DNA BARCODING AND CLASSICAL TAXONOMY: AN INTEGRATED
APPROACH TO THE IDENTIFICATION OF ONISCIDEAN ISOPODS
Bianca M. Lombardo2, Michela Barbuto1, Giuseppe Montesanto2, Nicolò Alfano1, Andrea Galimberti1,
Adriana Bellati3, Domenico Caruso2, Maurizio Casiraghi1
Dipartimento di Biotecnologie e Bioscienze, Università di Milano Bicocca, Milan, Italy. E-mail:
Dipartimento di Biologia “Marcello La Greca”, Università di Catania, Catania, Italy. E-mail:
Dipartimento di Biologia Animale, Università degli Studi di Pavia, Pavia, Italy
ABSTRACT: The crustacean order Isopoda includes thousands of species inhabiting a wide variety of habitats,
both aquatic and terrestrial. In particular the suborder Oniscidea includes about 4,000 species of terrestrial
isopods but probably this number represents only a fraction of their real diversity. A morphological
identification at the species level is often difficult to reach and usually requires a careful dissection of animals by
experienced taxonomists. In this study an integrated approach, based on traditional taxonomy and DNA
barcoding, has been used to characterize some species of Sicilian terrestrial isopods (suborder Oniscidea). We
produced coxI barcode sequences for all the analyzed specimens, and we compared them with reference entries
retrieved from GenBank. We calculated an Optimum Threshold (OT) of molecular divergence to discriminate
among species by using a bioinformatic approach previously developed by our laboratory research team. Results
show a rather high level of coherence (near 70%) between classical and molecular approaches. All of the
ambiguous cases refer to species with problematic taxonomy due to false positive or false negative errors.
Keywords: Oniscidea, DNA barcoding, coxI, integrated taxonomy
1. Introduction stored in ethanol 95% following the procedures
specified in the Biorepositories initiative
The suborder Oniscidea includes about 4,000
(http://www.biorepositories.org/) and catalogued
species of terrestrial isopods, worldwide distributed
into the collection of Milano Bicocca (MIB)
and well adapted to almost all terrestrial
institution. DNA extracts were prepared using the 5
environments, ranging from deserts to alpine fields
PRIME, ArchivePure DNA Purification Kit.
(Klossa-Kilia et al., 2006).
Oniscidea coxI amplifications and sequencing were
However it is thought that this number of species
obtained using the standard primer pair LCO1490
represents only a fraction of their real diversity. In
and HCO2198 (Folmer et al., 1994). PCRs products
fact those studies based only on morphology may
were gel purified and directly sequenced. K2P
underestimate the true level of divergence among
molecular distances (i.e. intraspecific, interspecific,
oniscidean populations due to the controversial use
overall mean K2P) and relative standard errors have
of few secondary sexual male characters as
been calculated using MEGA 4.0 (Tamura et al.,
diagnostic characters (e.g. first and second pair of
2007). In order to perform a DNA barcoding
pleopods, seventh pereopod etc).
analysis based on an integrated approach (with both
In the last years molecular techniques become more
morphological and molecular evidences), an
and more significant in taxonomic studies, mainly
optimum threshold value of genetic divergence
when they are integrated with traditional taxonomic
(OT) was calculated following Ferri et al. (2009).
approaches (e.g. morphology). In particular, DNA
The OT parameter represents the value that
barcoding, a technique based on the analysis of the
minimizes the cumulative error (i.e.: the total
level of molecular variability in a short and
amount of identification mismatches when
standardized nucleotide sequence (e.g. the
morphological and molecular data are compared)
mitochondrial coxI for almost all metazoans), is a
and at the same time maximizes the coherence
useful tool to evaluate differences between species
between morphological and molecular
(Hebert et al., 2003) especially for cryptic species
identification. A phenetic tree was generated using
complexes (Yassin et al., 2008; Stahls et al., 2008).
MEGA 4.0 (Tamura et al., 2007).
In this work, an integrated approach based on
traditional taxonomy and DNA barcoding was used
3. Results & Discussion
to characterize the terrestrial isopods diversity with
a major attention to Sicilian species. We firstly analyzed the most comprehensive
molecular dataset that included 119 barcode
2. Materials & Methods sequences from 25 oniscidean species already
present in GenBank, integrated with the 44 coxI
In the present work a total of 44 biological samples
sequences belonging to 12 species produced in this
from Oniscidean isopods were collected during
study. The OT obtained from this dataset was 9.8%
2009 in Italy (Sicily). The biological samples were
and shows a minimum cumulative error of 9.7%. It
showed a rather high level of coherence between
morphological and molecular approach. In fact, for
26 species out of 37 (70.3%) the molecular data
completely overlapped with morphological
identifications. However, we detected some
inconsistencies. In 11 cases we found a problematic
situation due to two types of error: false positives
(eight cases) and false negatives (three cases). In
the first category the intraspecific K2P genetic
divergence was higher than OT, that is, specimens
belonging to the same species were identified as
different MOTUs (Molecular Operational
Taxonomical Unit, sensu Floyd et al., 2002). For
most of these species (Ligidiuim spp. a high
intrapopulation variation was already reported in
literature (Klossa-Kilia et al. 2006; Montesanto et
al., 2007). The false negatives were due to
sequences belonging to different species that were
identified as being part of the same MOTU. This
was the situation of three species belonging to the
Porcellio imbutus complex (Viglianisi et al., 1992).
Considering these inconsistencies we excluded
from the dataset all those barcode sequences
belonging to the species suffering from taxonomic
inconsistencies or incongruences and we generated
a new molecular dataset that includes a total of 97
coxI sequences relative to 26 oniscidean species. Figure 1: Neighbour-Joining tree of 96 cytochrome c
The classical DNA barcoding analysis based on oxidase subunit I sequences from 26 Onicidean species,
K2P distance matrix produce the following results: using Kimura 2-parameter distances (bootstrap values
coxI mean nucleotide distance within species 1% greater than 70 shown). In round brackets is reported the
(standard error 0.2%; range 0-5.4%); coxI mean number of specimen for each species.
nucleotide distance between species 26.0%
(standard error 2.3%; range 9.4-35%); coxI overall 4. References
mean diversity 23.2% (standard error 1.2%).
Ferri E, Barbuto M, Bain O, Galimberti A, Uni S,
The OT obtained from this second dataset is 8.8%
Guerriero R, Ferté H, Bandi C, Martin C,
and generates a minimum cumulative error of 0%.
Casiraghi M. 2009. Frontiers in Zoology 6.
At the threshold level of 8.8% any types of errors
Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek
(e.g. false positive and false negative) occurred.
R. 1994. Mol. Marine Biol. Biotechnol. 3: 294–
A schematic view of the relationship among the
oniscidean species was reported in the phenetic tree
Floyd R, Abebe E, Papert A, Blaxter M. 2002.
obtained with the NJ approach (Fig. 1).
Molecular Ecology. 11(4): 839-850.
Hebert PDN, Cywinska A, Ball SL, deWaard JR.
In conclusion, our results clearly show that DNA
2003. Proc. R. Soc. Lond. B. 270: 313–321.
barcoding represents a potentially useful tool for
Klossa-Kilia E, Kilias G, Tryfonopoulos G,
taxonomy investigations on oniscidean isopods, but
Koukou K, Sfenthourakis S, Parmakelis A. 2006.
without the integration of traditional approaches
Zool. Scr. 35: 459–472.
could become a simple collection of MOTUs.
Montesanto G, Caruso D, Lombardo BM. 2007.
Given this assumption, it is reasonable to affirm
Crustaeana. 80(8): 917-938.
that the establishment, improvement and
Stahls G, Savolainen E. 2008. Molecular
maintenance of DNA barcoding as a taxonomic tool
Phylogenetics and Evolution. 46: 82–87.
cannot prescind from a long-lasting interaction
Tamura K., Dudley J, Nei M, Kumar S. 2007.
between traditional taxonomy and DNA-based
MEGA4: Molecular Biology and Evolution. 24:
Viglianisi F, Lombardo BM, Caruso D. 1992.
Animalia (Catania). 19: 235-273.
Yassin A, Capy P, Madi-Ravazzi L, Ogereau D,
David J.R. 2008. Molecular Ecology Resources.