... Schachbretter ...

From Africa to Europe and back: refugia and

range shifts cause high genetic differentiation in
the Marbled White butterfly Melanargia galathea
Habel et al.
Habel et al. BMC Evolutionary Biology 2011, 11:215
http://www.biomedcentral.com/1471-2148/11/215 (21 July 2011)
Habel et al. BMC Evolutionary Biology 2011, 11:215
http://www.biomedcentral.com/1471-2148/11/215

RESEARCH ARTICLE Open Access

From Africa to Europe and back: refugia and

range shifts cause high genetic differentiation in
the Marbled White butterfly Melanargia galathea
Jan C Habel1*, Luc Lens2, Dennis Rödder3,4 and Thomas Schmitt3

Abstract
Background: The glacial-interglacial oscillations caused severe range modifications of biota. Thermophilic species
became extinct in the North and survived in southern retreats, e.g. the Mediterranean Basin. These repeated
extinction and (re)colonisation events led to long-term isolation and intermixing of populations and thus resulted
in strong genetic imprints in many European species therefore being composed of several genetic lineages. To
better understand these cycles of repeated expansion and retraction, we selected the Marbled White butterfly
Melanargia galathea. Fourty-one populations scattered over Europe and the Maghreb and one population of the
sibling taxon M. lachesis were analysed using allozyme electrophoresis.
Results: We obtained seven distinct lineages applying neighbour joining and STRUCTURE analyses: (i) Morocco, (ii)
Tunisia, (iii) Sicily, (iv) Italy and southern France, (v) eastern Balkans extending to Central Europe, (vi) western
Balkans with western Carpathian Basin as well as (vii) south-western Alps. The hierarchy of these splits is well
matching the chronology of glacial and interglacial cycles since the Günz ice age starting with an initial split
between the galathea group in North Africa and the lachesis group in Iberia. These genetic structures were
compared with past distribution patterns during the last glacial stage calculated with distribution models.
Conclusions: Both methods suggest climatically suitable areas in the Maghreb and the southern European
peninsulas with distinct refugia during the last glacial period and underpin strong range expansions to the North
during the Postglacial. However, the allozyme patterns reveal biogeographical structures not detected by
distribution modelling as two distinct refugia in the Maghreb, two or more distinct refugia at the Balkans and a
close link between the eastern Maghreb and Sicily. Furthermore, the genetically highly diverse western Maghreb
might have acted as source or speciation centre of this taxon, while the eastern, genetically impoverished Maghreb
population might result from a relatively recent recolonisation from Europe via Sicily.
Keywords: climatic oscillations, barriers, phylogeography, Melanargia galathea, Melanargia lachesis, allozyme elec-
trophoresis, climate envelope modelling

Background these taxa exclusively survived glacial periods south of the

The impacts of climatic oscillations on the earth’s biota European high mountain chains in the Iberian, Italian and
have been intensively studied [1]. In the western Palaearc- Balkan peninsulas, and some even in additional extra-
tic, thermophilic organisms went extinct over major parts Mediterranean refugia [7,8]. The long-term isolation of
of Central and North Europe during cold stages and sur- populations in these retreats over many thousands of years
vived in the lowlands of lower latitudes in often distinct resulted in genetic differentiation [5]. During the warmer
refugia [2-6]. Molecular studies revealed that most of interglacial periods, species expanded their distribution
ranges northwards and extended their different genetic
lineages over more northern areas [9,10].
* Correspondence: [email protected] In contrast to the three more intensively studied Medi-
1
Musée national d’histoire naturelle Luxembourg, L-2160 Luxembourg,
Luxembourg terranean refugia of southern Europe (Iberia, peninsular
Full list of author information is available at the end of the article

© 2011 Habel et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 2 of 14
http://www.biomedcentral.com/1471-2148/11/215

Italy and the Balkans), little is known about North African (iii) Which routes of expansion and retraction fol-
refugia and the biogeographical relation between the lowed the butterfly throughout time?
Maghreb and southern Europe separated by the two
narrow sea straits of Gibraltar and Sicily. It has been Results
shown that the Maghreb is often sub-structured following Phylogeographic analyses
an east-west [e.g. [11-13]] or south-north differentiation All enzyme loci had banding patterns consistent with
pattern [e.g. [5,14]]; in some cases, genetic continuity was known quaternary structures. While most loci were inher-
demonstrated between the Maghreb and Sicily [e.g. ited autosomally, 6PGDH and ME were located on the Z
[15,16]]. Other studies underline the important role of chromosome so that hemizygous females (but not males)
Sicily as diversification centre for European taxa unravel- had a single copy [27]. No general linkage disequilibrium
ling deep genetic splits between this island and peninsular was observed for any locus (all p > 0.05 after Bonferoni
Italy (e.g. Erinaceus europaeus: [17]; Pseudepidalea viridis: correction). A total of 13 analysed loci were polymorphic,
[18,19]). Few molecular analyses also reveal the outstand- but two loci (FUM, GPDH) were monomorphic through-
ing importance of North Africa as a refugium for thermo- out all samples. Allele frequencies for each enzyme and
philic species during glacial periods [e.g. [12,14,20,21]]. population are given in an additional file 1.
However, most studies focus either on the Maghreb or the When calculating parameters of genetic diversity, all
southern European refugia and do not combine the distri- 15 loci were used. The genetic diversities of populations
bution of species all over north-western Africa and showed strong variability among populations of different
throughout Europe. regions, and standard deviations were high if compared
To study the biogeographical importance of the Maghreb against means (ratio standard deviation against means:
region and its connection with Europe, we selected the A 9.0%; H e 12.9%; H o 18.3%; P tot 18.1%; P 95 15.6%).
Marbled White butterfly species complex Melanargia Values for all populations analysed are given in Table 1,
galathea (Linnaeus, 1758) and Melanargia lachesis (Hüb- overall means in Table 2.
ner, 1790) as a model system using two analytical tools A neighbour joining phenogram based on allele fre-
(allozyme polymorphisms and distribution modelling). quencies (Figure 1) showed a first split between M.
Today, M. galathea is widely distributed from the Maghreb galathea and M. lachesis with a genetic distance [28] of
region (mountain ranges of Morocco, Algeria and Tunisia) about 0.9. The second split between M. galathea popula-
[22,23] to the English Midlands [24], and from the Pyre- tions from Tunisia and Sicily on the one hand and all
nees [25] to the Baltic Sea in Poland [26]. On the Iberian remaining M. galathea populations on the other was in
Peninsula, M. galathea is replaced by its sibling species M. average about half of the genetic distance of the first
lachesis. Thus, the Italian peninsula is the only possible split. The outgroup M. lachesis routing the tree also sup-
link between North Africa and Europe for M. galathea. ports this split being the first one in M. galathea. The
Previous molecular studies on these butterflies based Tunisia - Sicily group showed a further genetic differen-
on allozyme polymorphisms supported the sibling species tiation between these two geographic regions. All these
status of both taxa and revealed two genetic groups in splits are supported by bootstrap values. The remaining
M. galathea, one western and one eastern group, indicat- populations split into five groups, the populations from
ing an Italian and a Balkan refugium [27], with further Morocco and four European groups: (i) mainland Italy
substructures in the Balkan region [28]. Preliminary data and southern France (Condat), (ii) western Balkans and
for the western Maghreb showed the highest known western Carpathian Basin, (iii) eastern Balkans, Romania
values of genetic variability in this region [29]. Based on and Central Europe, as well as (iv) south-western Alps
these data, we hypothesized a Maghreb origin of the spe- (Col de Tende). The latter group is the only one well sup-
cies and colonisation of Europe via Sicily and Italy. How- ported by bootstrapping. Unexpectedly, the populations
ever, these previous studies lack populations from the from Morocco are not well distinguishable in this tree
eastern Maghreb, Sicily, Italy and southern France. In from the western Balkan group. Three populations are
this article we combine an allozyme data set covering not matching any of these groups: the north-eastern
most of the recent distribution of the species with climate French population Lorry and the southern German
envelope models to test the previously postulated biogeo- population Bossler are intermediate between groups iii
graphical scenario of refugia and barriers during the last and iv, and the southern Calabrian population St. Giorgio
ice ages until today, addressing the following questions: shows some traits of the Sicily-Tunisia group thus not
(i) Which refugia are of importance for the glacial sur- clustering together with the other populations from
vival of the M. galathea /lachesis species complex dur- mainland Italy. These three populations are thought to
ing the subsequent glacial periods? be of hybrid origin between the respective genetic groups.
(ii) Is there any evidence of genetic structuring within STRUCTURE plots (for K = 2 to K = 8) support the
the North African and Italian refugia? topology of the neighbour joining phenogram (Figure 2);
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 3 of 14
http://www.biomedcentral.com/1471-2148/11/215

Table 1 Sampling location and five parameters of genetic diversity for 41 populations of Melanargia galathe a from
its western Palaearctic distribution area and one population of M. lachesis from the Pyrenees: number of individuals
analysed (N), mean number of alleles per locus (A), percentage of expected (He) and observed (Ho) heterozygosity,
percentage of polymorphic loci not exceeding 95% (P95) and total number of polymorphic loci (Ptot)
Region Location Running Nr. Long. (N) Date of sampling N A He (%) Ho (%) P95 (%) Ptot (%)
Lat. (E/W)
Morocco M-Oukaimeden 1 31.12; 7.52W 23-V-05 36 2.33 20.5 20.6 46.7 46.7
M-Naour 2 32.29; 5.56W 26-V-05 36 2.33 17.0 13.5 33.3 53.3
M-Bekrite 3 33.05; 5.13W 29-V-05 36 2.53 17.2 18.5 33.3 53.3
M-Timhadite 4 33.14; 5.03W 29-V-05 36 2.33 18.7 17.7 33.3 60.0
Tunisia T-Ain Draham 5 36.46; 8.42E 23-V-10 27 2.20 17.6 15.6 40.0 53.3
T-Nebeur 6 36.18; 8.46E 26-V-10 40 2.00 12.3 9.9 26.7 60.0
T-Table de Yagurta 7 35.45; 8.23E 25-V-10 40 2.13 13.6 10.6 40.0 53.3
T-Thala 8 35.34; 8.41E 25-V-10 40 1.80 12.5 12.1 26.7 40.0
T-Béja 9 36.44; 8.54E 28-V-10 40 1.86 13.6 9.8 33.3 40.0
Sicily I-Valledomo 10 37.45; 13.53E 4-VII-07 40 2.13 14.5 13.2 40.0 53.3
I-Francavilla 11 37.53; 15.07E 5-VII-07 30 2.26 17.8 14.8 26.7 66.7
I-Reitano 12 37.58; 14.20E 26-VI-07 40 2.00 15.0 12.2 33.3 46.7
Italy I-St. Giorgio 13 38.17; 15.59E 6-VII-07 40 2.13 17.1 17.2 33.3 40.0
I-Mormanno 14 39.54; 15.58E 7-VII-07 40 2.66 21.3 15.6 53.3 73.3
I-Napoli 15 41.08; 14.19E 7-VII-07 20 2.13 18.2 13.3 40.0 46.7
I-Rieti 16 42.25; 12.53E 8-VII-07 40 2.20 17.8 17.4 33.3 53.3
I-Consuma 17 43.48; 11.36E 8-VII-07 40 2.20 19.2 17.5 40.0 60.0
I-Verona 18 45.31; 10.55E 9-VII-07 40 2.13 17.7 13.5 40.0 46.7
France F-Col de Tende 19 44.09;7.34E 31-VII-03 40 2.13 17.7 12.1 33.3 40.0
F-Condat 20 45.20; 2.45E 11-VII-08 27 1.86 13.8 10.6 26.7 46.7
F-Lorry 21 49.00; 6.06E 17-VI-03 40 2.06 17.6 15.3 33.3 40.0
Central Europe L-Niederanven 22 49.39; 6.16E 13-VI-03 32 2.00 18.4 17.1 33.3 40.0
D-Niederehe 23 50.18; 6.48E 18-VII-03 18 1.86 18.2 16.7 33.3 33.3
D-Bossler 24 48.36; 9.36E VII-04 40 2.00 18.5 17.0 40.0 40.0
D-Buchenberg 25 50.43; 11.40E VII-03 40 2.33 19.3 16.4 33.3 40.0
A-Jadersdorf 26 46.40; 13.18E 18-VIII-04 40 2.00 13.5 12.6 33.3 33.3
A-Hochobir 27 46.30; 14.30E 23-VIII-04 40 2.06 14.6 13.3 26.7 46.7
A-Schöckl 28 47.12; 15.28E 22-VIII-04 40 2.13 17.0 15.7 33.3 40.0
A-Leithagebirge 29 47.57; 16.39E 31-VII-03 14 1.86 14.8 13.6 33.3 40.0
Hungary H-Csákvár 30 47.24; 18.26E 13-VII-04 40 2.33 17.2 15.9 33.3 46.7
western Balkans SLO-Postojna 31 45.47; 14.12E 15-VII-04 40 2.00 14.4 13.5 33.3 40.0
MNE-Durmitor 32 43.10; 19.08E 10-VII-03 12 1.86 14.8 14.5 26.7 33.3
Romania RO-Hoteni 33 47.38; 24.02E 25-VII-04 40 1.93 17.9 20.5 33.3 33.3
RO-Cluj 34 46.45; 23.35E 23-VII-04 40 2.13 18.3 17.7 40.0 40.0
RO-Voslobeni 35 46.40; 25.38E 29-VII-04 40 1.86 18.5 18.6 33.3 40.0
RO-Pasul Predelus 36 45.35; 26.09E 31-VII-04 40 2.33 20.4 18.8 33.3 40.0
RO-Porta di Fier Transilvanici 37 45.31; 22.39E 15-VII-04 40 2.06 19.1 19.2 33.3 40.0
RO-Inelet 38 44.58; 22.29E 9-VIII-04 19 1.80 19.2 16.0 33.3 40.0
Bulgaria BG-Milanovo 39 43.05; 23.23E 10-VIII-04 40 2.06 17.8 17.3 33.3 40.0
BG-Karandila 40 42.43; 26.20E 3-VIII-04 40 2.06 18.4 17.1 33.3 40.0
BG-Trigrad 41 41.36; 24.17E 7-VIII-04 40 2.06 16.5 16.5 33.3 33.3
lachesis E-Col de Perbes* 42 42.23; 1.13E 19-VII-03 40 2.26 20.8 17.6 40.0 40.0
* M. lachesis
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 4 of 14
http://www.biomedcentral.com/1471-2148/11/215

Table 2 Means of sample sizes and genetic diversities of the different genetic groups of Melanargia galathea and
M. lachesis; p values of Kruskal Wallis ANOVAs among groups are given
N A He Ho P95 Ptot
All 36.0 ± 7.9 2.10 ± 0.19 17.0 ± 2.2 15.3 ± 2.8 34.5 ± 5.4 45.2 ± 9.4
Morocco 36.0 ± 0.0 2.38 ± 0.10 18.4 ± 1.6 17.6 ± 3.0 36.7 ± 6.7 53.3 ± 5.4
Tunisia 37.4 ± 7.8 2.00 ± 0.17 13.9 ± 2.1 11.6 ± 2.4 33.3 ± 6.7 49.3 ± 8.9
Sicily 37.7 ± 5.8 2.13 ± 0.13 15.8 ± 1.8 13.4 ± 1.3 33.3 ± 6.6 55.6 ± 10.2
Italy + SE France 35.3 ± 8.3 2.19 ± 0.24 17.9 ± 2.3 15.1 ± 2.6 38.1 ± 8.3 52.4 ± 11.2
SW Alps 40 2.13 17.7 12.1 33.3 40.0
eastern Balkans + Central Europe 36.8 ± 7.4 2.04 ± 0.15 17.8 ± 1.7 16.9 ± 2.0 33.7 ± 2.9 38.7 ± 3.6
western Balkans 26.5 ± 15.6 2.01 ± 0.22 15.3 ± 1.3 14.4 ± 1.1 31.6 ± 3.3 40.0 ± 5.5
p (Kruskal Wallis ANOVA) 0.33 0.044 0.019 0.004 0.65 < 0.001
eastern Balkans + Romania 35.7 ± 8.4 2.05 ± 0.15 17.1 ± 2.2 15.5 ± 1.8 33.3 ± 3.8 39.0 ± 4.6
Central Europe 37.7 ± 7.0 2.03 ± 0.16 18.5 ± 1.1 18.0 ± 1.4 34.0 ± 2.2 38.5 ± 2.9
p (Kruskal Wallis ANOVA) 0.56 > 0.99 0.31 0.011 0.75 0.87
M. lachesis 40 2.26 20.8 17.6 40.0 40.0

the sequence of splits of STRUCTURE groups is mostly Giorgio (13) is grouped with all other populations from
reflecting the genetic distances in the tree. Additionally, Italy, Bossler (24) with all populations from Central and
the samples from Morocco are consistently separated as south-eastern Europe and Lorry (21) is part of the
one group from all other M. galathea populations from south-western Alps group.
K = 5 onwards. However, the STRUCTURE analysis is The overall differentiation among all populations includ-
not able to distinguish the western Balkan group from ing M. lachesis was strong (FST: 0.179, p < 0.001); exclud-
the eastern Balkans, Romania and Central Europe ing this outgroup population only decreased this value
group. Furthermore, STRUCTURE did not reflect the marginally (FST: 0.169, p < 0.001) (Table 3). Hierarchical
hybrid origin of three populations mentioned above: St. variance analyses well supported the hierarchical

M-Naour

M-Oukaimeden

te
di 0.1
ha r
A-Jadersdorf im vá th a itor Nei (1972)
-T k i ur m
A-Hochobir M Csá A-LeMN-D
- SLO-Postojna T-Thala
RO-Voslobeni H 64.1 T-Table de Yahurta
T-Beja
D-
Ni

70.2 M-Bekrite
e

D-Bossler T-Ain Draham

de

RO-Cluj 57.8
re

F-Lorry T-Nebeur
he

I-Napoli I-Pazza I-St. Giorgio

66.2 66.9
D-Buchenberg
F-Col de A-Schöckl 64.0 I-Si-Valledomo
Tende BG-Trigrad nov o 60.2
53.6 I-Si-Francavilla
ila F-Condat
L -N

BG-Mdila t
ns ilva lu s

I-Rieti I-Si-Reitano
nic i

an ele
de

r
ie d

K a I-Consuma
BG- - In
re

RO

e ra
-P

RO I-Mormanno
RO

-H o

nv
F Tra

en
ten

0.9 Nei (1972)

i
R O-P

M. lachesis
Figure 1 Neighbour-joining dendrogram of Melanargia galathea and M. lachesis. The phenogram is based on the genetic distances (Nei,
1972) [56] of 41 populations of M. galathea and one population of M. lachesis.
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 5 of 14
http://www.biomedcentral.com/1471-2148/11/215

lachesis
Morocco Tunisia Sicily Italy France Central and SE Europe
2

Figure 2 Individual based genetic classification of Melanargia galathea and M. lachesis. Bayesian analysis of all M. galathea and M. lachesis
populations performed using the STRUCTURE software [61]. Analyses for K = 2 to K = 8 are depicted.

Table 3 Analyses of molecular variance for all Melanargia galathea populations and one population of M.lachesis
Group among populations among individuals within populations within individuals
(FST) (FIS)
All galathea and lachesis 0.1789*** 0.0875*** (0.7910)
(0.1889) (0.0758)
All galathea 0.1685*** 0.0858*** (0.7968)
(0.1767) (0.0748)
Sicily and Tunisia 0.0623*** 0.1111*** (0.7059)
(0.0527) (0.0882)
Sicily 0.0476*** 0.0827* (0.6325)
(0.0345) (0.0570)
Tunisia 0.0425*** 0.1252*** (0.7513)
(0.0381) (0.1075)
Morocco 0.0334*** 0.0408 (0.8427)
(0.0304) (0.0358)
Italy, Balkans, Central Europe 0.0951*** 0.0853*** (0.8174)
(0.0939) (0.0763)
Italy 0.0116 0.1092*** (0.8438)
(0.0112) (0.1034)
Italy and S France 0.0142* 0.1143*** (0.8258)
(0.0135) (0.1066)
west Balkan group 0.0500*** 0.0999** (0.7028)
(0.0411) (0.0780)
east Balkan and Central Europe 0.0614*** 0.0673*** (0.8282)
(0.0580) (0.0597)
F statistics (top line) with their respective variance values (in parenthesis below). Groupings following neighbour joining analysis and STRUCTURE plots (see
Figures 1 and 2).
*: p < 0.05; **: p < 0.01; ***: p < 0.001
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 6 of 14
http://www.biomedcentral.com/1471-2148/11/215

structures of the neighbour joining phenogram and of On average, the ‘temperature annual range’ had the
STRUCTURE analyses (Table 4). Thus, these analyses highest explanatory power (30.3%), followed by the
strongly support (i) the genetic break in the Maghreb, (ii) ‘minimum temperature of the coldest month’ (16.8%),
the break between Sicily and mainland Italy, (iii) the differ- the ‘precipitation of the warmest quarter’ (14.8%), the
entiation into four genetic lineages in continental Europe, ‘annual precipitation’ (10.2%), the ‘maximum tempera-
(iv) the cohesiveness between Sicily and Tunisia, (v) the ture of the warmest month’ (8.2%) and the ‘precipitation
lack of differentiation from the eastern Balkans via Roma- of the driest quarter’ (7.2%). All other variables contrib-
nia to Central Europe and (vi) the strong genetic similarity uted less than 5% each. The average minimum training
between mainland Italy and southern France. The genetic presence was 0.05, and the lowest 10 percentile training
differentiation within the seven groups at the lowest hier- omission threshold was 0.36.
archical level was low to moderate ranging from FST values The current potential distribution suggested by the
of 0.0142 to 0.0614 (Table 3). SDM is highly coincident with the butterfly’s recent
The genetic diversities among these genetic lineages range. The recent climatic niche over North Africa is
showed significant differences (Table 2). Thus, the Mor- displayed as two separate areas, in the West and East.
occo group showed the highest values achieved for A, He Under palaeoclimatic conditions assumed to have pre-
as well as Ho, and the means for P95 and Ptot were above vailed 21,000 y BP (CCSM scenario), the potential distri-
average. On the other extreme, Tunisia had the lowest bution may have been much more restricted in Europe:
means for A, He and Ho, and the mean for P95 was well Major parts of Central Europe changed into climatically
below average; the genetic diversities of Tunisia were lower unsuitable areas for M. galathea during the glacial per-
(A, He, Ho, Ptot) or equal than in the otherwise rather simi- iod, while the southern European peninsulas (Iberia,
lar populations from Sicily. The four groups from mainland Italy and the Balkan) retained suitable climatic condi-
Europe all have mostly intermediate genetic diversities tions. Major parts of the Maghreb had a suitable climate
scatted around the respective mean values. for the butterfly being geographically more extended
than today (Figure 3).
Species distribution modelling
According to the classification of Swets [30], we Discussion
received ‘excellent’ AUC values in our 100 models (aver- The obtained allozyme data displayed in neighbour-join-
age training AUC = 0.927, average test AUC = 0.902). ing phenograms, structure plots and hierarchical

Table 4 Hierarchical variance analyses of Melanargia galathea and M.lachesis among genetic groups
Groups among within prop. of among groups variance of total variance
groups groups among pops.
(FCT) (FSC)
galathea vs lachesis 0.1811*** 0.1694*** 56.6%
(0.2308) (0.1768)
Tunisia + Sicily vs rest of Europe 0.2256*** 0.0892*** 76.6%
(0.2786) (0.0852)
Tunisia vs Morocco 0.2153*** 0.0387*** 87.6%
(0.2475) (0.0349)
Tunisia vs Sicily 0.0338 0.0445*** 44.0%
(0.0291) (0.0370)
Sicily vs Italy 0.1797*** 0.0203*** 91.5%
(0.1928) (0.0179)
Italy vs S France (Condat) 0.0107*** 0.0119*** 47.4%
(0.0102) (0.0113)
Italy vs SW Alps (Col de Tende) 0.1726*** 0.0117*** 94.7%
(0.1993) (0.0112)
east Balkans + Central Europe vs SW Alps (Col de Tende) 0.0615* 0.0611*** 51.8%
(0.0621) (0.0579)
Italy vs SW Alps vs west Balkans vs east Balkans + 0.0798*** 0.0479*** 64.4%
Central Europe (0.0813) (0.0449)
east Balkans + Romania vs Central Europe 0.0066 0.0581*** 10.3%
(0.0063) (0.0548)
Morocco vs continental Europe 0.0269* 0.0896*** 23.6%
(0.0271) (0.0877)
F statistics (top line) with their respective variance values (in parenthesis below). Groupings following neighbour joining analysis and STRUCTURE plots (see
Figures 1 and 2).
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 7 of 14
http://www.biomedcentral.com/1471-2148/11/215

A
A

Maxent score High, 1 > min train presence

value Low, 0
> 10% train omission
glaciers
Figure 3 Distribution modelling of Melanargia galathea. Potential distributions of Melanargia galathea under (a) current and (b)
palaeoclimatic conditions (CCSM). Species records used for model training are indicated as white dots (3a). Dark grey areas refer to occurrence
probabilities above the average minimum training presence; black areas represent those above the 10 percentile training omission.

variance analyses indicate a profound genetic split (iii) remarkable differences in DNA sequences of the
between the two taxa, M. galathea and M. lachesis. nuclear wg gene between M. lachesis and M. galathea,
Nazari et al. [31] supported this pattern by three lines of but no major differentiation between M. galathea sam-
evidence: (i) differences of the male genitalia between M. ples from Europe and the Maghreb. However, the
lachesis and M. galathea, (ii) a stronger difference in sequences of the two mtDNA genes cox1 and 16S con-
wing patterns between these two taxa than between M. tradict the common pattern of allozymes, genital struc-
galathea population in Europe and the Maghreb and tures, wing patterns and nuclear DNA sequences: This
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 8 of 14
http://www.biomedcentral.com/1471-2148/11/215

marker is not well distinguishing M. galathea from Eur- for vicariance and thus the beginning of this differentia-
ope and M. lachesis, but shows remarkable differences tion. As (i) Iberia was continuously blocked for the
between Europe and the Maghreb with this split being expansion of M. galathea to Europe by M. lachesis [cf.
dated back to the Messinia Salinity Crises more than 5 27] and (ii) all European M. galathea populations except
My ago [31]. Having in mind the differentiation pattern Sicily are more similar to populations from Morocco
in all known marker systems, we believe that these two than from Tunisia, a scenario with this split taking place
mtDNA lineages in the entire species complex might in the Maghreb is little likely. This assumption is further
have originated at that time horizon, but were distribu- supported by SDMs for ice age conditions predicting
ted to different geographical regions only much later by mostly continuous distributions over North Africa (Fig-
lineage sorting, maybe hereby exemplifying one case of ure 3b) thus allowing vicariance in this region only dur-
the often observed difference between mtDNA on the ing the relatively short interglacial stages. For these
one hand and nuclear DNA sequences, morphological reasons, M. galathea must have reached Europe before
characteristics and allozyme pattern on the other [32]. the Riss glaciation.
Our allozyme data further show strong differentiation As the region of the eastern Sahara in Egypt appar-
within M. galathea into two major groups with respec- ently always have been too dry for an expansion of M.
tive subgroups: (i) Sicily - Tunisia with (i-a) Sicily and galathea, this first expansion of M. galathea to Europe
(i-b) Tunisia as well as (ii) all other M. galathea with must have been from Tunisia to Sicily (Figure 4a), a sea
(ii-a) Morocco, (ii-b) Italy with parts of southern France, strait known for biogeographical connections for many
(ii-c) western Balkan including the western Carpathian taxa [e.g. [15]; and references therein]. As the Strait of
Basin, (ii-d) eastern Balkans with Romania and Central Sicily was considerably narrower during glacial periods
Europe, and (ii-e) the south-western Alps. due to eustatic sea level lowering, the transition from
Mindel glaciation to Holstein interglacial with still low
Atlantic-Mediterranean origin of the M. galathea/lachesis sea level but already higher temperatures might have
species complex been a suitable time period for this dispersal. After arri-
The recent geographic restriction of M. lachesis to val to Sicily, the Holstein interglacial might have given
Iberia and the highest genetic diversity of M. galathea suitable condition for the expansion of M. galathea over
in Morocco support the idea of a centre of origin of the most parts of Europe, including the Balkans but exclud-
entire species complex in this area. This assumption is ing Iberia as this peninsula was already populated by M.
further supported by other Melanargia species mostly lachesis (Figure 4a).
endemic to the Atlantic-Mediterranean region (M. occi- With the climatic cooling of the Riss ice age, which
tanica, M. ines) and other endemics to further Mediter- was considerably longer than the following Würm gla-
ranean refugia (M. arge: peninsular Italy; M. pherusa: ciation and had longer durations of minimum tempera-
Sicily, M. larissa: Pontic-Mediterranean region and tures [33,37], M. galathea most probably was nearly
Iran). The onset of the differentiation between these sis- extinct in Europe only surviving in the southernmost
ter species should be due to vicariance events most possible retreats in Sicily and the southern Balkans
likely correlated with the onset of an ice age. If giving (Peleponnesos), but also in the Maghreb; M. lachesis
one glacial-interglacial cycle for the lowest level of dif- could survive in southern Iberia (Figure 4b). This vicar-
ferentiation (i.e. the subgroups within the two major M. iance might be the origin of the two major European
galathea lineages), the most likely time horizon of this lineages of M. galathea with the eastern one by chance
vicariance event is the onset of the Günz glaciation evolving similarly in allele frequencies as the Morocco
some 560,000 years BP [33] (Figure 4a). Since then, M. lineage, with this similarity therefore not representing
lachesis most likely has never expanded out of Iberia recent biogeographical connection between them. Riss
whereas M. galathea colonised most of Europe from its vicariance events most likely have also been responsible
Maghreb expansion centre. Similar splits between Iberia for other differentiation processes as e.g. in the Polyom-
and the Maghreb are commonly observed in many spe- matus coridon /hispana complex [e.g. [38]].
cies groups [e.g. [13,34-36]].
...and back to the Maghreb
From the Maghreb to Europe As the time for differentiation between the four M.
The deepest split in the M. galathea populations is galathea lineages from continental Europe is assumed to
between the Sicily - Tunisia group and all the other be the result of one glacial cycle (see above) and as the
populations. As this split is about twice the genetic dif- differentiation between populations from Sicily and
ferentiation among their subgroups and less than half of Tunisia are in the same order of magnitude, we assume
the distance against M. lachesis, the onset of the Riss that the onset of this differentiation is in the same time
glaciation (about 310 ky BP) [33] might be the trigger frame. As the genetic diversity is significantly higher in
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 9 of 14
http://www.biomedcentral.com/1471-2148/11/215

warm stages
4a) Günz to Holstein

cold stages
age (Ma)
2 Würm (2-4
4 5a
Eem (5a-5e)
0.1
5e
6
4b) Riss and Eem 7a
0.2 Riss (6-8)
7c
8
9a
0.3 Holstein (9a-9e)
9e
10
Mindel (10-12
0.4 11
12

4c) Würm and

Postglacial Cromer (13)
0.5
14
Günz (14)
15a

0.6
15c Waal (15a-15e)
16

Figure 4 Biogeographic scenario of Melanargia galathea and M. lachesis. Range expansions and retractions of M. galathea and M. lachesis
(on the Iberian Peninsula) during the past ice-ages (a-c) and fluctuations of marine isotope stages (d) (redrawn after Gibbard & van Kolfschoten
[33]). Refugia are marked by grey areas, expansions/retractions by arrows. 4a) Günz and Mindel refugia in Iberia (M. lachesis) and the Maghreb (M.
galathea); expansion from the eastern Maghreb to Sicily during the Mindel/Holstein transition; Holstein expansion over Europe, but retraction in
the Maghreb to the west. 4b) Riss refugia in Iberia (M. lachesis) and the western Maghreb, Sicily and southern Balkans (M. galathea; Riss/Eem
transition expansion from Sicily to the eastern Maghreb; Eem expansion of the southern Balkan group over major parts of Europe including
peninsular Italy. 4c) Würm refugia in Iberia (M. lachesis) as well as western and eastern Maghreb, Sicily, peninsular Italy, southwestern Alps and
Balkan area (M. galathea); postglacial expansion from Iberia (M. lachesis), peninsular Italy and the Balkan area (M. galathea).

Sicily than in Tunisia and the warm and dry interglacial most probably could colonise most parts of Europe
climatic conditions in Tunisia generally unsuitable for apart from Iberia and Sicily, which were occupied by
the survival of M. galathea, we assume that a colonisa- other genetic lineages of this species complex (Figure
tion most likely has taken place from Sicily to Tunisia. 4b).
While the sea level was still considerably lowered at the
transition from Riss to Eem thus facilitating dispersal The existence of extra-Mediterranean refugia for
between these two areas, this time period might be the thermophilic taxa
most likely for this expansion event. During the follow- During the Würm ice age, which was not more severe
ing Eem interglacial, the Balkan refuge of M. galathea than the two previous glaciations but with a shorter
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 10 of 14
http://www.biomedcentral.com/1471-2148/11/215

maximum [33], the Marbled White butterflies were not origin of these populations rather likely and thus expan-
that much pushed to the South than in the previous sion of the southern Alps lineage over the chains of the
cases. This is well matching the remarkable differentia- Alps.
tion of the species in Europe allowing to distinguish five Also the Italian lineage could expand beyond its refu-
lineages (see above), which most likely are the result of gium to southern France. Therefore the entire region of
survival of the Würm ice age in a larger number of dif- northern France and southern Germany might be a
ferent refugia. zone of mixing between these three lineages. Hybrid
This pattern implies at least two different refugia at the zones between different taxa are frequently observed in
Balkan Peninsula at the western and the eastern flank; this region [e.g. [9,55]]. Furthermore, the southernmost
more in detail analyses also support a third Balkan centre population in Calabria (southern Italy) has an intermedi-
in the peninsula’s southern parts [29] (Figure 4c). This ate genetic texture between the Italian and the Sicily
pattern of multiple refugia in the Balkans was already group thus speaking for a postglacial contact and inter-
erected by Reinig [39] postulating different centres of mixing between these two groups in this region.
survival in the western, southern and eastern Balkans and
was later supported by genetic analyses showing genetic Conclusion
divergences between these areas for a variety of different The hierarchical structure of our allozyme data set on
animal species [e.g. [18,40-42]]. M. galathea and M. lachesis is consistent with the chron-
Furthermore, different Würm refugia have to be pos- ology of the last four glacial-interglacial cycles. Based on
tulated for Sicily and peninsular Italy, a pattern also this consistency, we derive the following scenario, which
repeated by other genetic analyses [e.g. [17,43]]. Further- in our opinion is the most likely one: (i) The beginning of
more, other genetic studies show a remarkable genetic the Günz ice age might have affected the vicariance
differentiation in the southernmost parts of peninsular between the two species. (ii) M. galathea might have
Italy [e.g. [34,44,45]]. crossed from Tunisia to Sicily at the transition from
The last remaining lineage of M. galathea in the south- Mindel ice age to Holstein interglacial and (iii) subse-
western Alps most likely is not representing a Mediterra- quently spread all over Europe, but retreated in the
nean refuge of this species, but an extra-Mediterranean Maghreb to the higher elevations of the Atlas mountains.
refuge area at the southern slopes of the glaciated Alps (iv) The members of this species complex survived the
(Figure 4c). As already shown by Steward and Lister [46], coldest periods of the Riss glaciation only in southern
glacial survival of temperate species in Europe was not Iberia, Morocco, Sicily and the southern Balkans (Pele-
only possible in the classical Mediterranean refugia sensu ponnesos). (v) At the transition from Riss ice age to Eem
de Lattin [47], but also in small climatically buffered interglacial, Tunisia was recolonised from Sicily. (vi) The
pockets in more northern regions [8,48,49]. Recent works southern Balkan group might have colonised major parts
especially highlight the southern and south-eastern parts of Europe during the Eem interglacial including Italy and
of the Alps of particular importance for additional Würm Central Europe. (vii) Populations of this group survived
ice age refugia for temperate species [e.g. [42,50,51]], and the Würm ice age in Italy, the southern margin of the
also for species formerly thought to be of exclusive Medi- Alps, the western and eastern flank of the Balkan penin-
terranean origin [e.g. [52,53]]. This apparently was also sula; members of other lineages survived in Sicily, Tuni-
the case for the Marbled White. sia, Morocco and Iberia. (viii) During the Postglacial,
only the eastern Balkan and the Italian lineage showed
Postglacial expansion major northwards range expansion. (ix) Hybridisation
During the Postglacial, several lineages of M. galathea between lineages most probably occurred in western
were mostly blocked in their expansion by other lineages Central Europe and southern Calabria.
representing the respective leading edges [cf. 54]. In the
case of M. galathea in Morocco, their northwards expan- Methods
sion was blocked by M. lachesis distributed in Iberia. The Allozyme electrophoresis
lineage surviving in the eastern Balkans apparently had We scored banding patterns of allozyme polymorphisms
the most important impact in the recolonisation of more analysed for 1,463 M. galathea specimens from 41
northern parts of Europe as its dispersal was not ham- populations sampled across major parts of the Maghreb
pered by any major mountain obstacle [cf. 9] so that this and Europe and one populations of M. lachesis (40 indi-
lineage could expand throughout Central Europe to the viduals) from the Spanish Pyrenees (see Figure 5 and
western parts of Germany (Figure 4c). However, the sam- Table 1). In total we analysed 15 enzyme systems:
ples of north-eastern France and southern Germany 6PGDH, ACON, FUM, G6PDH, GAPDH, AAT2,
show an intermediate genetic structure between this line- GPDH, PGI, HBDH, IDH1, IDH2, MDH1, MDH2, ME,
age and the south-western Alps lineage, making hybrid and PEP. The data of 26 M. galathea populations and of
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 11 of 14
http://www.biomedcentral.com/1471-2148/11/215

23 25
22
21
24
29 33
28 30
26 27 34 35
20 18 31 37 36
38
19
17
32 39
42 16 40
15
41

14

13
10 11 12
5 9
6
7 8

3 4
2
1

Figure 5 Sampling design of Melanargia galathea and M. lachesis. Geographical location of the sampled populations of M. galathea and M.
lachesis (sample 42). Given numbers coincide with Table 1.

M. lachesis were taken from Habel et al. [27,28] and detecting differences of means of genetic diversities
Schmitt et al. [29]. Specimens were netted in the field among genetic lineages and sublineages, we calculated
from mid-May to the beginning of August between U-tests using STATISTICA. Conventional F statistics,
2004 and 2010, frozen alive in liquid nitrogen or in a AMOVAs, hierarchical genetic variance analysis, tests of
freezer and stored under these conditions until analysis. Hardy-Weinberg equilibrium and linkage disequilibrium
Standard procedures of allozyme electrophoresis were were calculated with ARLEQUIN 3.1 [58]. Phenograms
performed as described in Habel et al. [27]. using the neighbour joining algorithm [59] were con-
structed with PHYLIP [60], including bootstrap-values
Statistics (calculated based on 1,000 iterations). To define indivi-
Alleles were labelled according to their relative mobility, dual based genetic clusters we performed STRUCTURE
starting with “1” for the slowest. All laboratory results analyses [61]. As burn-in and simulation lengths we
were stored on cellulose acetate plates. These banding used 100,000 and 300,000 iterations per run based on
patterns were (re)analysed by one person (JCH). Allele the admixture model with correlated gene frequencies
frequencies, Nei’s standard genetic distances [56] and comparing different groupings (from K = 2 to K = 10).
parameters of genetic diversity (i.e. mean number of
alleles per locus, A, expected heterozygosity, H e , and Species Distribution Modelling
observed heterozygosity, Ho, total percentage of poly- Over the last few decades, Geographic Information Sys-
morphic loci, Ptot, and percentage of polymorphic loci tem (GIS) based Species Distribution Models (SDMs)
with the most common allele not exceeding 95%, P95) have become vital tools used to predict the potential
were computed with G-Stat [57]. As sample sizes do not distribution of species under current conditions and cli-
differ significantly, the calculation of allelic richness cor- mate change scenarios [62-64]. In combination with
recting for population sizes was not necessary. For palaeoclimatological data, SDMs have been suggested as
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 12 of 14
http://www.biomedcentral.com/1471-2148/11/215

a mean of inferring species’ past distributions [65,66], Acknowledgements

We acknowledge financial support by the German Academic Exchange
especially when combined with phylogeographic techni-
Service (PostDoc Programme) and the Musée national d’histoire naturelle
ques [67]. Luxembourg to JCH as well as from the Ministry of Education, Science,
We compiled a set of 3,483 species records of M. Youth and Culture of the Rhineland-Palatinate state of Germany to DR
(project: ‘Implications of global change for biological resources, law and
galathea from online data bases (Global Biodiversity standards’). We thank Claas Damken (Auckland, New Zealand) and Marc
Information Facility - GBIF; http://www.gbif.org) and our Meyer (Luxembourg) for field assistance. We thank the Fonds National de la
own field surveys. The accuracy of all records was Recherche Luxembourg for covering the publication fees.
checked in DIVA-GIS 5.4 [68] and only those which
Author details
could be unambiguously assigned to a single grid cell 1
Musée national d’histoire naturelle Luxembourg, L-2160 Luxembourg,
with a resolution of 2.5 arc min (ca. 4 km in the study Luxembourg. 2Terrestrial Ecology Unit, Department of Biology, Gent
University, B-9000 Gent, Belgium. 3Department of Biogeography, Trier
area) were used for further processing. Since unequal
University, D-54296 Trier, Germany. 4Zoologisches Forschungsmuseum
spatial clumping of species records may cause problems Alexander Koenig, Adenauerallee 160, 53113 Bonn, Germany.
when computing SDMs, the species records were filtered
Authors’ contributions
in geographic space, leaving only 1 record per 10 arc JCH sampled and analysed M. galathea-individuals, performed statistics, data-
min. The final data set comprised 535 records (Figure 3a) interpretation, and has written major parts of the manuscript. TS sampled
scattered all over the known range of the species in Eur- individuals, and supported the work by data-interpretation and writing. DR
performed species distribution models, and LL strengthened the significance
ope and North Africa.
of the manuscript by data-interpretation and by upgrading the linguistic
We obtained information on current and past climate style. All authors red and approved the final version of the manuscript.
as describedby the Community Climate System Model
Received: 27 January 2011 Accepted: 21 July 2011
(CCSM; http://www.ccsm.ucar.edu) with a spatial resolu- Published: 21 July 2011
tion of 2.5 arc min from the Worldclim data base ([69];
http://www.worldclim.org). Original palaeoclimatological References
data were previously processed as described by Peterson 1. Lomolino MV: Biogeography. Sinauer Assn 2005, 465.
2. Reinig WF: Die Holarktis. Gustav-Fischer-Verlag 1937, Jena.
and Nyári (2007) [70]. A total of 19 BIOCLIM variables
3. De Lattin G: Grundriß der Zoogeographie. Verlag Gustav Fischer 1967, Jena.
were previously suggested as suitable for SDM computa- 4. Dennis RLH, Williams WR, Shreeve TG: A multivariate approach to the
tion [71,72]. However, inclusion of too many inter-corre- determination of faunal units among European butterfly species
(Lepidoptera: Papilionoidea, Hesperioidea). Zoological Journal of the
lated variables or biologically irrelevant predictors may
Linnean Society 1991, 101:1-49.
hamper the transferability of SDMs through space and 5. Hewitt GM: Some genetic consequences of ice ages, and their role in
time [73-76]. Therefore, we first computed a pair-wise divergence and speciation. Biological Journal of the Linnean Society 1996,
58:247-276.
correlation matrix based on Pearson’s correlation coeffi- 6. Taberlet P, Fumagalli L, Wust-Saucy A-G, Cosson J-F: Comparative
cients among all 19 predictor variables and excluded phylogeography and postglacial colonization routes in Europe. Molecular
those with R2 > 0.75. Subsequently, we chose a final set Ecology 1998, 7:453-464.
7. Hewitt GM: Genetic consequences of climatic oscillation in the
of eleven predictors describing biologically relevant cli-
quaternary. Phil Trans R Soc Lond B 2004, 359:183-195.
mate conditions for the long-term persistence of M. 8. Schmitt T: Molecular biogeography of Europe: Pleistocene cycles and
galathea populations (i.e. annual mean temperature, postglacial trends. Frontiers in Zoology 2007, 4:11.
9. Hewitt GM: Post-glacial re-colonization of European biota. Biological
maximum temperature of warmest month, minimum Journal of the Linnean Society 1999, 68:87-112.
temperature of coldest month, temperature annual range, 10. Hewitt GM: The genetic legacy of the Quaternary ice ages. Nature 2000,
mean temperature of wettest quarter, mean temperature 405:907-913.
11. Cosson J-F, Hutterer R, Libois R, Sara M, Taberlet P, Vogel P:
of driest quarter, annual precipitation, precipitation of
Phylogeographical footprint of the Strait of Gibraltar and Quaternary
wettest quarter, precipitation of driest quarter, precipita- climatic fluctuations in the western Mediterranean: a case study with
tion of warmest quarter, precipitation of coldest quarter). the greater white-toothed shrew, Crocidura russula (Mammalia:
Soricidae). Molecular Ecology 2005, 14:1151-1162.
For SDM computation, Maxent 3.3.2 was applied
12. Carranza S, Arnold EN, Pleguezuelos JM: Phylogeny, biogeography, and
[77,78] using the default program settings. Random evolution of two Mediterranean snakes, Malpolon monspessulanus and
background records were automatically sampled by Hemorrhois hippocrepis (Squamata, Colubridae), using mtDNA sequences.
Molecular Phylogenetics and Evolution 2006, 40:532-546.
Maxent within the study area. Species records were split 13. Carranza S, Harris DJ, Arnold EN, Batista V, Gonzalez de la Vega JP:
100 times into 70% used for model training and 30% for Phylogeography of the lacertid lizard, Psammodromus algirus, in Iberia
model evaluation via the area under the receiver operat- and across the Strait of Gibraltar. Journal of Biogeography 2006,
33:1279-1288.
ing characteristic curve (AUC; [79]). Subsequently, the
14. Fritz U, Barata M, Busack SD, Fritsch G, Castilho R: Impact of mountain
average predictions for current and past conditions of chains, sea straits and peripheral populations on genetic and taxonomic
the logistic output of the 100 models were computed structure of a freshwater turtle, Mauremys leprosa (Reptilia, Testudines,
Geoemydidae). Zoologica Scripta 2006, 35:97-108.
and transformed into presence/absence maps applying
15. Franck P, Garnery L, Loiseau A, Oldroyd BP, Hepburn HR, Solignac M,
the average minimum training presence and average Cornuet J-M: Genetic diversity of the honeybee in Africa: microsatellite
10% training omission as thresholds. and mitochondrial data. Heredity 2001, 86:420-430.
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 13 of 14
http://www.biomedcentral.com/1471-2148/11/215

16. Habel JC, Rödder D, Scalercio S, Meyer M, Schmitt T: Strong genetic discriminatory value of genetics in the taxonomy. Organisms, Diversity
cohesiveness between Italy and the Maghreb in four butterfly species. and Evolution 2005, 5:297-307.
Biological Journal of the Linnean Society 2010, 99:818-830. 39. Reinig WF: Chorologische Voraussetzungen für die Analyse von
17. Santucci F, Emerson B, Hewitt GM: Mitochondrial DNA phylogeography of Formenkreisen. Syllegomena Biologica, Festschrift für O Kleinschmidt 1950,
European hedgehogs. Molecular Ecology 1998, 7:1163-1172. 346-378.
18. Seddon JM, Santucci F, Reeve NJ, Hewitt GM: DNA footprints of European 40. Lenk P, Fritz U, Joger U, Winks M: Mitochondrial phylogeography of the
hedgehogs, Erinaceus europaeus and E. concolor. Pleistocene refugia, European pond turtle, Emys orbicularis (Linnaeus 1758). Molecular Ecology
postglacial expansion and colonization routes. Molecular Ecology 2001, 1999, 8:1911-1922.
10:2187-2198. 41. Seddon JM, Reeve N, Hewitt GM: Caucasus Mountains divide postulated
19. Colliard C, Sicilia A, Turrisi GF, Arculeo M, Perrin N, Stöck M: Strong postglacial colonization routes in the white-breasted hedgehog,
reproductive barriers in a narrow hybrid zone of West-Mediterranean Erinaceus concolor. Journal of Evolutionary Biology 2002, 15:463-467.
green toads (Bufo viridis subgroup) with Plio-Pleistocene divergence. 42. Pinceel J, Jordaens K, Pfenninger M, Backeljau T: Rangewide
BMC Evolutionary Biology 2010, 10:232. phylogeography of a terrestrial slug in Europe: evidence for Alpine
20. Veith M, Mayer C, Samraoui B, Barrosso DD, Bogaerts S: From Europe to refugia and rapid colonization after the Pleistocene glaciations. Molecular
Africa and vice versa: evidence for multipe intercontinental dispersal in Ecology 2005, 14:1133-1150.
ribbed salamanders (Genus Pleurodeles). Journal of Biogeography 2004, 43. Cooper SJ, Ibrahim KM, Hewitt GM: Postglacial expansion and genome
31:159-171. subdivision in the European Grasshopper Chorthippus parallelus.
21. Paulo OS, Pinto I, Bruford MW, Jordan WC, Nichols RA: The double origin Molecular Ecology 1995, 4:49-60.
of Iberian peninsular chameleons. Biological Journal of the Linnean Society 44. Podnar M, Mayer W, Tvrtkovic N: Phylogeography of the Italian wall lizard,
2008, 75:1-7. Podarcis sicula, as revealed by mitochondrial DNA sequences. Molecular
22. Tennent J: The butterflies of Morocco, Algeria and Tunisia. Gem Ecology 2005, 14:575-588.
publishing Company Wallingford; 1996. 45. Canestrelli D, Cimmaruta R, Costantini V, Nascetti G: Genetic diversity and
23. Tolman T, Lewington R: Field guide butterflies of Britain and Europe. phylogeography of the Apennine yellow-bellied toad Bombina pachypus,
Harper Collins Publishers London; 1997. with implications for conservation. Molecular Ecology 2006, 15:3741-3754.
24. Asher J, Warren M, Fox R, Harding P, Jeffcoate G, Jeffcoate S: The 46. Steward JR, Lister AM: Cryptic northern refugia and the origins of the
millennium atlas of butterflies in Britain and Ireland Oxford University Press modern biota. Trends in Ecology and Evolution 2001, 16:608-613.
Oxford; 2001. 47. de Lattin G: Beiträge zur Zoogeographie des Mittelmeergebietes.
25. García-Barros E, Munguira ML, Martín Cano J, Romo Benito H, Garcia- Verhandlungen der deutschen Zoologischen Gesellschaft, Kiel 1949, 143-151.
Pereira P, Maravalhas ES: Atlas de las mariposas diurnas de la Península 48. Nève G, Verlaque R: Genetic differentiation between and among refugia.
Ibérica e islas Baleares (Lepidoptera: Papilionoidea and Hesperioidea). In In Relict species - phylogeography and conservation biology. Edited by: Habel
Monografias Sociedad Entomológica. Volume 11. Aragonesa; 2004. JC, Assmann T. Springer Heidelberg; 2010:277-294.
26. Buszko J: Atlas rozmieszczenia motyli dziennych w Polsce (Lepidoptera: 49. Habel JC, Augenstein B, Nève G, Rödder D, Assmann T: Population
Papilionidae, Hesperiidae). Edycja Turpress 1997, Torun. genetics and ecological niche modelling reveal high fragmentation and
27. Habel JC, Schmitt T, Müller P: The fourth paradigm pattern of postglacial potential future extinction of the endangered relict butterfly Lycaena
range expansion of European terrestrial species: The phylogeography of helle. In Relict species - phylogeography and conservation biology. Edited by:
the Marbled White butterfly (Satyrinae, Lepidoptera). Journal of Habel JC, Assmann T. Springer Heidelberg; 2010:417-440.
Biogeography 2005, 32:1489-1497. 50. Schmitt T, Seitz A: Allozyme variation in Polyommatus coridon
28. Habel JC, Meyer M, El Mousadik A, Schmitt T: Africa goes Europa: The (Lepidoptera: Lycaenidae): identification of ice-age refugia and
complete phylogeography of the Marbled White butterfly species reconstruction of post-glacial expansion. Journal of Biogeography 2001,
complex Melanargia galathea /lachesis. Organisms, Diversity and Evolution 28:1129-1136.
2008, 8:121-129. 51. Gratton P, Konopinski MK, Sbordoni V: Pleistocene evolutionary history of
29. Schmitt T, Habel JC, Zimmermann M, Müller P: Genetic differentiation of the Clouded Apollo (Parnassius mnemosyne): genetic signatures of
the marble white butterfly, Melanargia galathea, accounts for glacial climate cycles and a ‘time-dependent’ mitochondrial substitution rate.
distribution patterns and postglacial range expansion in southeastern Molecular Ecology 2008, 17:4248-4262.
Europe. Molecular Ecology 2006, 15:1889-1901. 52. Magri D: Patterns of post-glacial spread and the extent of glacial refugia
30. Swets K: Measuring the accuracy of diagnostic systems. Science 1988, of European beech (Fagus sylvatica). Journal of Biogeography 2008,
240:1285-1293. 35:450-463.
31. Nazari V, Hagen WT, Bozano GC: Molecular systematics and phylogeny of 53. Magri D, Vendramin GG, Comps B, Dupanloup I, Geburek T, Gomory D,
the Marbled Whites (Lepidoptera: Nymphidae, Satyrinae, Melanargia Latalowa M, Litt T, Paule L, Roure JM, Tantau I, van der Knaap WO, Petit RJ,
Meigen). Sysematics Entomology 2009, 35:132-147. de Beaulieu JL: A new scenario for the Quaternary history of European
32. Hein J, Schierup MH, Winf C: Gene genealogies, variation and evolution - beech populations: palaeobotanical evidence and genetic
A primer in coalescent theory. Oxford University Press, Oxford; 2005. consequences. New Phytologist 2006, 71:199-221.
33. Gibbard P, van Kolfschoten T: The Pleistocene and Holocene epochs. In A 54. Hampe A, Petit RJ: Conserving biodiversity under climate change: the
geologic time scale. Volume 98. Edited by: Gradstein FM, Ogg JG, Smith AG. rear edge matters. Ecology Letters 2005, 8:461-467.
Cambridge University Press, Cambridge; 2004:441-452. 55. Habel JC, Dieker P, Schmitt T: Biogeographical connections between the
34. Steinfartz S, Veith M, Tautz D: Mitochondrial sequence analysis of Maghreb and the Mediterranean peninsulas of southern Europe.
Salamandra taxa suggests old splits of major lineages and postglacial Biological Journal of the Linnean Society 2009, 98:693-703.
recolonizations of Central Europe from distinct source populations of 56. Nei M: Genetic distances between populations. The American Naturalist
Salamandra salamandra. Molecular Ecology 2000, 9:397-410. 1972, 106:283-291.
35. Carranza S, Wade E: Taxonomic revision of Algero-Tunisian Pleurodeles 57. Siegismund HR, Müller J: Genetic structure of Gammarus fossarum
(Caudata: Salamandridae) using molecular and morphological data. populations. Heredity 1991, 66:419-436.
Revalidation of the taxon Pleurodeles nebulosus (Guichenot, 1850). 58. Excoffier L, Larval G, Schneider S: Arlequin ver. 3.0: an integrated software
Zootaxa 2004, 488:1-24. package for population genetics data analysis. Evolutionary Bioinformatics
36. Fromhage L, Vences M, Veith M: Testing alternative vicariance scenarios Online 2005, 1:47-50.
in Western Mediterranean discoglossid frogs. Molecular Phylogenetics and 59. Saitou N, Nei M: The neighbor-joining method: a new method for
Evolution 2003, 31:308-322. reconstructing phylogenetic trees. Molecular Biology and Evolution 1987,
37. Quante M: The changing climate: future. In Relict species - phylogeography 4:406-425.
and conservation biology. Edited by: Habel JC, Assmann T. Springer 60. Felsenstein J: PHYLIP (Phylogeny Inference Package) Ver. 3.5.c.
Heidelberg; 2010:9-56. Department of Genetics, University of Washington Seattle, Washington; 1993.
38. Schmitt T, Varga Z, Seitz A: Are Polyommatus hispana and Polyommatus 61. Pritchard JK, Stephens M, Donnelly P: Inference of population structure
slovacus bivoltine Polyommatus coridon (Lepidoptera: Lycaenidae)? The using multilocus genotype data. Genetics 2000, 155:945-955.
Habel et al. BMC Evolutionary Biology 2011, 11:215 Page 14 of 14
http://www.biomedcentral.com/1471-2148/11/215

62. Guisan A, Zimmermann N: Predictive habitat distribution models in

ecology. Ecological Modelling 2000, 135:147-186.
63. Jeschke JM, Strayer DL: Usefulness of bioclimatic models for studying
climate change and invasive species. Annals of the New York Academy of
Sciences 2008, 1134:1-24.
64. Elith J, Leathwick JR: Species distribution models: Ecological explanation
and prediction across space and time. Annual Reviews in Ecology, Evolution
and Systematics 2009, 40:677-697.
65. Waltari E, Hijmans RJ, Peterson AT, Nyári AS, Perkins SL, Guralnick RP:
Locating pleistocene refugia: Comparing phylogeographic and
ecological niche model predictions. PLoS one 2007, 7:1-11.
66. Hijmans RJ, Guarino L, Jarvis A, O’brien R, Mathur P, Bussink C, Cruz M,
Barrantes I, Rojas E: DIVA-GIS version 5.2 manual. 2005.
67. Nogués-Bravo D: Predicting the past distributions of species climatic
niches. Global Ecology and Biogeography 2009, 18:521-531.
68. Rödder D, Weinsheimer F, Lötters S: Molecules meet macroecology -
combining species distribution modells and phylogeographic studies.
Zootaxa 2010, 2426:54-60.
69. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A: Very high resolution
interpolated climate surfaces for global land areas. International Journal
of Climatology 2005, 25:1965-1978.
70. Peterson AT, Nyári ÁS: Ecological niche conservatism and pleistocene
refugia in the thrush-like mourner, Shiffornis sp., in the neotropics.
Evolution 2007, 62:173-183.
71. Busby JR: Bioclim - a bioclimatic analysis and prediction system. In
Nature conservation: Cost effective biological surveys and data analysis. Edited
by: CR Margules, MP Austin. CSIRO, Melbourne; 1991:64-68.
72. Beaumont LJ, Hughes L, Poulsen M: Predicting species distributions: Use
of climatic parameters in bioclim and its impact on predictions of
species’ current and future distributions. Ecological Modelling 2005,
186:250-269.
73. Heikkinen RK, Luoto M, Araújo MB, Virkkala R, Thuiller W, Sykes MT:
Methods and uncertainties in bioclimatic envelope modeling under
climate change. Progress in Physical Geography 2006, 30:751-777.
74. Rödder D, Lötters S: Niche shift versus niche conservatism? Climatic
charactersistics of the native and invasive ranges of the mediterranean
house gecko (Hemidactylus turcicus). Global Ecology and Biogeography
2009, 18:674-687.
75. Rödder D, Schmidtlein S, Veith M, Lötters S: Alien invasive slider turtle in
unpredicted habitat: A matter of niche shift or of predictors studied?
PLoS one 2009, 4:e7843.
76. Rödder D, Lötters S: Explanative power of variables used in species
distribution modelling: An issue of general model transferability or
niche shift in the invasive greenhouse frog (Eleutherodactylus
planirostris). Naturwissenschaften 2010, 97:781-796.
77. Phillips SJ, Anderson RP, Schapire RE: Maximum entropy modeling of
species geographic distributions. Ecological Modelling 2006, 190:231-259.
78. Phillips SJ, Dudík M: Modeling of species distributions with Maxent: new
extensions and a comprehensive evaluation. Ecography 2010, 31:161-175.
79. Pearce J, Ferrier S: An evaluation of alternative algorithms for fitting
species distribution models using logistic regression. Ecological Modelling
2000, 128:128-147.

doi:10.1186/1471-2148-11-215
Cite this article as: Habel et al.: From Africa to Europe and back: refugia
and range shifts cause high genetic differentiation in the Marbled
White butterfly Melanargia galathea. BMC Evolutionary Biology 2011
11:215.
Submit your next manuscript to BioMed Central
and take full advantage of:

• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit