Page 4 - BMC_1471

BMC Evolutionary Biology 2008, 8:56 http://www.biomedcentral.com/1471-2148/8/56

substantially different from other western Mediterranean terior support: 97% and 100%) and demonstrated that
green toad forms. These phylogeographic relationships nuclear and mitochondrial markers show essentially the
substantiate scarce knowledge on western Mediterranean same signals.
terrestrial biogeography and may have implications for
comparative research on the phylogeography of other ter- The control region tree also revealed three well-supported
restrial animals in the region. subclades: one comprising toads from Corsica and Sar-
dinia which exhibit some geographic intermixing between
Results genetically differentiated lineages within the islands; a
Phylogeographic structure reveals a relationship across the second much-less structured clade including toads from
Strait of Sicily the entire Apennine Peninsula from Turin, Pavia and
We found genetic markers in green toads (Additional file Marche (loc. 13, 14, 55, 56) in the north to Apulia (loc.
1) within our geographic scope (Figure 1) to indicate five 17) and Calabria (loc. 19) in the far south; and a third
major spatially structured lineages (Figure 2). (I) The first subclade (nested between the two others) containing
lineage was found on Corsica, Sardinia, the Balearic toads from the Balearic Islands. Subclade structure is evi-
Islands, Apennine Peninsula and the northeastern dent in the average F values between each of these subc-
ST
extreme of Sicily. This clade (balearicus) was different from lades (Table 1): Analyses of population structure
other Eurasian mainland green toads (lineages II and III: confirmed high genetic differentiation (pairwise F ST =
viridis and variabilis), whose ranges border northeast to the 0.7585, p = 0.00000) between Italian mainland and
Po River drainage and belong to widespread mono- island toads (Corsica, Sardinia). Demographic analyses
phyletic groups in Eurasia [28]. (IV) Although geographi- performed in Fluctuate estimated a greater than ten-fold
cally neighboring I, on most of Sicily and its surrounding exponential growth rate for the clade on the Apennine
islands another lineage (Bufo n. sp.) occurs that is substan- Peninsula than for the mitochondrial group on Corsica
tially different from the first. (V) The entire North African and Sardinia (767.1 > 51.7), suggesting population
range and the off-coast islands constitutes a fifth lineage expansion on the mainland. However, log-likelihood tests
(boulengeri). The details of the results defining each line- did not reject a scenario of zero growth (Table 2). Growth
age are described below. of the mainland population was also not significantly
supported by the mismatch distribution analysis in which
In particular, our analyses shows a deep sister relationship the observed distribution matched an expected distribu-
between African (V) and Sicilian (VI) groups and there- tion for an expanding population; however, the shape and
fore a rarely studied phylogeographic connection across magnitude of the observed and expected distributions are
the Strait of Sicily. very similar, suggesting some support for expansion on
the mainland (Figure 3c). By contrast, this analysis, both
Characterization of the groups identified based on the graphical and statistical, supports stable populations on
Bayesian phylogram Corsica and Sardinia (Figure 3a, b). Tajima's D estimates
Here we consider and name lineages that maintain their for all balearicus groupings were not significant, and thus
evolutionary integrity with respect to other lineages population expansion was not indicated by this test.
through both time and space and name them as species
under the phylogenetic species concept. The Bayesian phylogenetic analysis shows B. balearicus to
be reciprocally monophyletic on the mitochondrial level,
Bufo balearicus Boettger, 1880 albeit using a subsample of individuals within each
For details on taxonomy and nomenclature of Bufo balear- group, with respect to all other green toads; thus our data
icus [45] see supporting data in Additional file 2 (d). Anal- corroborate the proposal that B. balearicus be given the
yses of mitochondrial 16S (not shown in tree) and control status of a separate species (see Additional file 2 for details
region sequences of 50 green toads from the Balearic on nomenclature).
Islands (loc. 5–7), Corsica (loc. 8–9), Sardinia (loc. 10–
12), the entire Apennine Peninsula (loc. 13–16), Apulia Bufo boulengeri Lataste, 1879 [46]
(loc. 17), Calabria (loc. 18, 19), Marche (loc. 55, 56) and While exhibiting some internal structure, all North Afri-
northeastern Sicily (loc. 20) form one very well-supported can control region (Figure 2) and 16S mitochondrial
haplotype group (Bayesian posterior support: 100%; Fig- sequences (Additional file 1) of toads between the Atlan-
ure 1, 2). Cloned tropomyosine intron alleles (Figure 2) tic coast of Morocco (loc. 1–3) and the Nile Valley of
from four representatives each from Sardinia (loc. 10), Egypt (loc. 37–40), including the off-coast islands of
Calabria (loc. 18) and northeastern Sicily (loc. 20), as well Kerkennah (loc. 30, 31), Djerba (loc. 32) and Lampedusa
as fragments of RAG-1 in three individuals each from Cor- (loc. 26), form a well-supported monophyletic cluster
sica (loc. 9), Sardinia (loc. 10), and Calabria (loc. 18) (posterior probability = 99%) that is a sister clade to mito-
formed similarly highly supported clusters (Bayesian pos- chondrial sequences of toads from Sicily (loc. 21–25), but

Page 4 of 19
(page number not for citation purposes)