Phylogeography of Helleria brevicornis Ebner 1868 (Crustacea, Oniscidea): Old
and recent differentiations of an ancient lineage
, Alessandro Campanaro
, Monica Carosi
, Valerio Sbordoni
, Roberto Argano
Department of Biology, University ‘‘Tor Vergata”, Via della Ricerca Scientiﬁca, 1, 00133 Rome, Italy
Department of Animal and Human Biology, University ‘‘La Sapienza”, Viale dell’Università, 32, 00185 Rome, Italy
Centro Nazionale per lo Studio e la Conservazione della Biodiversità Forestale ‘‘Bosco Fontana”Corpo Forestale dello Stato, Via Carlo Ederle 16/a, I-37100 Verona, Italy
Received 30 March 2009
Revised 21 September 2009
Accepted 5 October 2009
Available online 21 October 2009
Helleria brevicornis has a disjunct, peri-Tyrrhenian distribution that mirrors that of several organisms, for
which geographic vicariance is invoked, due to the geological events started with the Oligocene split of
the Sardo-Corsican microplate from the Pyrenees, and successive separation between Sardinia and Cor-
sica. Molecular phylogenetic analyses demonstrated that such a biogeographic model does not apply
to Helleria. The original split of the Sardinian and Corsican lineages originated in the Early Pliocene. Fur-
ther diversiﬁcation occurred later. The colonization of the Tuscan archipelago, French, and Italian main-
land occurred most recently, but a possible time dependency bias suggests that such colonization was
driven by human-transport.
Ó2009 Elsevier Inc. All rights reserved.
Oniscidea is certainly the largest group of Crustaceans that
Oniscidea, Helleria brevicornis is one of the most intriguing species.
Its origin and phylogenetic relationship have been debated for long
time, as many morphological characters separate H. brevicornis from
the rest of Oniscidea. For such features, this species has been consid-
ered the remnant of a branch that diverged early from the lineage
that led to other Oniscidea (see literature in Vandel, 1960).
Current systematics reﬂects such a hypothesis. H. brevicornis is
the only recognized species of the genus. So far, Helleria and Tylos
(which comprises several species) are the only two genera included
in the family Tylidae, although in two separate subfamilies (Vandel,
1960). In the case of Helleria, mitochondrial DNA provided evidence
for an early divergence from the rest of Oniscidea and suggested a
distinctive pattern of molecular evolution, as indicated by an extre-
mely high adenine–thymine (AT) richness (Michel-Salzat and Bou-
chon, 2000). However, those data did not resolve the phylogenetic
relationships of this genus, which still remain to be investigated.
Helleria lives in the litter and humus horizons of soil, in forests
from the sea level up to 1200 m (Vandel, 1960). This taxon is ende-
mic to the North Tyrrhenian area. In fact, its geographic distribution
is limited to spotted locations in Sardinia, Corsica, continental France
(Esterel), a few islands of the Tuscan archipelago, and continental
Italy (Massoncello, Tuscany), as indicated in Fig. 1. Such a disjunct
peri-Tyrrhenian distribution, which mirrors that of several organ-
isms, as for example earthworms (Omodeo and Rota, 2008), insects
(Caccone and Sbordoni, 2001), salamanders (Caccone et al., 1997),
is generally explained in terms of geographic vicariance. Often,
vicariance is interpreted as due to the geological events started with
the Oligocene split of the Sardo-Corsican microplate from the
Pyrenees, followed by the separation between Sardina and Corsica
(Caccone et al., 1997; Caccone and Sbordoni, 2001; Grill et al.,
2007). Paleomagnetic and tectono-sedimentary data suggest that,
between 33 Million years ago (Ma) and 21–20.5 Ma, Sardinia–
Corsica microplate rotated 35–30°clockwise, with a successive
counterclockwise rotation, terminated about 17.5–18 Ma (Edel
et al., 2001). The microplate may have started breaking into Sardinia
and Corsica (Boccaletti et al., 1990), which separated completely
about 9 Ma (Alvarez, 1972; Bellon et al., 1977; Bonin et al., 1979).
Such a biogeographic model of evolution implies that the iso-
lated populations may undergo independent evolution which
may ultimately lead to speciation, generating closely related but
distinct species without sympatric occurrence. Phylogenetic rela-
tionships between these species should reﬂect the temporal conse-
cution of geological events.
Although the old origin of the Helleria lineage and its present
distribution could potentially ﬁt the vicariance model, the phyloge-
ography of this species has never been studied and the predictions
of the model have not been tested, yet.
1055-7903/$ - see front matter Ó2009 Elsevier Inc. All rights reserved.
These authors equally contributed to the work
*Corresponding author. Fax: +39 06 72595965.
E-mail address: email@example.com (G. Gentile).
Molecular Phylogenetics and Evolution 54 (2010) 640–646
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Here we use mtDNA sequence data to investigate the phyloge-
ographic relationships between multiple populations from the en-
tire area of distribution of H. brevicornis. We combined classical
phylogenetic and network analyses with a coalescence approach
to elucidate times and modes of evolution and to test the predic-
tions of the vicariance model.
The 54 individuals of H. brevicornis used in this study were col-
lected from 21 localities, representing the entire area of distribu-
tion of the species. Sampling locations and the number of
individuals investigated are reported in Fig. 1. Two individuals
belonging to Tylos europaeus were also collected from the seashore
near Brindisi (Apulia).
2.2. DNA extraction, PCR ampliﬁcation and sequencing
Whole DNA was extracted from 2–4 legs using C-Tab protocol
(Doyle and Doyle, 1987).
Fragments of the Cytochrome Oxydase I gene (COI), the large
ribosomal subunit (16S) and the small ribosomal subunit (12S) of
mtDNA were PCR ampliﬁed. PCR ampliﬁcations were performed
in a volume of 50
L containing 1X Reaction buffer, 5 mM MgCl
0,2 mM dNTP, 0.5
M each primer, 1, 5 units EuroTaq (Euroclone,
UK). The following speciﬁc primer pairs were designed and used
for both PCR amplifying and sequencing fragments of COI, 12S,
and 16S, respectively: (i) heCOIfor TCCTTTAGCTTCAAATATTGC,
heCOIrev CATAATGGAAATGAGCTACAAC; (ii) 12S aih1 GGATTAGAT
ACCCTACTATTTATAAAC, 12S bih1 GACGGGCGATGTGTACTTATTC;
(iii) 16S tyhefor ATATTGACTGTGCTAAGGTAGC, 16S tyherev CT
TAATCCAACATCGAGGTC. These primers were successfully used
for H. brevicornis and T. europaeus. PCR conditions were: 2 min at
94 °C followed by 35 cycles of 30 s at 94 °C, 45 s at 46 °C, 1 min
at 72 °C and a ﬁnal elongation of 5 min at 72 °C. PCR products were
puriﬁed by using ExoSap-IT (USB). Sequences were run on ABI
PRISM 3100 sequencer and edited with SEQUENCHER 4.1.2 (Gene
Codes). The alignment obtained by using CLUSTAL X (Thompson
et al., 1997) was checked by eye. The ﬁnal dataset consisted in
631 bp of COI, 365 bp of 12S, and 331 bp of 16S. GenBank accession
numbers for the sequences generated and used in this study are
GU097457–GU097510 for the 12S, GU097511–GU097564 for
16S, and GU097565–GU097618 for COI of H. brevicornis. Accession
numbers are GU097623–GU097626 for 12S, GU097627–GU097630
for 16S, and GU097619–GU097622 for COI of T. europaeus.
As it has been shown that the occurrence of Wolbachia may alter
the levels and patterns of mtDNA diversity, leading to biased infer-
ence of organisms’ evolutionary history (Hurst and Higgins, 2005),
we used the primer pair wsp81F and wsp691R (Zhou et al., 1998)
to test for the occurrence of Wolbachia in our samples.
2.3. Phylogenetic analyses
Phylogeny inference was performed using the three datasets
both separately and combined. We tested for phylogenetic incon-
gruence of different datasets by performing the incongruence
length difference test (ILD; Farris et al., 1994), after removing the
invariant characters (Cunningham, 1997). Analyses were con-
ducted using maximum parsimony (MP; Farris, 1970), maximum
likelihood (ML; Felsenstein, 1981), and Bayesian inference (Bayes,
1763), as implemented in PAUP* 4.0b10 (Swofford, 2002), TREE-
FINDER 2008 (Jobb, 2008) and MRBAYES 3.1.2 (Ronquist and Huel-
senbeck, 2003), respectively. PAUP* 4.0b10 was also used to
estimate uncorrected p-distances. MP heuristic parameters were
as follows: starting trees obtained by random-addition (10 repli-
cates) and TBR branch swapping. Gaps in the 12S were single-base
deletions and were coded as ﬁfth character state.
ML trees were sought via a genetic algorithm, less prone to get
trapped in local optima (Jobb, 2004). Different models were used
for the COI, 12S, and the 16S data. MODELTEST 3.7 (Posada and
Crandall, 1998) and the corrected Akaike information criterion
(AICc; Hurvich and Tsai, 1989) was used to select the HKY85 + I
for the 12S (Ti/Tv rate = 19.6907; I= 0.8612) and 16S (Ti/Tv
rate =9.5184; I= 0.8791) genes. The GTR +
Imodel was chosen
by the ‘‘model proposer” in TREEFINDER and was used for the
COI, with all parameters estimated separately for ﬁrst, second,
and third positions. The same models were used in the Bayesian
analysis. Gaps were recoded as binary data and considered as a
separate partition to which a binary (restriction) model of evolu-
tion was applied as implemented in MRBAYES. For each partition,
parameters’ values were estimated during the search. First, second
and third positions of the COI were treated as separate partitions.
For MP and ML, nodal support was tested by bootstrapping (Fel-
senstein, 1985), with 1000 pseudoreplicates. T. europaeus was used
as outgroup within the family Tylidae.
Average genetic distances were calculated between populations
within and between each group resulting from the phylogenetic
analysis. They were estimated starting from pairwise tip-to-tip dis-
tances along the ML tree.
Network analysis was performed using the TCS software (Clem-
ent et al., 2000), which implements the method of statistical parsi-
mony developed by Templeton (Templeton et al., 1992). This
method is not constrained by many of the assumptions of phyloge-
netic reconstruction techniques as it does not assume that the
ancestral sequence is missing and does not require bifurcating
relationships. Since this method performs poorly in case of phylog-
Fig. 1. Sampling localities. For Corsica, gray area is South Corsica and white area is
Upper Corsica. Numbers within square brackets indicate individuals used for COI,
12S, and 16S analyses, respectively. (1) Esterel: rue du Pic del’Ours [2, 4, 3]. (2)
Esterel: carfour de la Colle Douce [2, 4, 4]. (3) Capraia: port [4, 4, 4]. (4) Masson-
cello: Populonia [4, 4, 4]. (5) Elba: Portoferraio [3, 3, 3]. (6) Pianosa [1, 3, 3]. (7) Col
de Saint Nicolas [4, 4, 4]. (8) Marinca [3, 1, 2]. (9) Crosciano [3, 2, 1]. (10) Aregno
[2, 2, 2]. (11) Murato [2, 0,1]. (12) Castirla [4, 4, 4]. (13) Vizzavona: Plateau Monte
d’Oro [7, 5, 8]. (14) Bocognano [1, 1, 1]. (15) Tolla [2, 2, 2]. (16) Col de Bavella
[4, 4, 4]. (17) Casalabriva [7, 5, 5]. (18) Foret de l’Ospedale [4, 2, 4]. (19) Ierzu
[3, 3, 3]. (20) Fluminimaggiore [2, 4, 4]. (21) Burcei [4, 4, 4].
G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646 641
enies with long branches, it is particularly suitable only for popu-
lation level analysis.
Based on results of phylogenetic analyses, we investigated the
hypothesis of molecular clock for the COI gene by applying the
‘‘global molecular clock test — all roots”, as implemented in
HY-PHY (ver.0.99 beta, Pond et al., 2005). We used the same set-
tings as for the ML phylogenetic analysis. The test was run start-
ing from a ML unrooted tree with six terminal nodes obtained
by eliminating poorly diverging sequences (less than 1%). The
test performed the likelihood ratio test of the null hypothesis
of global molecular clock, versus the alternative hypothesis of
no molecular clock, on every possible rooted version of the tree
(i.e. 2N3, for a binary unrooted, where Nis the number of
We applied coalescence to estimate divergence times for the
monophyletic clades found by the phylogenetic analyses. We used
the BEAST package v1.4.7 to run an analysis for 20 million steps of
the MCMC, using the GTR +
Imodel partitioned by codon posi-
tion, under the assumption of strict clock (Drummond et al.,
2006). Convergence was assessed with ESS (effective sample size)
values, after a burning of 5 million steps. As no obvious calibration
point is available for Helleria, we used, as a ﬁx mean substitution
rate, the rate estimated by Ketmaier and collaborators (2003) for
the COI gene in Crustacea Isopoda Asellota. We used the ‘‘constant
size” option, as a parametric demographic function of population
size. This choice was suggested by the statistics D(Tajima, 1989)
(Fu, 1997) which, under neutrality, can identify the effects
of demographic changes, with demographic expansions being re-
ﬂected in large negative values. Dand F
were estimated for the
monophyletic clades, identiﬁed by the phylogenetic and network
analyses, for which it was reasonable to hypothesize a population
expansion. Dand F
statistics were calculated by using DNASP
v4.50.3 (Rozas et al., 2003).
The PCR ampliﬁcation of Wolbachia genes was unsuccessful in
all our samples.
> 0.99) and separate phylogenetic analyses
of the 12S, 16S, and COI datasets as well as different methods
of inference all produced very similar results. However, indepen-
dent of the approach used, the disproportionate AT richness and
multiple substitutions, typical of Helleria, impeded the correct
rooting of the Helleria tree, when Tylos was used as outgroup. In
fact, the root of the Helleria tree, as suggested by Tylos appeared
inconsistent (Fig. 2). Despite we could not use an outgroup to root
our trees, all phylogenetic analyses indicated four homogeneous
clusters: A, B, C, and D. Clusters A and B included only popula-
tions from Sardinia, whereas clusters C and D comprised the
remaining populations. Average ML genetic distances within clus-
ters A, B, C, and D were: 0.0150 ± 0.0124 (standard deviation);
0.0285 ± 0.0004; 0.0045 ± 0.0081; and 0.0042 ± 0.0051, respec-
tively. Average ML genetic distances between clusters were all
very high, ranging between 0.0719 ± 0.0024 (B vs. C) and
0.0949 ± 0.0035 (A vs. D). A complete synopsis of ML and p-dis-
tances between groups, as resulting from the phylogenetic analy-
sis, is shown in Table 1.
As expected, the molecular clock hypothesis was rejected when
reduced trees were rooted at terminal nodes (Pranging from
0.0003 to 0.0273, with only one exception: P= 0.2389). However,
the hypothesis could not be rejected when rooting at internal
nodes (Pranging from 0.2750 to 0.4561). Thus, rooting the Helleria
tree at the midpoint seemed reasonable. When the root was placed
at the midpoint Sardinian and Corsican populations formed two
separate and monophyletic clades (Figs. 2 and 4a).
The network analysis indicated the genealogical relationships
between the 25 haplotypes found (Fig. 3). They were distributed
Fig. 2. Phylogenetic tree based on maximum likelihood. The tree was rooted at the midpoint, after testing for the hypothesis of global clock. The black arrow indicates the
position of the root. The white arrow indicates the position of the root when Tylos was used as outgroup. Such a rooting is a statistical artifact. Robustness of nodes is reported
for MP, ML, and Bayesian analyses, respectively.
642 G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646
in six haplogroups, when the statistical signiﬁcance level was set at
0.95%. Eight haplotypes exclusive of Sardinia were distributed in
haplogroups 1, 2, (which together corresponded to cluster A of
phylogenetic analyses) and 3 (cluster B). Eight haplotypes from
Upper Corsica, continental France, Tuscan islands, and continental
Italy were distributed in haplogroups 4 and 5 (together corre-
sponding to cluster D), whereas haplogroup 6 (cluster C) com-
prised nine haplotypes occurring in an area mostly
corresponding to South Corsica (only haplotypes 5 and 6 were
found in Upper Corsica). Haplogroups 1 and 2 were linked by a
27-step long path, whereas 25 steps separated haplogroups 4 and
5. However, these connections were not statistically supported.
Despite the star-like shape of haplogroups 5 and 6 could suggest
a recent expansion, both D and F
estimates were not statistically
Results of the BEAST analysis are summarized in Fig. 4. For each
monophyletic clade, the posterior probability distributions of
divergence time were bell-shaped, with minor, negligible skew-
ness occurring in a few cases (Fig. 4a). The means of such distribu-
tions indicate that three separate periods of diversiﬁcation may be
identiﬁed, along the Helleria lineage. Such periods are better illus-
trated in Fig. 4b. Although with large interval of conﬁdence, three
events appear to have occurred in the ﬁrst period, between 4 and
5 Ma. The split of the basal lineages (Sardinia and Corsica) dates
back to about 4.8 Ma. Perhaps 700–800 thousand years later, two
major lineages in Sardinia and two in Corsica have originated
almost contemporarily, leading to the clades A, B and C, D, respec-
tively. A second period of differentiation occurred between 1 and
1.5 Ma, within clades A, C, and D. A third, more recent, period
would set the timing of the diversiﬁcation within clades C and D
(haplogroups 6, and [4 + 5], respectively) between 200 and 450
thousand years ago (Ka).
The PCR survey failed to detect the occurrence of Wolbachia in
our samples. However, this can not permit to completely exclude
the presence of Wolbachia in Helleria, as our sample is limited
and different populations of arthropods may show different levels
of infection (Hurst and Jiggins, 2005).
Our results conﬁrm the distinctiveness of Helleria, which per-
haps could be assigned to a separate family. Clearly, these data
unambiguously falsify the hypothesis that current distribution of
Helleria may be explained by vicariance related to the Oligocene–
Miocene dynamics. Most likely, the basal split between Corsican
and Sardinia lineages occurred in the Early Pliocene. It was a con-
sequence of insularity conditions, restored after the Mediterranean
was re-ﬂooded at the end of the Messinian crisis of salinity
(5.33 Ma; Popov et al., 2006). In fact, according to the double-phase
model of the Messinian crisis (Clauzon et al., 2001), the Mediterra-
nean sea level lowered initially by less than 100 m during a ﬁrst
phase (5.8 Ma) that affected only the Mediterranean basin margins.
A ﬂooding event followed. In a second phase (5.6 Ma) a drastic
ML and uncorrected p-distances between groups, as resulting from the phylogenetic
ML distance Uncorrected p-distance
Average SD Average SD
12S + 16S + COI
Cvs. D 0.0802 0.0019 0.0496 0.0020
(A + B) vs. (C + D) 0.0842 0.0094 0.0498 0.0038
Avs. B 0.0721 0.0030 0.0458 0.0009
Avs. D 0.0949 0.0035 0.0545 0.0012
Bvs. D 0.0782 0.0019 0.0486 0.0012
Avs. C 0.0885 0.0038 0.0508 0.0014
Bvs. C 0.0719 0.0024 0.0443 0.0006
Cvs. D 0.0983 0.0060 0.0607 0.0037
(A + B) vs. (C + D) 0.1202 0.0151 0.0685 0.0067
Avs. B 0.1199 0.0021 0.0695 0.0012
Avs. D 0.1319 0.0025 0.0732 0.0047
Bvs. D 0.1029 0.0017 0.0685 0.0028
Avs. C 0.1343 0.0061 0.0719 0.0043
Bvs. C 0.1053 0.0058 0.0586 0.0015
Cvs. D 0.0502 0.0040 0.0369 0.0025
(A + B) vs. (C + D) 0.0415 0.0080 0.0306 0.0046
Avs. B 0.0362 0.0050 0.0256 0.0031
Avs. D 0.0403 0.0049 0.0325 0.0028
Bvs. D 0.0298 0.0028 0.0254 0.0028
Avs. C 0.0467 0.0050 0.0291 0.0030
Bvs. C 0.0474 0.0030 0.0351 0.0035
Cvs. D 0.0663 0.0026 0.0427 0.0024
(A + B) vs. (C + D) 0.0586 0.0114 0.0354 0.0060
Avs. B 0.0312 0.0009 0.0230 0.0015
Avs. D 0.0738 0.0021 0.0430 0.0023
Bvs. D 0.0593 0.0019 0.0360 0.0009
Avs. C 0.0567 0.0020 0.0343 0.0024
Bvs. C 0.0421 0.0018 0.0271 0.0019
Fig. 3. Statistical parsimony network. The genealogical relationships between the
25 mtDNA haplotypes (12S + 16S + COI) are shown. Six groups of haplotypes are
identiﬁed at the 95% signiﬁcance level groups 1, 2, and 3 include haplotypes from
Sardinia. Haplogroups 4 and 6 comprise only haplotypes found in Corsica. Group 5
includes haplotypes occurring in Upper Corsica, continental France, some Tuscan
islands, and continental Italy.
G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646 643
drop in sea level occurred, that exceeded 1500 m and affected the
whole basin. After the crisis, Corsica and Sardinia joined again dur-
ing the Pliocene (Meulenkamp and Sissingh, 2003). Connections
occurred also in the Pleistocene and are reﬂected in the vertebrate
fauna, which suggests a ﬂow, from Italian mainland to Sardina, via
Corsica, ended in the Middle-Pleistocene (Masini et al., 2002). Our
data, however, do not provide evidence that Helleria might have
dispersed along that route.
Events of differentiation occurred between 1 and 1.5 Ma could
be a reﬂection of the climatic changes that occurred in the Mediter-
ranean at the end of the Pliocene through the Early Pleistocene. In
fact, Mediterranean-type climate, with a marked seasonality, con-
solidated in the Mediterranean region about 3.2–3.6 Ma (Suc,
1984; Tzedakis, 2007), although it might have intermittingly ap-
peared in the Tertiary. Such a climate was characterized by a per-
iod of obliquity-dominated cycles of glaciations (2.8–1 Ma), which
determined marked environmental transformations and ecological
successions in the Mediterranean area (Tzedakis, 2007). Pliocene
and Pleistocene climatic events might have maintained the two
lineages, leading to the separated haplogroups [4 + 5] and 6. In fact,
during the Pleistocene, the main mountain range, crossing Corsica
from North-West to South-East, was repeatedly and extensively
covered by glaciers (Hughes et al., 2006) impeding Helleria dis-
persal. For its overall high altitude, culminating with Monte Cinto
(2710 m a.s.l.), such a barrier would be difﬁcult to overcome even
during interglacial periods (haplotypes 5 and 6 indicate a recent
The interpretation of the latest events requires caution. On one
hand, our data would indicate that climatic changes of the Pleisto-
cene, as documented by isotopic (Deuterium concentration) data
(EPICA, 2004) in the marine isotope stage (MIS) 12 and between
MIS 6–10 (formerly recognized as alpine Riss and Mindel glacia-
tions; Fig. 4b) could have driven the diversiﬁcation within haplo-
groups 5 and 6. According to this model, marine regressions,
associated with glacial maxima (Kukla, 2005), should have pro-
vided facilitated conditions for dispersal to continental France (Est-
erel) and for the colonization of some of the islands of the Tuscan
archipelago, some of which were joint by land bridges and con-
nected with the Italian mainland. Both phylogenetic and network
analyses indicate that such colonization might have started from
Corsican Cape, in the North-East region of Upper Corsica. Some
support to this scenario, that implies recent colonization and dif-
ferentiation, would also be provided by a few haplotypes (10, 13,
14 and 15) with narrow distribution. However, we must remark
Fig. 4. (a) Coalescence tree, based on COI sequence data, scaled along the time axis. At each node, 95% High Posterior Density intervals and TMRCA (Time to Most Recent
Common Ancestor) posterior probability densities are also shown. Colors indicate the same clades as in Fig. 2. (b) Sea level variation (meters), stratigraphy (in terms of Marine
Isotopic Stages, MIS), and DeltaD/permil concentration (EPICA, 2004), over the last 0.5 Ma. Sea level variations and stratigraphy are redrawn after Kukla (2005).
644 G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646
that these haplotypes are only one-two steps different from the
most widely distributed haplotype 11. Additionally, their narrow
distribution might be an effect of limited sampling.
Although similar processes of dispersal have been invoked to
explain patterns of distribution and genetic variation in several
organisms during the Late Pleistocene (Dapporto et al., 2007;
Fattorini, 2009), the case of Helleria remains controversial. In fact,
the direct colonization of continental France via a hypothetical
Pleistocene land bridge is not supported by any palinspastic recon-
struction, or by vertebrate fossil records.
While we believe that our data certainly demonstrate that the
colonization of the Tuscan archipelago, Italian and French main-
land occurred recently, we suspect that our time estimate for such
event might suffer from a ‘‘time-dependency” bias, as recently de-
scribed by Ho et al. (2005). They provided evidence of a relation-
ship between the rate of molecular evolution and the age of
calibration that can be described by vertically translated exponen-
tial decay curves. Such curves identify measurable transitions be-
tween short- and long-term substitution rate, around 1–2 Ma.
Thus, as the substitution rate that we used was obtained by setting
two calibration points at 29 and 9 Ma (Ketmaier et al., 2003), our
estimates of the time for the most recent events of divergence
could be severely overestimated. Recently, experimental evidence
showed that the use of rates extrapolated from deep phylogenies
may lead to overestimation in molecularly-based timing of Pleisto-
cene events (Saarma et al., 2007; Gratton et al., 2008). Indirect evi-
dence of this bias is provided here by some discrepancies between
time estimates and the paleogeographic implications of the model.
As the Pleistocene bathymetry of the Mediterranean was similar to
that of today, variations in sea levels were produced by the glacio-
eustatic processes of the World Ocean (Emig and Geistdoerfer,
2004). In fact, during the entire Pleistocene, drops in the sea levels
were cyclically interspersed by rises. Pronounced increases (about
20 m a.s.l.) occurred three times, in the last 450,000 years (Kukla,
2005; see Fig. 4b), enough to submerge Pianosa island, where Hel-
leria occurs at present.
A further complication, not easy to reconcile with the hypothe-
sis of a glacial colonization, arises from the current distribution of
Helleria. In fact, if Helleria was able to expand its area during a gla-
cial marine regression, it is very difﬁcult to explain why the occur-
rence of Helleria in continental France and Italy remained restrict
to point locations, although the Holocene reforestation (Tinner
and Kaltenrieder, 2005) formed suitable habitats along all Italian
peninsula and in large part of continental France.
Based on the previous considerations we suggest, as a possible
working hypothesis, an alternative scenario that implies historic
human-mediated transportation to explain the presence of Helleria
in the French and Italian mainland and in the Tuscan archipelago.
In fact, during the VI and V century B.C. a documented increase in
commercial trades occurred between Corsica, Tuscany (including
islands), and southern France (where the Greek colony Massalía,
the modern Marseille, was expanding). In this period, Etruscans ex-
panded and intensiﬁed their marine commercial trades to the
northern Tyrrhenian area of the Mediterranean Sea. This was a
consequence of (i) the substantial control of Corsica, gained after
winning, in alliance with Carthago, the battle at Alalíe (the modern
Aleria, Corsica) against the Greeks Phocei (540–535 B.C.); (ii) the
dramatic naval debacle at Cuma (near Naples, southern Italy; 474
B.C.) against the Greek colony Syracuse, which terminated the
expansion of the Etruscan political inﬂuence (and piracy) in the
southern Italy. In this period, Corsica island was for Etruscans a pri-
mary source of timber, very expensive at those times yet funda-
mental for urban and architectonic use, naval purposes and as
combustible to operate metal melting furnaces (Gras, 1985).
In conclusion, the current distribution of this species in not due
to Oligocene–Miocene paleogeographic events, but rather to more
recent processes of diversiﬁcation datable, at the latest, to Early
Pliocene. Thus, the causes that determined the occurrence of Helle-
ria in Corsica and Sardinia remain obscure and, most likely, will
never be uncovered. The colonization of the French and Italian
mainland and of the Tuscan archipelago occurred very recently,
but it is still unclear if such colonization was driven primarily by
climatic changes of the Late Pleistocene or by human-transport
in historic times.
We are grateful to A. Caccone for her support in this research
and to T. Kotsakis for providing valuable and illuminating advice
in discussing Pleistocene scenarios. We also thank G. Cifani for dis-
cussing Etruscan commercial routes and I.M. Grimaldi for critically
reviewing Helleria literature. Thanks to M. Oliverio for reading an
early version of this paper. We wish to thank an anonymous re-
viewer for suggesting to investigate the occurrence of Wolbachia
and its possible effect on our phylogeographic analysis.
This work was supported by a post-doc fellowship from the
University ‘‘La Sapienza” (1997) to G. Gentile, the grant MIUR
2005 (Calibrating molecular clocks by means of vicariant specia-
tion events and paleogeographical dating) to V. Sbordoni, and the
Faculty grant 2005 (University ‘‘La Sapienza”) to R. Argano.
Alvarez, W., 1972. Rotation of the Corsica–Sardinia microplate. Nature 235, 103–
Bayes, T., 1763. Studies in the history of probability and statistics: IX. Thomas
Bayes’s essay towards solving a problem in the doctrine of chances. Biometrika
Bellon, H., Coulon, C., Edel, J.B., 1977. Le deplacement de la Sardigne: synthese des
donnees geochronologiques, magmatiques et paleomagnetiques. Bull. Soc. Geol.
France 7, 825–831.
Boccaletti, M., Ciaranﬁ, N., Cosentino, D., et al., 1990. Palinspastic restoration and
paleogeographic reconstruction of the peri-Tyrrhenian area during the
Neogene. Palaeogeogr. Palaeocl. 77, 41–50.
Bonin, B., Chotin, P., Giret, A., Orsini, J.B., 1979. Etude du bloc corsosarde sur
documents satellites: le probleme des mouvements differentiels entre les deux
iles. Rev. Geogr. Phys. Geol. Dynam. 21, 147–154.
Caccone, A., Milinkovitch, M., Sbordoni, V., Powell, J.R., 1997. Phylogeny,
biogeography, and molecular rates in European newts (genera Euproctus and
Triturus), inferred from mitochondrial DNA sequences. Syst. Biol. 46, 126–144.
Caccone, A., Sbordoni, V., 2001. Molecular biogeography, evolutionary rates, and
morphological adaptation to cave life: a case study using Bathysciine beetles
and sequence data from the mitochondrial COI gene. Evolution 55, 122–130.
Clauzon, G., Rubino, J.L., Casero, P., 2001. Regional modalities of the Messinian
salinity crisis in the framework of a two phases model. Late Miocene to Early
Pliocene environments and ecosystems. In: 2nd EEDEN Workshop. Sabadell,
Spain, pp. 17–18.
Clement, M., Posada, D., Crandall, K.A., 2000. TCS: a computer program to estimate
gene genealogies. Mol. Ecol. 9, 1657–1660.
Cunningham, C.W., 1997. Can three incongruence tests predict when data should be
combined? Mol. Biol. Evol. 14, 733–740.
Dapporto, L., Wolf, H., Strumia, F., 2007. Recent geography determines the
distribution of some ﬂying Hymenoptera in the Tuscan Archipelago. J. Zool.
Doyle, J.J., Doyle, J.L., 1987. A rapid DNA isolation procedure for small quantities of
fresh leaf tissue. Phytochem. Bull. 19, 11–15.
Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics
and dating with conﬁdence. PLoS Biol. 4, e88.
Ebner, V.v., 1868. Helleria, eine neue Isopodengattung aus der Familie der
Oniscoiden. Verhandlungen der zoologischbotanischen Gesellschaft Wien 18,
Edel, J.B., Dubois, D., Marchant, R., Hernandez, J., Cosca, M., 2001. La rotation
miocene inferieur du bloc corso-sarde; nouvelles contraintes paleomagnetiques
sur la ﬁn du mouvement. Bull. Soc. Geol. France 172, 275–283.
Emig, C.C., Geistdoerfer, P., 2004. The Mediterranean deep-sea fauna: historical
evolution, bathymetric variations and geographical changes. Carnets de
Géologie/Notebooks on Geology, Article 2004/01 (CG2004_A01_CCE-PG).
EPICA community members, 2004. Eight glacial cycles from an Antarctic ice core.
Nature 429, 623–628.
Farris, J.S., 1970. Methods for computing Wagner trees. Syst. Zool. 18, 374–385.
Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1994. Testing signiﬁcance of
incongruence. Cladistics 10, 315–319.
Fattorini, S., 2009. Both Recent and Pleistocene geography determine animal
distributional patterns in the Tuscan Archipelago. J. Zool., 1–11.
G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646 645
Felsenstein, J., 1981. Evolutionary trees from DNA sequences: a maximum
likelihood approach. J. Mol. Evol. 17, 368–376.
Felsenstein, J., 1985. Conﬁdence limits on phylogenies: an approach using the
bootstrap. Evolution 39, 783–791.
Fu, Y.X., 1997. Statistical tests of neutrality of mutations against population growth,
hitchhiking and background selection. Genetics 147, 915–925.
Gras, M., 1985. Traﬁcs tyrrhéniens archaïques, Bibliothèque des Écoles francaises
d’Athènes et Rome, Rome.
Gratton, P., Konopinski, M.K., Sbordoni, V., 2008. Pleistocene evolutionary history of
the Clouded Apollo (Parnassius mnemosyne): genetic signatures of climate
cycles and a ‘time-dependent’ mitochondrial substitution rate. Mol. Ecol. 17,
Grill, A., Casula, P., Lecis, R., Menken, S., 2007. Endemism in Sardinia. In: Weiss, S.,
Ferrand, N. (Eds.), Phylogeography of Southern European Refugia. Springer,
Netherlands, pp. 273–296.
Ho, S.Y.W., Phillips, M.J., Cooper, A., Drummond, A.J., 2005. Time dependency of
molecular rate estimates and systematic overestimation of recent divergence
times. Mol. Biol. Evol. 22, 1561–1568.
Hughes, P.D., Woodward, J.C., Gibbard, P.L., 2006. Quaternary glacial history of the
Mediterranean mountains. Prog. Phys. Geogr. 30, 334–364.
Hurst, G.D.D., Jiggins, F.M., 2005. Problems with mitochondrial DNA as a marker in
population, phylogeographic and phylogenetic studies: the effects of inherited
symbionts. Proc. R. Soc. B 272, 1525–1534.
Hurvich, C.M., Tsai, C.L., 1989. Regression and time series model selection in small
samples. Biometrika 76, 297–307.
Jobb, G., 2004. TREEFINDER: a powerful graphical analysis environment for
molecular phylogenetics. BMC Evol. Biol. 4 (18).
Jobb, G., 2008. TREEFINDER version of April 2008. Munich, Germany. Distributed by
the author at: <www.treeﬁnder.de>.
Ketmaier, V., Argano, R., Caccone, A., 2003. Phylogeography and molecular rates of
subterranean aquatic Stenasellid Isopods with a peri-Tyrrhenian distribution.
Mol. Ecol. 12, 547–555.
Kukla, G., 2005. Saalian supercycle, Mindel/Riss interglacial and Milankovitch’s
dating. Quatern. Sci. Rev. 24, 1573.
Masini, F., Bonﬁglio, L., Abbazzi, L., Delﬁno, M., Fanfani, F., Ferretti, M., Kotsakis, T.,
Petruso, D., Marra, A.C., Torre, D., 2002. Vertebrate assemblages of the central-
western Mediterranean islands during the Pliocene and Quaternary: reﬂecting
on extinction events. In: Vth Deia International Conference of Prehistory —
World Island in Prehistory: 2001. British Archaeological Research Reports,
International Series No. 1095, pp. 437–444.
Meulenkamp, J.E., Sissingh, W., 2003. Tertiary palaeogeography and
tectonostratigraphy evolution of the Northern and Southern Peri-Tethys
platforms and the intermediate domains of the African-Eurasian convergent
plate boundary zone. Palaeogeogr. Palaeocl. 196, 209–228.
Michel-Salzat, A., Bouchon, D., 2000. Phylogenetic analysis of mitochondrial LSU
rRNA in oniscids. CR Acad. Sci. Ser. III Sci. 323, 827–837.
Omodeo, P., Rota, E., 2008. Earthworm diversity and land evolution in three
Mediterranean districts. Proc. Calif. Acad. Sci. 59 (Suppl. I), 65–83.
Pond, S.L.K., Frost, S.D.W., Muse, S.V., 2005. HyPhy: hypothesis testing using
phylogenies. Bioinformatics 21, 676–679.
Popov, S.V., Shcherba, I.G., Ilyina, L.B., et al., 2006. Late Miocene to Pliocene
palaeogeography of the Paratethys and its relation to the Mediterranean.
Palaeogeogr. Palaeocl. 238, 91–106.
Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution.
Bioinformatics 14, 817–818.
Ronquist, F., Huelsenbeck, J.P., 2003. MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, 1752–1754.
Rozas, J., Sánchez-DelBarrio, J.C., Messeguer, X., Rozas, R., 2003. DnaSP, DNA
polymorphism analyses by the coalescent and other methods. Bioinformatics
Saarma, U., Ho, S.Y.W., Pybus, O.G., et al., 2007. Mitogenetic structure of brown
bears (Ursus arctos L.) in northeastern Europe and a new time frame for the
formation of European brown bear lineages. Mol. Ecol. 16, 401–413.
Suc, J.-P., 1984. Origin and evolution of the Mediterranean vegetation and climate in
Europe. Nature 307, 429–432.
Swofford, D.L., 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods). Version 4. Sinauer Associates, Sunderland, MA.
Tajima, F., 1989. Statistical method for testing the neutral mutation hypothesis by
DNA polymorphism. Genetics 123, 585–595.
Templeton, A.R., Crandall, K.A., Sing, C.F., 1992. A cladistic analysis of phenotypic
associations with haplotypes inferred from restriction endonuclease mapping
and DNA sequence data. III. Cladogram estimation. Genetics 132, 619–633.
Thompson, J.D., Gibson, T.J., Plewniak, F., et al., 1997. The CLUSTAL_X windows
interface: ﬂexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res. 25 (24), 4876–4882.
Tinner, W., Kaltenrieder, P., 2005. Rapid responses of high-mountain vegetation
to early Holocene environmental changes in the Swiss Alps. J. Ecol. 93, 936–
Tzedakis, P.C., 2007. Seven ambiguities in the Mediterranean palaeoenvironmental
narrative. Quatern. Sci. Rev. 26, 2042–2066.
Vandel, A., 1960. Faune de France — Isopodes Terrestres. P. Lechevalier, Paris.
Zhou, W., Rousset, F., O’Neill, S., 1998. Phylogeny and PCR-based classiﬁcation of
Wolbachia strains using wsp gene sequences. Proc. R. Soc. Lond. B 265, 509–
646 G. Gentile et al. / Molecular Phylogenetics and Evolution 54 (2010) 640–646