-
gerprinting as an identification tool is not a new idea, DNA
barcoding has earned remarkable success attributable to the
standardization of the procedure by the use of a universal
barcode sequence across a wide range of organisms (Hebert
et al. 2004). The proposal of using DNA barcoding as a new
identification tool turned on a heated debate between the ad-
vocates and the opponents to the potential uses of this techni-
que because of some theoretical and methodological
weakness (Will and Rubinoff 2004; Will et al. 2005; Hicker-
son et al. 2006). The ambitious idea of using the polymor-
phism information in a short sequence of DNA to
distinguish every species in the world has already been trans-
lated into a powerful tool in the animal kingdom (Ward et al.
2005), even if other studies demonstrated that some taxa are
problematic for the application of DNA barcoding (Brower
2006; Meier et al. 2006; Wiemers and Fiedler 2007). Regard-
ing the utility of the approach for land plants, biologists have
been slower in adapting a universal gene region as a barcode
because of the difficulty of finding a region analogous to the
animal COI gene (also known as cox1). Recently, the CBOL
Plant Working Group (2009) recommended the combination
of the chloroplast genic regions rbcL and matK as the plant
barcode. This core, two-locus DNA-barcoding approach has
been proposed as a universal framework for the routine use
of DNA sequence data to identify specimens and contribute
to the discovery of unknown species of land plants. In the
same publication, a minority position of the CBOL Plant
Working Group supported the inclusion of the trnH-psbA in-
tergenic spacer in the plant barcode following earlier publica-
tions that outlined practical difficulties related to the
acquisition of matK sequences (Kress and Erickson 2007; Fa-
zekas et al. 2008). The combination of the rbcL gene with
the trnH-psbA intergenic spacer, a more rapidly evolving re-
gion than rbcL and matK, seems to be a valid alternative to a
simple two-locus model: the former distinguishes distantly re-
lated plants, and the latter recognizes closely related sister
species or species groups that have only recently diverged
(Kress and Erickson 2007). Finally, even if organellar DNA
sequences are used as the main source of information for a
barcoding system, then one or more nuclear genes may also
be required for the supplemental analysis of hybrids. Nuclear
genes such as internal transcribed spacers (ITS), which are
frequently used for phylogenetic analyses and single-copy nu-
clear regions, have been considered by some research groups
(as, for instance, Cowan et al. 2006), even if with some re-
serves (see also http://www.kew.org/barcoding/).
Several DNA fingerprinting and genotyping assays based
on molecular markers such as RFLPs and SNPs have been
developed in the past and are still used in plant genetics and
breeding (Mohler and Schwarz 2008). DNA barcoding could
provide an additional system to identify not only species but
also crop varieties and germplasm resources to assess the dis-
tinctiveness of genotypes and relatedness among genotypes
(Pallottini et al. 2004). Assessment of the potential of DNA
barcoding to distinguish between plant varieties of agri-food
interest would be valuable for both breeders and farmers.
Whereas the utility of DNA barcoding in species identifica-
tion has been widely investigated, the intraspecific discrimi-
nation of single varietal genotypes, such as clones, pure
lines, and hybrids, has been poorly investigated, and few
studies have focused on the use of DNA barcoding as a suffi-
ciently informative technique to be exploited for the genetic
identification of closely related crop varieties (Tsai et al.
2008).
Our work focuses on the application of DNA barcoding to
cultivated bean germplasm as a new tool for discrimination
among Phaseolus spp. and, most of all, for identification of
Phaseolus vulgaris L. varieties. Phaseolus is a genus in the
family Fabaceae, the third largest family of flowering plants
(Gepts et al. 2005), and it represents multiple domestications
of distinct, but related, species and multiple populations
within the same species, e.g., as found in P. vulgaris and
Phaseolus lunatus L. The original natural distribution of this
species, before its introduction throughout Europe and Africa
in the post-Columbian period, consists of a fragmented area
throughout Central and South America. On the basis of the
available data, at least two primary centers of origin have
been recognized: a relatively heterogeneous one in the Andes
(Colombia, Ecuador, Peru, Bolivia, Chile, and Argentina) and
a more homogeneous one in Mesoamerica (primarily Mexico,
Guatemala, Honduras, El Salvador, Nicaragua, and Costa
Rica). These two centers of origin are called the Andean and
Mesoamerican gene pools, respectively (Chacón et al. 2005;
Papa et al. 2006).
In this paper, we present results on the use of DNA bar-
coding in several pure lines of wild, domesticated, and culti-
vated common beans, using both coding and noncoding
regions from the chloroplast and nuclear genomes. Our ob-
jectives were the following: (i) analysis of the performance
of different markers as DNA barcodes, primarily below the
species level (i.e., Andean and Mesoamerican gene pools);
and (ii) evaluation of the effectiveness of different methods
(i.e., tree based versus character based) of DNA barcoding.
Materials and methods
Germplasm sampling of Phaseolus
In total, 33 varieties of P. vulgaris were selected as repre-
sentative of the Mesoamerican and Andean gene pools, based
on morphological seed traits, plant descriptors, and molecular
markers (Rossi et al. 2009). Eight wild and nine domesticated
accessions from Central America (Mexico, Costa Rica, Hon-
duras, and El Salvador) and ten wild and six domesticated
accessions from South America (Argentina, Bolivia, Brazil,
Colombia and Peru) were used, including two wild acces-
sions from northern Peru and Ecuador characterized by the
ancestral phaseolin type I (Debouck et al. 1993; Kami et al.
530 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
1995). These accessions were obtained from the germplasm
banks held at the International Center for Tropical Agricul-
ture (CIAT) and the United States Department of Agriculture
(USDA) (Table 1). In addition, 22 Italian, cultivated, com-
mercially available accessions from unknown progenitor
gene pools were obtained from the Agricultural Research
Council (CRA), Research Unit for Horticulture of Montanaso
Lombardo (Fig. 1). Several Phaseolus coccineus L., P. luna-
tus, and Vigna unguiculata (L.) Walp accessions were used
as reference standards and outgroups. A list of varieties and
landraces with information on their origins can be found in
Table 1.
Genomic DNA extraction
Genomic DNA was isolated from 0.5–1.0 g of powdered,
frozen, young leaf tissue using the Nucleon PhytoPure DNA
extraction kit (Amersham Biosciences, Little Chalfont, Buck-
inghamshire, UK), following the manufacturer’s instructions.
A purification step with NaOAc was performed to remove
excess salts, and the DNA pellets were resuspended in 80–
100 µL of 1× TE buffer (100 mmol/L Tris–HCl, 0.1 mmol/L
EDTA, pH 8). DNA concentration was estimated by electro-
phoresis on an 0.8% agarose/TAE gel using the 1 kb Plus
DNA ladder (Invitrogen, Carlsbad, California) as a size
standard.
DNA barcode markers and PCR assays
To employ a multilocus barcoding technique (Kress and
Erickson 2007; Newmaster et al. 2006), a subset of bean
samples was tested at several genomic regions to determine
the markers that provided the highest polymorphism informa-
tion content at the intraspecific level. Only 7 of 12 chloro-
plast gene regions, including both coding (rbcL and matK)
and noncoding regions (the atpB-rbcL,trnH-psbA,trnT-trnL,
and rpoB-trnC intergenic spacers and the trnL intron), proved
variable and informative, whereas the other regions (rpl32-
trnL,ndhF-rpl32,trnD-trnT,trnS-trnG, and rpoC1) were
found to be monomorphic and were not adopted for further
analysis (data not shown). ITS1 and ITS2, the two ITS that
separate the 5.8S ribosomal gene from the 18S and 25S loci
in rDNA, were used to compare the utility of the nuclear and
chloroplast genomes for resolving relationships at the variety
level. For three of the selected chloroplast DNA (cpDNA)
barcode regions, rbcL,trnL, and atpB-rbcL, primers were de-
signed based on the sequences in the National Center for Bi-
otechnology Information (NCBI) databases for the Fabaceae
(legume) family. After removal of redundant and unverified
entries, serial local multiple sequence alignments were per-
formed by the Vector NT software. We used the PRIMER3
software to design specific primer pairs, ranging from 18 to
28 base pairs (bp) and located in highly conserved short
stretches (300–500 bp) flanking the most variable portions
of each region. In the other cases, universal primers were
adopted (Table 2).
All PCR experiments were performed in duplicate using
the GeneAmp PCR System 9700 (Applied Biosystems, Fos-
ter City, California) with an initial denaturation step of 5 min
at 95 °C; followed by 35 cycles of 30 s at 95 °C, 1.10 min at
54 °C or 56 °C, and 1.20 min at 72 °C; followed by 7 min at
72 °C; and then held at 4 °C. PCR conditions were modified
for the matK marker: an initial denaturation step of 5 min at
95 °C; followed by 40 cycles of 30 s at 95 °C, 1 min at 56 °
C, and 2 min at 72 °C; followed by 7 min at 72 °C. The 25
µL PCR volume included 1× PCR buffer (100 mmol/L Tris–
HCl pH 9.0, 15 mmol/L MgCl2, and 500 mmol/L KCl),
0.2 mmol/L dNTPs, 0.2 µmol/L of each primer, 0.5 U of
Taq DNA polymerase, 15 ng of genomic DNA as template,
and 1× Hi Specific Additive (Bioline, London, UK) to facili-
tate amplification. The PCR products were resolved on 2%
agarose/TAE gels and visualized under UV light using ethi-
dium bromide staining. When faint double bands indicating
the presence of nonspecific products were visualized on a
gel, a second PCR was performed using more stringent con-
ditions (higher annealing temperatures and fewer cycle num-
bers). Positive and negative controls were used as references.
All amplification products were purified enzymatically by di-
gestion with exonuclease I and shrimp alkaline phosphatase
(Amersham Biosciences) and then sequenced using forward
and reverse primers according to the original Rhodamine ter-
minator cycle sequencing kit (ABI PRISM; Applied Biosys-
tems). For some regions, an additional forward or reverse
primer located outside the amplified region was adopted for
sequencing replicates. For sequencing matK, dimethyl sulfox-
ide at 4% of the reaction volume was used to overcome some
secondary structural problems.
Tree-based analysis
DNA sequences were visualized and manually edited using
Sequencer 4.8 software to minimize sequencing errors and
remove gaps in the coding regions that could cause shifts in
the open reading frames of rbcL.
The BLASTn algorithm (http://www.ncbi.nlm.nih.gov/
BLAST) was used to perform sequence similarity searches
against the nonredundant nucleotide databases of NCBI.
Then, the correspondence between the sequences of the PCR
amplicons and the known sequences was tested. We carried
out separate data analyses for each individual sequence and
for the combined chloroplast and nuclear data sets, individu-
ally and together. Multiple sequence alignments were per-
formed by the Se–Al v2.0a11 software, and the inter- and
intraspecific genetic divergences were calculated by the
MEGA 4.1 beta software (Tamura et al. 2007) according to
the Kimura 2-parameter distance model (Kimura 1980).
Based on the pairwise nucleotide sequence divergences, the
neighbor-joining (NJ) tree was estimated and rooted using
the accessions from different species as outgroups. A boot-
strap analysis was conducted to measure the stability of the
computed branches with 1000 resampling replicates. All nu-
cleotide positions containing gaps and missing bases were
eliminated from the data set (the complete deletion option).
To assign each accession to the correct gene pool, we used a
phenetic approach based on the computation of genetic dis-
tance to detect the “barcode gap”, a discontinuity between intra-
and interspecific variation (Hebert et al. 2003; Barrett and
Hebert 2005), and the derived “10× rule”in Phaseolus
spp. polymorphism analysis was performed on the complete
sequence, a combination of the cpDNA regions, and the nu-
clear ITS regions.
Character-based analysis
The character-based technique was employed to look for
unique sets of diagnostic characters related to single varieties
Nicolè et al. 531
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Table 1. List of 63 bean entries with the common name, accession number, origin area, and voucher information.
Sample Species Accessions Classification Origin Gene pool Voucher No.
PvF8wanc Phaseolus vulgaris G23585 Wild-ancestral South America (Peru) Ancestral i.p.
PvG8wanc Phaseolus vulgaris G23587 Wild-ancestral South America (Peru) Ancestral i.p.
PvH2mw Phaseolus vulgaris G23652 Wild Central America (Mexico) Mesoamerican i.p.
PvA3mw Phaseolus vulgaris G12979 Wild Central America (Mexico) Mesoamerican i.p.
PvC3mw Phaseolus vulgaris G23463 Wild Central America (Mexico) Mesoamerican i.p.
PvD3mw Phaseolus vulgaris G22837 Wild Central America (Mexico) Mesoamerican i.p.
PvB7mw Phaseolus vulgaris G12873 Wild Central America (Mexico) Mesoamerican 3901-8
PvG7mw Phaseolus vulgaris G12922 Wild Central America (Mexico) Mesoamerican i.p.
PvB8mw Phaseolus vulgaris G11050 Wild Central America (Mexico) Mesoamerican i.p.
PvC8mw Phaseolus vulgaris G12949 Wild Central America (Mexico) n.d. i.p.
PvD8aw Phaseolus vulgaris G21113 Wild South America (Colombia) Mesoamerican i.p.
PvE6aw Phaseolus vulgaris G23445 Wild South America (Bolivia) Andean i.p.
PvF6aw Phaseolus vulgaris G23444 Wild South America (Bolivia) Andean i.p.
PvG6aw Phaseolus vulgaris W618821 Wild South America (Bolivia) Andean i.p.
PvH6aw Phaseolus vulgaris G23455 Wild South America (Peru) Andean i.p.
PvG3aw Phaseolus vulgaris G23420 Wild South America (Peru) Andean i.p.
PvB6aw Phaseolus vulgaris G19893 Wild South America (Argentina) Andean i.p.
PvC6aw Phaseolus vulgaris G19898 Wild South America (Argentina) Andean i.p.
PvD6aw Phaseolus vulgaris G21198 Wild South America (Argentina) Andean i.p.
PvH5aw Phaseolus vulgaris W617499 Wild South America (Argentina) n.d. i.p.
PvF7md Phaseolus vulgaris PI201349 Domesticated Central America (Mexico) Mesoamerican i.p.
PvG1md Phaseolus vulgaris PI165435 Domesticated Central America (Mexico) Mesoamerican 3901-10
PvH1md Phaseolus vulgaris PI165440 Domesticated Central America (Mexico) Mesoamerican i.p.
PvA2md Phaseolus vulgaris PI309785 Domesticated Central America (Mexico) Mesoamerican i.p.
PvH4md Phaseolus vulgaris PI207370 Domesticated Central America (Mexico) Andean i.p.
PvE7md Phaseolus vulgaris PI309885 Domesticated Central America (Costa Rica) Mesoamerican i.p.
PvD1md Phaseolus vulgaris PI309831 Domesticated Central America (Costa Rica) Mesoamerican i.p.
PvF1md Phaseolus vulgaris PI310577 Domesticated Central America (Honduras) Mesoamerican i.p.
PvE1md Phaseolus vulgaris PI304110 Domesticated Central America (El Salvador) n.d. i.p.
PvC1ad Phaseolus vulgaris BAT93–1 Domesticated South America (Colombia) Mesoamerican i.p.
PvC2ad Phaseolus vulgaris BAT93–2 Domesticated South America (Colombia) Mesoamerican i.p.
PvH8ad Phaseolus vulgaris BAT881 Domesticated South America (Colombia) n.d. 3901-11
PvB4ad Phaseolus vulgaris MIDAS Domesticated South America (Argentina) Andean i.p.
PvD5ad Phaseolus vulgaris PI290992 Domesticated South America (Peru) Andean 3901-9
PvA7ad Phaseolus vulgaris JALOEEP558 Domesticated South America (Brasile) Andean 3901-7
Pv1itc Phaseolus vulgaris Cannellino rosso Cultivated Italy —3901-16
Pv3itc Phaseolus vulgaris Montalbano Cultivated Italy —3901-18
Pv6itc Phaseolus vulgaris Munachedda nera Cultivated Italy —3901-19
Pv9itc Phaseolus vulgaris San Michele Cultivated Italy —i.p.
Pv10itc Phaseolus vulgaris Nasieddu viola Cultivated Italy —i.p.
Pv13itc Phaseolus vulgaris Maruchedda Cultivated Italy —i.p.
Pv14itc Phaseolus vulgaris Riso bianco Cultivated Italy —3901-20
Pv16itc Phaseolus vulgaris Cannellino Cultivated Italy —3901-21
Pv19itc Phaseolus vulgaris Verdolino Cultivated Italy —3901-22
Pv22itc Phaseolus vulgaris Blu Lake Cultivated Italy —3901-23
Pv23itc Phaseolus vulgaris Goldrush Cultivated Italy —3901-24
Pv24itc Phaseolus vulgaris Borlotto Clio Cultivated Italy —i.p.
Pv27itc Phaseolus vulgaris Lena Cultivated Italy —3901-25
532 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
or variety groups of P. vulgaris. Rather than using hierarchies
or distance trees, character-based analysis classifies taxo-
nomic groups based on shared specific informative character
states, SNPs or insertions or deletions (indels), at either one
or multiple nucleotide positions (DeSalle et al. 2005). Analy-
sis of polymorphism distribution was performed using the
DnaSP v.4 software (Rozas et al. 2003) to generate a map
containing haplotype data without considering sites with
alignment gaps. This program detects positions characterized
by the presence of specific character states that are limited to
a particular subgroup within P. vulgaris species and shared
by all the members of that cluster. In addition, the haplotype
number, Hn, and the haplotype diversity, Hd(Nei 1987), were
estimated.
Population structure analysis
The population structure of the P. vulgaris germplasm was
investigated using the Bayesian model-based clustering algo-
rithm implemented in the STRUCTURE software (Pritchard
et al. 2000; Falush et al. 2003), which identifies subgroups
according to combination and distribution of molecular
markers. This software was also used to assign each DNA
sample of varieties and landraces, predefined according to
geographical origin and (or) gene pool, to an inferred cluster.
All simulations were executed assuming the admixture
model, with no a priori population information. Analyses of
SNP data were performed with 500000 iterations and 500
000 burn-ins by assuming the allele frequencies among pop-
ulations to be correlated (Falush et al. 2003). Ten replicate
runs were performed, with each run exploring a range of K
spanning from 1 to 16. The most likely value of Kwas esti-
mated using DK, as reported in other studies (Evanno et al.
2005). Individuals with membership coefficients of qi≥0.7
were assigned to a specific group, whereas individuals with
qi< 0.7 were identified as admixed.
Results
DNA barcoding success and levels of variability
For the selected chloroplast and nuclear markers examined
in all 63 accessions of Phaseolus spp., our PCR amplifica-
tions were successful 100% of the time, although low quality
sequences were sometimes produced because of specific gene
regions (Table 3). For all dubious amplicons and sequences,
the reactions were repeated. The only particularly problem-
atic barcode marker was matK, with multiple failed amplifi-
cations and low sequence quality. Similar difficulties have
been reported by others (Kress and Erickson 2007; Fazekas
et al. 2008). Therefore, we removed this region from our
analyses.
The primer pairs designed for trnT-trnL and trnH-psbA
proved highly universal with a 100% success rate for both
PCR and sequencing, whereas primers for the other markers
(i.e., rbcL,atpB-rbcL,trnL, and rpoB-trnC) were also highly
universal but unreliable in sequence quality. Although double
PCR products were usually not detectable in the gel, se-
quencing problems likely arose from multiple comigrating
amplicons of similar size but different sequence. When non-
specific amplicons of unexpected length were visible in the
gel (i.e., for rbcL and atpB-rbcL), a second, more stringent
PCR was performed, or new primer pairs were adopted for
Table 1 (concluded).
Sample Species Accessions Classification Origin Gene pool Voucher No.
Pv28itc Phaseolus vulgaris Giulia Cultivated Italy —3901-26
Pv29itc Phaseolus vulgaris Saluggia Cultivated Italy —3901-27
Pv31itc Phaseolus vulgaris Borlotto Lamon Cultivated Italy —3901-28
Pv32itc Phaseolus vulgaris Saluggia Cultivated Italy —3901-29
Pv33itc Phaseolus vulgaris Cannellini Cultivated Italy —3901-30
Pv34itc Phaseolus vulgaris Verdoni Cultivated Italy —3901-34
Pv35itc Phaseolus vulgaris S. Matteo Cultivated Italy —3901-31
Pv36itc Phaseolus vulgaris Zolferini Rovigotti Cultivated Italy —3901-32
Pv37itc Phaseolus vulgaris Neri Messicani Cultivated Italy —3901-33
PcA1mw Phaseolus coccineus PI417608 Wild Central America (Mexico) n.d. i.p.
Pc30itc Phaseolus coccineus Venere Cultivated Italy —i.p.
Pc39itc Phaseolus coccineus Spagna Cultivated Italy —i.p.
PlB1md Phaseolus lunatus PI310620 Domesticated Central America (Guatemala) n.d. i.p.
Pl38itc Phaseolus lunatus Lima Cultivated Italy —3901-2
Vu40itc Vigna unguiculata Fagiolino dall'occhio Cultivated Italy —3905-2
Note: Voucher No., plants with flowers and pods are conserved in the herbarium of the Botanical Garden of the University of Padua (Italy); i.p., voucher attainment in progress; n.d., not determined.
Nicolè et al. 533
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Fig. 1. Seeds of the common bean (Phaeolus vulgaris L.) varieties analyzed in this study as representatives of the Italian cultivated germ-
plasm (1, Cannellino rosso; 2, Riso giallo; 3, Montalbano; 4, Munachedda nera; 5, San Michele; 6, Nasieddu Viola; 7, Maruchedda; 8, Riso
bianco; 9, Cannellino nano; 10, Verdolino; 11, Blu lake; 12, Goldrush; 13, Clio; 14, Zolferino rovigotto; 15, Lena; 16, Giulia; 17, Saluggia
nano; 18, Venere; 19, Borlotto Lamon; 20, Saluggia; 21, Cannellino; 22, Verdone; 23, San Matteo; 24, Nero messicano; 25, BAT881 (refer-
ence breeding line)). Also analyzed in this study seeds of Phaseolus lunatus L. (26, sieva bean from Lima), Phaseolus coccineus L. (27,
scarlet runner bean or Spanish bean), and Vigna unguiculata L. Walp. (28, blackeyed pea).
534 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Table 2. List of primers used for each chloroplast and nuclear marker with their nucleotide sequence, amplicon length, and reference source.
Amplicon length (bp)
Marker Phaseolus
vulgaris
Phaseolus
coccineus
Phaseolus
lunatus
Vigna
uguiculata Primer name Primer sequence (5′-3′)Ta(°C) References
rbcL gene 543 543 543 543 rbcL_F GCAGCATTYCGAGTAASTCCYCA 56 Nicolé et al. unpublished
rbcL_R GAAACGYTCTCTCCAWCGCATAAA Nicolé et al. unpublished
rbcL 724R* TCACATGTACCTGCAGTAGC Lledó et al. 1998
matK gene 695 695 695 695 matK4La CCTTCGATACTGGGTGAAAGAT 56 Wojciechowski et al. 2004
matK1932Ra CCAGACCGGCTTACTAATGGG Wojciechowski et al. 2004
trnL intron 350 350 296 357 trnL_F GGATAGGTGCAGAGACTCRATGGAAG 56 Nicolé et al. unpublished
trnL_R TGACATGTAGAATGGGACTCTATCTTTAT Nicolé et al. unpublished
5′trnLUAAF* CGAAATCGGTAGACGCTACG Taberlet et al. 1991
3′trnLUAAR* GGGGATAGAGGGACTTGAAC Taberlet et al. 1991
atpB-rbcL IGS 329 325 326 331 atpB_F GGTACTATTCAATCAATCCTCTTTAATTGT 56 Nicolé et al. unpublished
atpB_R ATGTAAATCCTAGATGTRAAAATAKGCAG Nicolé et al. unpublished
atpB_R2* CGCAACCCAATCTTTGTTTC Nicolé et al. unpublished
trnH-psbA IGS 365 365 365 369 psbA3′f GTTATGCATGAACGTAATGCTC 56 Sang et al. 1997
trnHf CGCATGGTGGATTCACAATCC Tate and Simpson 2003
rpoB-trnC IGS 1117 1117 1124 1136 rpoB_F CKACAAAAYCCYTCRAATTG 54 Shaw and Small 2005
trnCGCAR CACCCRGATTYGAACTGGGG Shaw and Small 2005
rpoB_R3* TTCTTTACAATCCCGAATGG Nicolé et al. unpublished
trnT-trnL IGS 813 837 823 871 trnTUGU2F CAAATGCGATGCTCTAACCT 56 Cronn et al. 2002
5′trnLUAAR TCTACCGATTTCGCCATATC Taberlet et al. 1991
Total length 3556 3576 3509 3627
ITS1 373 382 355–364 314 ITS5 GGAAGTAAAAGTCGTAACAAGG 54 White et al. 1990
ITS2 GCTGCGTTCTTCATCGATGC White et al. 1990
ITS2 419 418 413 401 ITS3 GCATCGATGAAGAACGCAGC 54 White et al. 1990
ITS4 TCCTCCGCTTATTGATATGC White et al. 1990
*Primers used only for sequencing.
Nicolè et al. 535
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
sequencing (see Table 2). Similar problems were experienced
and solved for the ITS1 and ITS2 markers (Table 3).
The sequences of accessions corresponding to different va-
rieties differed only at SNPs and were, therefore, easily
aligned, but the sequences corresponding to different species
or genera contained indels in some portions of the noncoding
cpDNA, requiring manual editing of the alignments. For the
ITS regions, heterozygosity was detected at only a few nu-
cleotide positions (see Table 3), and the sites of nucleotide
substitutions were recorded using the conventional code for
degenerate bases of the International Union of Biochemistry.
The single sequences analyzed for cpDNA markers ranged
from 328 to 1124 bp, covering a total length of 4229 bp,
whereas amplicons for ITS1 and ITS2 markers averaged 358
and 413 bp, respectively. The occurrence of polymorphisms
among P. vulgaris accessions was limited to single nucleoti-
des; 17 SNPs were documented across the six chloroplast
markers, and 10 SNPs were found for the two nuclear
markers (Table 3).
The tree-based genetic identification method
The distance matrices based on the K2P substitution model
for both chloroplast and nuclear regions were generated, and
the average values were calculated between Phaseolus spp.
and between subpopulations of P. vulgaris. Combined DNA
barcode sequences showed high interspecific and low intra-
specific variation rates (Table 4). The genetic distances be-
tween P. vulgaris and V. unguiculata, calculated over all
barcode regions, were 0.0618 and 0.1651 on the basis of
cpDNA and ITS polymorphisms, respectively. Moreover,
P. vulgaris proved to be more closely related to P. cocci-
neus than to P. lunatus, according to both chloroplast and
nuclear markers. The average genetic distance of the former
was 0.0104 and 0.0173, whereas with the latter it was
0.0231 and 0.0432 on the basis of cpDNA and ITS sequen-
ces, respectively (see Supplementary data,1Table S1). In P.
vulgaris, the genetic distance estimated within varietal
groups, classified on the basis of the known gene pool
membership, was 0.0011 and 0 for the Andean gene pool
according to cpDNA and ITS markers, respectively; for the
Mesoamerican gene pool it was 0.0021 for cpDNA and
0.0020 for ITS regions (Fig. 2).
Because our focus was on the detection of polymorphisms
useful for discriminating among P. vulgaris landraces and va-
rieties within Mesoamerican, Andean, and Italian plant mate-
rials, further analysis was based on the DNA markers scored
as polymorphic at the intraspecific level. The degree of nu-
cleotide differentiation between congeneric species was at
least 5-fold higher than were values estimated within species,
whereas no significant sequence divergence rate was scored
between the two different gene pools of P. vulgaris. Further-
more, out of 1600 intraspecific comparisons of the chloro-
plast and nuclear markers, 180 (11.25%) showed no
significant differences between varieties.
We used the NJ tree method to analyze genetic distinctive-
ness using cpDNA markers. The NJ tree allows the conver-
sion of sequence polymorphisms into genetic distances using
nucleotide substitution models (Wiemers and Fiedler 2007).
Based on the coalescence of conspecific populations with in-
complete sampling, the NJ tree assembles all the accessions
derived from one species into a single group. Separate analy-
ses for each marker yielded NJ trees that correctly distin-
guished sister species and different genera, forming separate
clusters for V. unguiculata,P. lunatus,P. coccineus, and
P. vulgaris (data not shown). In contrast, the NJ tree built
for each barcode sequence of P. vulgaris species was not
unique because of tie trees retrieved due to low divergence
values among common bean accessions. Moreover, the NJ
tree constructed from the whole set of cpDNA polymor-
phisms produced low discrimination among accessions
within the species P. vulgaris, owing to the complete lack
or paucity of informative characters in the investigated
chloroplast regions.
In the NJ tree constructed with a combination of sequence
polymorphisms of the four variable chloroplast markers,
members of the species P. vulgaris,P. coccineus, and P. lu-
natus were split into defined clusters, with bootstrap values
as high as 99%–100%, whereas the branching nodes of
P. vulgaris subgroups were weakly supported, with boot-
strap values ≤60% in most cases (see Supplementary data,
Figure S1). The accessions of P. vulgaris derived from ei-
ther Mesoamerican or Andean gene pools grouped together
and formed a few subclusters slightly separated from each
other, with several exceptions. In four cases the gene pool
Table 3. Basic information on the cpDNA and internal transcribed spacers (ITS) barcode regions, including sequence length of amplicons,
inter- and intraspecific number and frequency of SNPs, and insertions or deletions (indels).
rbcL matK trnL atpB-rbcL trnH-psbA trnT-trnL rpoB-trnC ITS1 ITS2
Total No. of Phaseolus entries 63 63 63 63 63 63 63 63 63
Average amplicon length (bp) 543 695 338 328 366 836 1124 358 413
No. of SNPs in Phaseolus spp. 8 n.d. 21 14 14 53 48 65 58
Interspecific frequency (SNPs/100 bp) 1.5 n.d. 6.0 4.3 3.8 6.5 4.2 17.4 13.8
No. of SNPs in P. vulgaris 0 n.d. 4 0 8 3 2 6 4
Intraspecific frequency (SNPs/100 bp) 0 n.d. 1.1 0 2.2 0.4 0.2 1.6 1.0
No. of indels in Phaseolus spp. 0 n.d. 1 4 0 5 5 10 5
Average indel size (bp) 0 n.d. 58 2 0 7 2 4 5
No. of heterozygous sites n.a. n.a. n.a. n.a. n.a. n.a. n.a. 3 7
Amplification success (%) 100 100 100 100 100 100 100 100 100
Sequencing success (%) 100 62 100 100 100 100 90 97 100
Note: n.d., not determined; n.a., not applicable. The percentage of sequence-tagged site PCR and sequencing success is also reported.
1Supplementary data are available with the article at www.nrcresearchpress.com/gen.
536 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Table 4. Consensus sequence related to the 17 individual SNPs detected in the target cpDNA regions with information on the haplotypes found across all common bean (Phaseolus
vulgaris L.) entries.
Halotype (no. of entries)
Ancestral Mesoamerican Andean
Marker SNP
position Consensus
sequence Hap16
(2) Hap09
(1) Hap01
(1) Hap03
(10) Hap08
(1) Hap12
(1) Hap13
(3) Hap06
(7) Hap14
(1) Hap15
(3) Hap02
(15) Hap04
(3) Hap10
(1) Hap11
(1) Hap07
(1) Hap05
(6)
trnL 14 G A AAAA
183 A C C
264 T G G G G
332 T A A A A A
trnH-psbA 156 A C C C
219 T C C
223 A T T
224 A T T
225 A T T
229 G A A
272 T G G G G
283 C A
trnT-trnL 85 A CCC
512 A G
673 T G G G
rpoB-trnC 478 G TTT
642 A n.d. C C C C n.d.
Note: Haplotypes are arranged in three main subgroups for ancestrals, Mesoamerican, and Andean gene pools. n.d., not determined. Hap01: PvA2md; Hap02: PvA7ad, PvG6aw, PvG3aw, PvB4ad, Pv1itc,
Pv6itc, Pv9itc, Pv10itc, Pv13itc, Pv14itc, Pv16itc, Pv19itc, Pv24itc, Pv27itc, Pv32itc; Hap03: PvC3mw, PvG1md, PvC1ad, PvH1md, PvC2ad, PvE7md, PvH8ad, PvF1md, Pv22itc, Pv23itc; Hap04: PvH5aw,
PvD6aw, Pv3itc; Hap05: PvH2mw, PvA3mw, PvB7mw, PvE6aw, PvF6aw, PvD1md; Hap06: PvH4md, Pv28itc, Pv29itc, Pv31itc, Pv33itc, Pv34itc, Pv36itc; Hap07: PvH6aw; Hap08: PvD3mw; Hap09:
PvD5ad; Hap10: PvB6aw; Hap11: PvC6aw; Hap12: PvE1md; Hap13: PvF7md, Pv35itc, Pv37itc; Hap14: PvG7mw; Hap15: PvB8mw, PvC8mw, PvD8aw; Hap16: PvF8wanc, PvG8wanc.
Nicolè et al. 537
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
was in disagreement with the geographic origin. In two of
these four cases, i.e., PvH4md (from Mexico but belonging
to the Andean gene pool, based on Rossi et al. (2009)) and
PvD8aw (from Colombia but belonging to the Mesoameri-
can gene pool after Rossi et al. (2009)), the positions of
the two accessions in the NJ tree were not in conflict with
those of the other genotypes. In fact, PvH4md grouped with
Italian cultivars and PvD8aw clustered with two Mesoamer-
ican accessions. In four different cases, there was no indica-
tion of a gene pool, but it was possible to recover this
information using NJ analysis. Two of these cases were
wild accessions (PvC8mw and PvH5aw), and for these gen-
otypes, the gene pool matched the geographic origin, as ex-
pected; the other two were domesticated accessions
(PvE1md and PvH8ad), and their position in the tree sug-
gests that they may have been transferred between regions,
possibly by human intervention (see Supplementary data,
Fig. S1). If all common bean accessions are classified ac-
cording to their position in the NJ tree, then it is evident
that 26 accessions belong to the Andean gene pool and
that the remaining 29 belong to the Mesoamerican gene
pool (see Table 1). It is worth noting that the ancestral
bean accessions were recognized as a separate subcluster
with a high confidence value and that they were grouped
with another accession from Peru (see Supplementary mate-
rials, Fig. S1), the putative primary center of the ancestral
wild gene pool (Debouck et al. 1993).
The NJ tree constructed using SNPs from the nuclear ITS
regions, based on a lower number of polymorphisms among
varieties compared with cpDNA regions, revealed an unstruc-
tured distribution of the SNPs with no subgroups for P. vul-
garis accessions (data not shown).
The character-based genetic characterization method
Owing to the paucity of results from the above genetic dis-
tance method, a second, character-based approach was em-
ployed to identify diagnostic attributes shared between the
members of a given taxonomic group but absent from a dif-
ferent clade that descends from the same node (Rach et al.
2008). This method does not consider indels (which were
not found at the intraspecific level anyway); hence, the infor-
mative characters employed in the character-based approach
were limited to SNPs.
Within P. vulgaris, the occurrence of SNPs depended on
the marker used: for rbcL and atpB-rbcL sequences, no
SNPs were detected, whereas for the other regions the num-
ber varied from two to eight (the latter for trnH-psbA).
Among the cpDNA markers, trnH-psbA and trnL showed
the highest number of SNPs, proving to be the most suitable
regions for discrimination of genotypes within a species,
along with the nuclear ITS1 and ITS2 markers. Of the other
four chloroplast regions, only trnT-trnL and rpoB-trnC exhib-
ited SNP markers among accessions, although at a lower fre-
quency (see Table 3). SNP analysis of the entire chloroplast
data set revealed 16 haplotypes out of the 57 accessions of
P. vulgaris (Table 4). It is worth noting that four of these
were the most common haplotypes, each being shared by 6–
15 accessions. Unique haplotypes were found for 8 of the 57
common bean accessions (Table 4); the number of haplo-
types (Hn) was nine for Central American, nine for South
American, and five for Italian varieties. The haplotype diver-
Fig. 2. Histograms representing the inter- and intraspecific diver-
gences calculated using chloroplast (A) and nuclear (B) markers. In
addition to the mean value, the standard deviation is reported for
each comparison within and between species.
538 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
sity (Hd) was 0.875, 0.908, and 0.688, respectively, for the
three regions (Table 5), with a mean Hdof 0.877 for P. vul-
garis.
The haplotypes based on chloroplast polymorphisms and
corresponding to varietal subgroups within P. vulgaris spe-
cies were used for the construction of a NJ tree (Fig. 3). The
majority of haplotypes nested together in tightly clustered
subgroups supported by low bootstrap values, with the excep-
tion of several haplotypes shared by the northern Peru and
Ecuador accessions characterized by the phaseolin type I
(e.g., haplotype number 16) and wild accessions. The latter
finding is particularly evident for some correlated haplo-
types such as Nos. 4, 10, and 11 that are linked to the An-
dean gene pool, as well as 6, 14, and 15 that are associated
with the Mesoamerican gene pool (see Fig. 3 and Table 5).
Accessions belonging to P. coccineus,P. lunatus, and V.
unguiculata revealed unique haplotypes that were grouped
separately for each species.
The number of segregating sites for chloroplast regions
was 9 out of 29 Mesoamerican accessions and 13 out of 26
Andean accessions. There were eight haplotypes (Hn)for
Mesoamerican accessions and nine for Andean accessions,
and the estimate of haplotype diversity (Hd) proved slightly
higher for the Mesoamerican (0.823) than the Andean gene
pool (0.665). Even without taking the 22 modern Italian vari-
eties into account, the haplotype diversity remained compara-
ble between true Mesoamerican and Andean common bean
accessions, with Hdvalues of 0.875 and 0.908, respectively
(Table 5).
The ITS data set for P. vulgaris was not informative; all
accessions, except the phaseolin type I entries that formed
two separate haplotypes, were grouped together in three hap-
lotypes, with one including 52 out of the 57 accessions (data
not shown). The Italian accessions did not show any poly-
morphic sites, whereas the South American accessions were
the most variable and scored a haplotype diversity much
higher than the Central American ones. The haplotype diver-
sity of the Mesoamerican gene pool was 0.204, but no haplo-
type diversity was found for the Andean gene pool (see
Table 5).
Investigation into the population structure of the P. vu l-
garis germplasm by estimation of DK(Evanno et al.
2005) suggested that our core collection of accessions is
most likely made up of three genetically distinguishable
subgroups (K=3),asshowninFig.4.Inparticular,23
of the 26 Andean accessions grouped separately from most
of the Mesoamerican accessions, showing a high genetic
homogeneity within this gene pool and a high estimated
membership for each individual. Of the 29 Mesoamerican
accessions, 24 were divided into two clearly distinguishable
subgroups of 14 and 10 individuals each, whereas the re-
maining 5 were clustered into a subgroup closely resem-
bling that of the Andean accessions (Fig. 4). On the
whole, this analysis showed that genetic diversity is low
among accessions of the Andean gene pool and that acces-
sions of the Mesoamerican gene pool are grouped into
three genetically differentiated clusters. Accessions with an
admixed ancestry were not detected as expected in absence
of recombination. It is notable that the two ancestral acces-
sions proved to be closely related to one of the Mesoamer-
ican clusters.
Table 5. Summary of genetic diversity computed separately for chloroplast (A) and nuclear (B) DNA markers for subgroups of geographically distinct accessions and over all accessions
of Phaseolus vulgaris L. and Phaseolus spp. (A, B) and for two different gene pools.
A
Germplasm source Geographical origin Gene pool
Genetic diversity statistics Phaseolus spp. Phaseolus vulgaris Central America South America Italy MesoamericanaAndeanb
No. of segregating sites (S) 122 17 9 14 7 9 13
Haplotype number (Hn)21 16 9 9 58 9
Haplotype diversity (Hd) 0.898 0.877 0.875 0.908 0.688 0.823 0.665
B
Germplasm source Geographical origin Gene pool
Genetic diversity statistics Phaseolus spp. Phaseolus vulgaris Central America South America Italy MesoamericancAndeanb
No. of segregating sites (S)69 9 5 7 0 6 0
Haplotype number (Hn)9 5 2 4 1 3 1
Haplotype diversity (Hd) 0.323 0.171 0.122 0.371 0 0.204 0
a29 accessions.
b26 accessions.
c28 accessions.
Nicolè et al. 539
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Discussion
Our results in Phaseolus spp. further support DNA barcod-
ing as a powerful technique for taxonomic identification and
phylogenetic analyses aimed at reconstructing evolutionary
patterns and genetic distances between tightly related species.
In addition to SNPs, several indels were discovered among
Phaseolus spp. Most of the interspecific phylogenetic rela-
tionships previously identified by Delgado-Salinas et al.
(1999) were confirmed by our data, with P. vulgaris more
closely related to P. coccineus than to P. lunatus.
Because the main goal of this study was to identify those
markers with the greatest polymorphism information and the
best performance in intraspecific barcoding, we focused on
the relevance of the nucleotide variation among accessions
of P. vulgaris. Considering the recent criticisms formulated
by the CBOL Plant Working Group of the effectiveness of
single barcodes and assuming that shallow nucleotide poly-
morphisms would have previously been detected within spe-
cies, a multilocus approach was adopted. To investigate the
genetic distinctiveness of pure lines, varietal groups, and
gene pools for the common bean, we used the following cri-
teria to select the DNA regions suitable for barcoding: (i)a
Fig. 3. Neighbor-joining tree based on the 16 haplotypes identified from the 57 bean accessions of Phaseolus vulgaris L. (for details on
haplotypes, see also Table 5).
540 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
high number of sequences available in public gene banks to
facilitate both primer design and the identification of species
by querying nucleotide databases; and (ii) an appropriate sub-
stitution rate for intraspecific studies on the basis of informa-
tion available in the literature.
To evaluate whether DNA barcoding is an efficient tool for
the analysis of intraspecific variation and for the identifica-
tion of landraces and cultivars within a species, two strat-
egies were tested: (i) a phenetic tree-building approach using
genetic distance data and the derived NJ tree to establish re-
lationships among accessions of P. vulgaris and Phaseolus
spp. and to determine the gene pool of origin for a set of Ital-
ian landraces; and (ii) a character-based system capable of re-
constructing haplotypes on the basis of diagnostic characters,
both fixed and variable among accessions and gene pools, for
the genetic identification of varietal groups without reference
to trees.
The standard tree-building approach proposed by Hebert et
al. (2003) to discriminate among closely related species en-
tails the use of sequence divergence values and the criterion
of reciprocal monophyly based on the NJ tree. The employ-
ment of the distance threshold derived from the barcode gap
as a tool for species delimitation is fundamental to DNA bar-
coding. This concept is controversial because a 10-fold
screening threshold of sequence difference is present in some
animals, such as birds and insects (Hebert et al. 2004; Haji-
babaei et al. 2006), but is absent in others, such as cowries
(Meyer and Paulay 2005). The latter observation supports
the hypothesis that the barcoding gap may be an artifact of
incorrect sampling (Meyer and Paulay 2005; Wiemers and
Fiedler 2007). An additional tool is the NJ tree profile that
allows the assignment of sequences to the correct species
based on the positions of the branches relative to the cluster
of the species (Wiemers and Fiedler 2007). In our study, this
type of system proved to be a powerful technique to correctly
cluster same-species accessions by the use of a standardized
genic or intergenic region as a molecular tag. All of the se-
quences, whether analyzed separately or together, supported
the distinctiveness of different species. In fact, even if we in-
vestigated a small number of genotypes of Phaseolus spp.,
the high nucleotide variability for these accessions, based on
the occurrence of both SNPs and indels, clearly indicated the
genetic distinctiveness of P. coccineus and P. lunatus from
P. vulgaris. In contrast, the NJ tree proved poorly informa-
tive for the genetic traceability of cultivars within P. vulga-
ris species. With the exceptions of the intergenic trnH-psbA
region and the trnL genic intron, the chloroplast sequences
contributed little or nothing toward resolving the genetic
identities of landraces and varieties. Although some con-
cerns have arisen about the difficulties associated to the
use of the trnH-psbA spacer (Whitlock et al. 2010), in the
present study we have never experienced problems with
this marker and, on the contrary, it proved to be the most
informative one, followed by the trnL. The NJ tree derived
from the chloroplast combined data set appeared to exhibit
a geographically related branching pattern, with the vast
majority of the Andean and Mesoamerican common bean
samples clustering separately. In this work, DNA barcoding
Fig. 4. Population structure of Phaseolus vulgaris L. germplasm core collection as estimated with STRUCTURE software. Each accession is
represented by a vertical histogram portioned into K= 3 colored segments that represent the estimated membership of each individual. Ac-
cessions were ordered by gene pool (i.e., Mesoamerican and Andean); improperly clustered accessions are indicated with an asterisk.
Nicolè et al. 541
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
failed to provide a clear separation between the Andean and
Mesoamerican gene pools, whereas several recent studies
successfully distinguished between the two groups by using
both chloroplast and nuclear SSR markers or genomic
AFLP markers alone (Kwak and Gepts 2009; Angioi et al.
2009; Rossi et al. 2009; Burle et al. 2010). Moreover, 12 of
the 22 Italian varieties clustered with the Andean gene pool,
whereas 10 accessions were classified as Mesoamerican.
This result confirms previous observations about the origin
and structure of European (Papa et al. 2006; Logozzo et al.
2007; Angioi et al. 2010) and Italian germplasm of P. vul-
garis (Sicard et al. 2005; Angioi et al. 2009).
Unlike the NJ tree based on cpDNA, the distance tree gen-
erated by combining the sequences of the nuclear markers
did not provide greater resolution. However, it confirmed
previous studies that discourage the use of ITS for intraspe-
cific phylogeny because of extensive intragenomic sequence
variation (Álvarez and Wendel 2003). The SNPs found in
ITS regions scored an average intraspecific frequency higher
than that of cpDNA regions (1.3 versus 0.65 SNPs/100 bp,
respectively). Nevertheless, the random distribution of ITS-
related SNPs negatively affected the genetic discrimination
between accessions and supports the likelihood of hybridiza-
tion among accessions, which may favor the occurrence of
intragenomic variation. In our study, intragenomic variation
is the strongest hypothesis because the inbreeding system of
P. vulgaris excludes a high frequency of heterozygous geno-
types.
The standard tree-building approach to discriminate be-
tween gene pools and the DNA barcoding method to identify
P. vulgaris varieties were not informative because of a slow
substitution rate. For this reason, a character-based system
was tested. For the DNA barcoding of multiple individuals
within a species, where the genetic distances are low, it has
been proposed that the character-based barcode is a more ap-
propriate approach than the phenetic system (Rach et al.
2008). The barcode method uses DNA sequence information
to generate discrete diagnostics for species identification.
To further explore intraspecific variability, the DnaSP soft-
ware was used to discover combinations of character states
both exclusive to a single variety and polymorphic among
varieties. For the 57 P. vulgaris accessions (landraces and va-
rieties), this approach allowed the detection of as few as 16
haplotypes over all cpDNA regions. These haplotypes corre-
sponded to an equal number of subgroups, each made up of
Mesoamerican or Andean accessions along with Italian ac-
cessions that clustered with either gene pool. The only excep-
tion was haplotype number 5, which was shared by mostly
wild accessions from both the Mesoamerican and Andean
groups. This finding raises concerns about the utility of
DNA barcoding for intraspecific genetic diversity analysis,
even when this technique is based on multiple loci. Although
it is true that a number of SNPs and haplotypes were recov-
ered for phaseolin type I, Mesoamerican, and Andean acces-
sion groups, it is also true that neither haplotypes nor
characters specific for single accessions were found (see Ta-
ble 4 for details).
In contrast to cpDNA regions, the nuclear ITS data set of
P. vulgaris proved, as expected, poorly informative; almost
all accessions clustered into a single group, except for the an-
cestral entries, which clustered apart. The corresponding NJ
tree revealed an unstructured distribution of SNPs with nei-
ther subgroups for P. vulgaris accessions (data not shown)
nor any segregating site among the Italian accessions. Con-
sistent discordances among molecular data sets (i.e., chloro-
plast versus nuclear markers) have been observed in other
taxa as well, e.g., in the Triticeae of the grasses (Mason-
Gamer and Kellogg 1996) and in the Anacardiaceae (Ting-
shuang et al. 2004).
The estimate of haplotype diversity deserves particular at-
tention because data based on cpDNA markers did not con-
flict with those based on nuclear ITS markers. When
cpDNA barcodes were used, accessions belonging to the
Mesoamerican gene pool exhibited a haplotype diversity
higher than that estimated for the Andean gene pool (Hd=
0.823 and 0.665, respectively). Conversely, when ITS
markers were used, no haplotype diversity was found for the
Andean gene pool, but for the Mesoamerican gene pool,
Hd= 0.204. Other works have demonstrated that the ge-
netic diversity within the two gene pools is, in general,
higher for the Mesoamerican gene pool compared with the
Andean one (see, e.g., Chacón et al. 2005; Kwak and Gepts
2009; Rossi et al. 2009). This finding was further supported
by independent cluster analyses with the STRUCTURE
software: genetic diversity was low among accessions of
the Andean gene pool that were grouped in tightly related
subclusters, whereas the accessions of the Mesoamerican
gene pool were grouped into three genetically differentiated
subclusters. In all cases, estimated membership values were
high, and admixed individuals were not present.
The 33 wild and domesticated common bean accessions
can be considered a core collection of Mesoamerican and
Andean gene pools, and the 22 commercial varieties are rep-
resentative of Italian cultivated germplasm. Both wild and
domesticated accessions within Mesoamerican and Andean
gene pools proved to be formed by pure lines that are poorly
distinguishable genetically from each other on the basis of
the cpDNA haplotypes and ITS polymorphisms.
To characterize the genetic diversity among common
beans, different approaches have been employed, from the
analysis of morphology and the seed protein phaseolin to the
examination of several types of molecular markers (for a re-
view see Papa et al. 2006). These methodologies have re-
vealed the existence of at least two major gene pools, the
Mesoamerican and the Andean, and several racial groups for
P. vulgaris (reviewed by Chacón et al. 2005; see also Rossi
et al. 2009). In our study, a new molecular tool, DNA bar-
coding combined with NJ tree-building, was tested to deter-
mine the genetic divergence of the modern common bean
cultivars and to relate them to wild and domesticated materi-
als from the original bean domestication centers. This techni-
que was shown to be highly reliable for identification
purposes at the species level but much less informative at
the variety level. Although DNA barcoding, using SNPs and
indels of genic or intergenic tagged regions, provided an ac-
curate method for the genetic identification of Phaseolus
spp., it should not be adopted for the genetic identification
of varieties within P. vulgaris.
The incorporation of multiple nuclear regions may be nec-
essary to reliably identify single common bean varieties, pri-
marily in groups that exhibit extensive hybridization and
repetitive introgression patterns. In addition to ITS, other tar-
542 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
get loci for genetic identification of cultivars within P. vulga-
ris could be single- or low-copy nuclear housekeeping genes.
However, the existence of high intragenomic variation can
limit the utility of ITS rDNA for phylogenetic reconstruc-
tions, especially between closely related taxa (Vollmer and
Palumbi 2004).
Molecular markers are applied in plant science to over-
come the absence of a standard characterization system and
appropriate legal protection of modern varieties and germ-
plasm resources, as previously demonstrated in the common
bean (Pallottini et al. 2004) and other major crop species
such as maize (Barcaccia et al. 2003). In this context, DNA
barcoding in plants could be profitably exploited for studying
biodiversity at the genus level, but it does not appear useful
for assessing the genetic identities of crop varieties and food-
stuffs within a species.
Acknowledgements
Thanks are due to the A. Gini Foundation (University of
Padova, Italy) to support S.N. during her internship at the
Smithsonian Institution (Washington DC). We also thank B.
Campion, Agricultural Research Council, Research Unit for
Vegetable Crops (CRA-ORL; Montanaso Lombardo, Italy),
for supplying the Italian bean varieties. Funding for this proj-
ect was provided by the Smithsonian Institution, the Ministry
of University, Research, Science, and Technology (Italy), and
the University of Padova (project CPDA087818/08).
References
Álvarez, I., and Wendel, J.F. 2003. Ribosomal ITS sequences and
plant phylogenetic inference. Mol. Phylogenet. Evol. 29(3): 417–
434. doi:10.1016/S1055-7903(03)00208-2. PMID:14615184.
Angioi, S.A., Desiderio, F., Rau, D., Bitocchi, E., Attene, G., and
Papa, R. 2009. Development and use of chloroplast microsatellites
in Phaseolus spp., and other legumes. Plant Biol. 11(4): 598–612.
doi:10.1111/j.1438-8677.2008.00143.x. PMID:19538398.
Angioi, S.A., Rau, D., Attene, G., Nanni, L., Bellucci, E., Logozzo, G.,
et al. 2010. Beans in Europe: origin and structure of the European
landraces of Phaseolus vulgaris L. Theor. Appl. Genet. 121(5):
829–843. doi:10.1007/s00122-010-1353-2. PMID:20490446.
Barcaccia, G., Lucchin, M., and Parrini, P. 2003. Characterization of
aflint maize (Zea mays var. indurata) Italian landrace. II. Genetic
diversity and relatedness assessed by SSR and Inter-SSR
molecular markers. Genet. Resour. Crop Evol. 50(3): 253–271.
doi:10.1023/A:1023539901316.
Barrett, R.D.H., and Hebert, P.D.N. 2005. Identifying spiders through
DNA barcodes. Can. J. Zool. 83(3): 481–491. doi:10.1139/z05-
024.
Brower, A.V.Z. 2006. Problems with DNA barcodes for speciesde-
limitation: ‘ten species’of Astraptes fulgerator reassessed
(Lepidoptera:Hesperiidae). Syst. Biodivers. 4(02): 127–132.
doi:10.1017/S147720000500191X.
Burle, M.L., Fonseca, J.R., Kami, J.A., and Gepts, P. 2010.
Microsatellite diversity and genetic structure among common
bean (Phaseolus vulgaris L.) landraces in Brazil, a secondary
center of diversity. Theor. Appl. Genet. 121(5): 801–813. doi:10.
1007/s00122-010-1350-5. PMID:20502861.
CBOL Plant Working Group. 2009. A DNA barcode for land plants.
Proc. Natl. Acad. Sci. U.S.A. 106(31): 12 794–12 797. doi:10.
1073/pnas.0905845106. PMID:19666622.
Chacón, M.I., Pickersgill, B., and Debouck, D.G. 2005. Domestica-
tion patterns in common bean (Phaseolus vulgaris L.) and the
origin of the Mesoamerican and Andean cultivated races. Theor.
Appl. Genet. 110(3): 432–444. doi:10.1007/s00122-004-1842-2.
PMID:15655667.
Cowan, R.S., Chase, M.W., Kress, W.J., and Savolainen, V. 2006.
300 000 species to identify: problems, progress and prospects in
DNA barcoding of land plants. Taxon, 55(3): 611–616. doi:10.
2307/25065638.
Cronn, R.C., Small, R.L., Haselkorn, T., and Wendel, J.F. 2002.
Rapid diversification of the cotton genus (Gossypium: Malvaceae)
revealed by analysis of sixteen nuclear and chloroplast genes. Am.
J. Bot. 89(4): 707–725. doi:10.3732/ajb.89.4.707.
DeSalle, R., Egan, M.G., and Siddall, M. 2005. The unholy trinity:
taxonomy, species delimitation and DNA barcoding. Philos. Trans.
R. Soc. Lond. B Biol. Sci. 360(1462): 1905–1916. doi:10.1098/
rstb.2005.1722.
Debouck, D.G., Toro, O., Paredes, O.M., Johnson, W.C., and Gepts,
P. 1993. Genetic diversity and ecological distribution of Phaseolus
vulgaris (Fabaceae) in northwestern South America. Econ. Bot. 47
(4): 408–423. doi:10.1007/BF02907356.
Delgado-Salinas, A., Turley, T., Richman, A., and Lavin, M. 1999.
Phylogenetic analysis of the cultivated and wild species of
Phaseolus (Fabaceae). Syst. Bot. 24(3): 438–460. doi:10.2307/
2419699.
Evanno, G., Regnaut, S., and Goudet, J. 2005. Detecting the number
of clusters of individuals using the software STRUCTURE: a
simulation study. Mol. Ecol. 14(8): 2611–2620. doi:10.1111/j.
1365-294X.2005.02553.x. PMID:15969739.
Falush, D., Stephens, M., and Pritchard, J.K. 2003. Inference of
population structure using multilocus genotype data: linked loci
and correlated allele frequencies. Genetics, 164(4): 1567–1587.
PMID:12930761.
Fazekas, A.J., Burgess, K.S., Kesanakurti, P.R., Graham, S.W.,
Newmaster, S.G., Husband, B.C., et al. 2008. Multiple multilocus
DNA barcodes from the plastid genome discriminate plant species
equally well. PLoS ONE, 3(7): e2802. doi:10.1371/journal.pone.
0002802. PMID:18665273.
Gepts, P., Beavis, W.D., Brummer, E.C., Shoemaker, R.C., Stalker,
H.T., Weeden, N.F., and Young, N.D. 2005. Legumes as a model
plant family. Genomics for food and feed report of the cross-
legume advances through genomics conference. Plant Physiol. 137
(4): 1228–1235. doi:10.1104/pp.105.060871. PMID:15824285.
Hajibabaei, M., Singer, G.A.C., and Hickey, D.A. 2006. Benchmark-
ing DNA barcodes: an assessment using available primate
sequences. Genome, 49(7): 851–854. doi:10.1139/G06-025.
PMID:16936793.
Hebert, P.D.N., Cywinska, A., Ball, S.L., and deWaard, J.R. 2003.
Biological identifications through DNA barcodes. Proc. Biol. Sci.
270(1512): 313–321. doi:10.1098/rspb.2002.2218. PMID:
12614582.
Hebert, P.D.N., Stoeckle, M.Y., Zemlak, T.S., and Francis, C.M.
2004. Identification of birds through DNA barcodes. PLoS Biol. 2
(10): e312. doi:10.1371/journal.pbio.0020312. PMID:15455034.
Hickerson, M.J., Meyer, C.P., and Moritz, C. 2006. DNA barcoding
will often fail to discover new animal species over broad parameter
space. Syst. Biol. 55(5): 729–739. doi:10.1080/
10635150600969898. PMID:17060195.
Kami, J., Velàsquez, V.B., Debouck, D.G., and Gepts, P. 1995.
Identification of presumed ancestral DNA sequences of phaseolin
in Phaseolus vulgaris. Proc. Natl. Acad. Sci. U.S.A. 92(4): 1101–
1104. doi:10.1073/pnas.92.4.1101. PMID:7862642.
Kimura, M. 1980. A simple method for estimating evolutionary rates
of base substitutions through comparative studies of nucleotide
sequences. J. Mol. Evol. 16(2): 111–120. doi:10.1007/
BF01731581. PMID:7463489.
Nicolè et al. 543
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Kress, W.J., and Erickson, D.L. 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(6): e508.
doi:10.1371/journal.pone.0000508 . PMID:17551588.
Kwak, M., and Gepts, P. 2009. Structure of genetic diversity in the
two major gene pools of common bean (Phaseolus vulgaris L.,
Fabaceae). Theor. Appl. Genet. 118(5): 979–992. doi:10.1007/
s00122-008-0955-4. PMID:19130029.
Lledó, M.D., Crespo, M.B., Cameron, K.M., Fay, M.F., and Chase,
M.W. 1998. Systematics of Plumbaginaceae based upon cladistic
analysis of rbcL sequence data. Syst. Bot. 23(1): 21–29.
Logozzo, G., Donnoli, R., Macaluso, L., Papa, R., Knupffer, H., and
Zeuli, P.S. 2007. Analysis of the contribution of Mesoamerican
and Andean gene pools to European common bean (Phaseolus
vulgaris L.) germplasm and strategies to establish a core
collection. Genet. Resour. Crop Evol. 54(8): 1763–1779. doi:10.
1007/s10722-006-9185-2.
Mason-Gamer, R.J., and Kellogg, E.A. 1996. Testing for phyloge-
netic conflict among molecular data sets in the tribe Triticeae
(Gramineae). Syst. Biol. 45(4): 524–545. doi:10.1093/sysbio/45.4.
524.
Meier, R., Shiyang, K., Vaidya, G., and Ng, P.K.L. 2006. DNA
barcoding and taxonomy in Diptera: a tale of high intraspecific
variability and low identification success. Syst. Biol. 55(5): 715–
728. doi:10.1080/10635150600969864. PMID:17060194.
Meyer, C.P., and Paulay, G. 2005. DNA barcoding: error rates based
on comprehensive sampling. PLoS Biol. 3(12): e422. doi:10.1371/
journal.pbio.0030422. PMID:16336051.
Mohler, V., and Schwarz, G. 2008. Genotyping tools in plant
breeding: from restriction fragment length polymorphisms to
single nucleotide polymorphisms. In Molecular marker systems in
plant breeding and crop improvement. Vol. 55. Edited by H. Lorz
and G. Wenzel. Springer, Berlin. pp. 23–38.
Nei, M. 1987. Molecular evolutionary genetics. Columbia University
Press, New York.
Newmaster, S.G., Fazekas, A.J., and Ragupathy, S. 2006. DNA
barcoding in land plants: evaluation of rbcL in a multigene tiered
approach. Can. J. Bot. 84(3): 335–341. doi:10.1139/B06-047.
Pallottini, L., Garcia, E., Kami, J., Barcaccia, G., and Gepts, P. 2004.
The genetic anatomy of a patented yellow bean. Crop Sci. 44(3):
968–977. doi:10.2135/cropsci2004.0968.
Papa, R., Nanni, L., Sicard, D., Rau, D., and Attene, G. 2006. The
evolution of genetic diversity in Phaseolus vulgaris L. In Darwin’s
Harvest: new approaches to the origins, evolution and conservation
of crops. Edited by T.J. Motley, N. Zerega, and H. Cross.
Columbia University Press, New York.
Pritchard, J.K., Stephens, P., and Donnelly, P. 2000. Inference of
population structure using multilocus genotype data. Genetics, 155
(2): 945–959. PMID:10835412.
Rach, J., DeSalle, R., Sarkar, I.N., Schierwater, B., and Hadrys, H.
2008. Character-based DNA barcoding allows discrimination of
genera, species and populations in Odonata. Proc. Biol. Sci. 275
(1632): 237–247. doi:10.1098/rspb.2007.1290. PMID:17999953.
Rossi, M., Bitocchi, E., Bellucci, E., Nanni, L., Rau, D., Attene, G.,
and Papa, R. 2009. Linkage disequilibrium and population
structure in wild and domesticated populations of Phaseolus
vulgaris L. Evol Appl. 2(4): 504–522. doi:10.1111/j.1752-4571.
2009.00082.x.
Rozas, J., Sánchez-DelBarrio, J.C., Messeguer, X., and Rozas, R.
2003. DnaSP, DNA polymorphism analyses by the coalescent and
other methods. Bioinformatics, 19(18): 2496–2497. doi:10.1093/
bioinformatics/btg359. PMID:14668244.
Sang, T., Crawford, D.J., and Stuessy, T.F. 1997. Chloroplast DNA
phylogeny, reticulate evolution, and biogeography of Paeonia
(Paeoniaceae). Am. J. Bot. 84(8): 1120–1136. doi:10.1111/j.1439-
0523.2005.01137.x.
Shaw, J., and Small, R.L. 2005. Chloroplast DNA phylogeny and
phylogeography of the North American plums (Prunus subgenus
Prunus section Prunocerasus, Rosaceae). Am. J. Bot. 92(12):
2011–2030. doi:10.3732/ajb.92.12.2011.
Sicard, D., Nanni, L., Porfiri, O., Bulfon, D., and Papa, R. 2005.
Genetic diversity of Phaseolus vulgaris L., and P. coccineus L.
landraces in central Italy. Plant Breed. 124(5): 464–472. doi:10.
1111/j.1439-0523.2005.01137.x.
Taberlet, P., Gielly, L., Pautou, G., and Bouvet, J. 1991. Universal
primers for amplification of three non-coding regions of
chloroplast DNA. Plant Mol. Biol. 17(5): 1105–1109. doi:10.
1007/BF00037152 .
Tamura, K., Dudley, J., Nei, M., and Kumar, S. 2007. MEGA4:
Molecular Evolutionary Genetics Analysis (MEGA) software
version 4.0. Mol. Biol. Evol. 24(8): 1596–1599. doi:10.1093/
molbev/msm092. PMID:17488738.
Tate, J.A., and Simpson, B.B. 2003. Paraphyly of Tarasa (Malvaceae)
and diverse origins of the polyploidy species. Syst. Bot. 28(4):
723–737. doi:10.1016/S0169-5347(02)00041-1.
Tautz, D., Arctander, P., Minelli, A., Thomas, R.H., and Vogler, A.P.
2003. A plea for DNA taxonomy. Trends Ecol. Evol. 18(2): 70–74.
doi:10.1016/S0169-5347(02)00041-1.
Tingshuang, Y., Miller, A.J., and Wen, J. 2004. Phylogenetic and
biogeographic diversification of Rhus (Anacardiaceae) in the
Northern Hemisphere. Mol. Phylogenet. Evol. 33(3): 861–879.
doi:10.1016/j.ympev.2004.07.006. PMID:15522809.
Tsai, L.-C., Wang, J.-C., Hsieh, H.-M., Liu, K.-L., Linacre, A., and
Lee, J.C. 2008. Bidens identification using the noncoding regions
of chloroplast genome and nuclear ribosomal DNA. Forensic Sci.
Int. Genet. 2(1): 35–40. doi:10.1016/j.fsigen.2007.07.005. PMID:
19083787.
Velzen, R., Bakker, F.T., and Loon, J.J.A. 2007. DNA barcoding
reveals hidden species diversity in Cymothoe (Nymphalidae).
Proc. Neth. Entomol. Soc. Meet. 18:95–103.
Vences, M., Thomas, M., Bonett, R.M., and Vieites, D.R. 2005.
Deciphering amphibian diversity through DNA barcoding:
chances and challenges. Philos. Trans. R. Soc. Lond. B Biol.
Sci. 360(1462): 1859–1868. doi:10.1098/rstb.2005.1717. PMID:
16221604.
Vollmer, S.V., and Palumbi, S.R. 2004. Testing the utility of
internally transcribed spacer sequences in coral phylogenetics.
Mol. Ecol. 13(9): 2763–2772. doi:10.1111/j.1365-294X.2004.
02265.x. PMID:15315687.
Ward, R.D., Zemlak, T.S., Innes, B.H., Last, P.R., and Hebert, P.D.N.
2005. DNA barcoding Australia’s fish species. Philos. Trans. R.
Soc. Lond. B Biol. Sci. 360(1462): 1847–1857. doi:10.1098/rstb.
2005.1716. PMID:16214743.
White, T.J., Bruns, T., Lee, S., and Taylor, J.W. 1990. Amplification
and direct sequencing of fungal ribosomal RNA genes for
phylogenetics. In PCR Protocols: a guide to methods and
applications. Edited by M.A. Innis, D.H. Gelfand, J.J. Sninsky,
and T.J. White. Academic Press, Inc., New York. pp. 315-322.
Whitlock, B.A., Hale, A.M., and Groff, P.A. 2010. Intraspecific
inversions pose a challenge for the trnH-psbA plant DNA barcode.
PLos ONE, 5(7): e11533. doi:10.1371/journal.pone.0011533.
Wiemers, M., and Fiedler, K. 2007. Does the DNA barcoding gap
exist? —a case study in blue butterflies (Lepidoptera: Lycanidae).
Front. Zool. 4: 8. doi:10.1186/1742-9994-4-8. PMID:17343734.
Will, K.W., and Rubinoff, D. 2004. Myth of the molecule: DNA
barcodes for species cannot replace morphology for identification
and classification. Cladistics, 20(1): 47–55. doi:10.1111/j.1096-
0031.2003.00008.x.
544 Genome, Vol. 54, 2011
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
Will, K.W., Mishler, B.D., and Wheeler, Q.D. 2005. The perils of
DNA barcoding and the need for integrative taxonomy. Syst. Biol.
54(5): 844–851. doi:10.1080/10635150500354878. PMID:
16243769.
Wojciechowski, M.F., Lavin, M., and Sanderson, M.J. 2004. A
phylogeny of legumes (Leguminosae) based on analysis of the
plastid matK gene resolves many well-supported subclades within
the family. Am. J. Bot. 91(11): 1846–1862. doi:10.3732/ajb.91.11.
1846.
Wong, E.H., and Hanner, R.H. 2008. DNA barcoding detects market
substitution in North American seafood. Food Res. Int. 41(8):
828–837. doi:10.1016/j.foodres.2008.07.005.
Nicolè et al. 545
Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by BIBLIO UNIVERSITARIA DI on 09/15/11
For personal use only.
