Page 13 - Colliard_ali_2010

Colliard et al. BMC Evolutionary Biology 2010, 10:232 Page 13 of 16
http://www.biomedcentral.com/1471-2148/10/232

Pliocene divergence (3.6 Mya) [20], indicating a lack of toads, and tail tips from tadpoles. Most adults were
current gene exchange. By contrast, phylogeographic released at the sampling sites, some vouchers were
studies [e.g. [16,17]] have shown that clades of more deposited in institutional collections [MVZ, NME,
recent divergence [1.33 Mya; [17]] form “wide hybrid ZFMK,details Appendix 1in23],and tissues were
zone(s) with a considerable genetic exchange between stored in 98% ethanol and/or at -20°C.
the two gene pools” [16], suggesting that reproductive
barriers are still low or inexistent. Crossing experiments
The observed post-mating barriers (hybrid breakdown) Toad crosses were performed with mature adult males
in Sicilian green toads certainly induce selection to act and females in a naturally reproductive state during the
against hybridization. Therefore, we expect some pre- breeding period in spring. Females were stimulated to
mating mechanisms to have evolved since the first con- spawn by injection of 0.1 ml of a 0.9% NaCl-solution
tact of these allopatrically evolved lineages, which might containing 500-1000 IU of human choriogonadotropin
also explain the absence of F 1 -hybrids. These two spe- (Sigma).
cies have already been shown to differ in breeding phe- A first crossing experiment was made in the laboratory
nology [23]. Pre-mating isolation between closely related in 2007 between a female B. balearicus (Si 41, pop. 11)
species (through mating calls and female choice) is and a male B. siculus (Si 11, pop. 22), from which 200 F 1 -
widespread in anurans, sometimes even in absence of hybrid tadpoles were raised. Ten randomly chosen off-
post-mating isolation [56]. The nature of barriers might spring were genotyped and added to our sampling
also affect the structure of hybrid zones, since, due to ("population 25”). Seven others crossing experiments
female choice, mtDNA is expected to introgress more were made in the laboratory in 2009 using two B. balear-
readily than nuclear DNA under pre-zygotic isolation icus individuals (male: Si 336; female: Si337; Sicily, Mar-
mechanisms (and even more so when the invading spe- ina S. Biagio, pop. 9), two B. siculus individuals (male: Si
cies is relatively rare compared to residents) [57]. 334; female:Si335;Sicily,Pergusa Lake,provinceof
Further insights on the evolutionary interactions Enna, 37°31’00.29” N, 14°18’04.34"E, W of pop. 18 and
between B. siculus and B. balearicus populations might 19) and two F 1 -hybrid individuals coming from the first
be gained by collecting bio-acoustic and mate-choice crossing experiment (male: Cross 13; female: Cross 11).
data, together with additional crossing experiments. As no proper control for intraspecific crosses could be
Whatever the exact causes, our data clearly show a performed (due to lack of available females), we provide
virtual absence of gene flow at the present contact zone results from intraspecific matings of other green toad
(corroborated with a hybrid breakdown in backcrosses lineages (Table 3: Control 1: B. turanensis,Control2: B.
and F 2 ), meaning that the speciation process can be pewzowi), which were raised under identical conditions.
considered as close to have reached complete reproduc- Crossing pairs from 2009 were observed during
tive isolation after some 2.7 My of divergence. These amplexus every hour. When 10-15 cm of clutch strings
field data contrast sharply with the results from experi- had been deposited, couples were removed from the
mental hybridization in anurans, which show that some tank and separated, animals were rinsed to avoid sperm
lineages may still produce viable F 1 offspring after ca. 20 contamination and arranged in new cross combinations
My of divergence (see Background). Information on F 1 - in another tank (see Table 3 for cross combinations).
hybridizability or viability gained under laboratory con- Clutch was left untouched until hatching or until visi-
ditions may thus grossly overestimate the time required ble signs of dying eggs/embryos were found that had to
for genetic isolation and speciation to occur in anurans. be removed to avoid a chain-reaction of embryo suffoca-
tion. After one month, 100 tadpoles were randomly cho-
Methods sen among the surviving (if so) offspring, raised in
Sampling shallow oxygenated aquaria and fed with Elodea plants
Samples from 323 specimens of green toads were col- and fish food (Tetramin) under identical conditions.
lected during intense fieldwork between 2004 and 2007 After the second month of development, percentage of
at 24 localities across the Italian Peninsula and Sicily surviving tadpoles was determined by individual counts.
(Figure 1 and Table 1). Italian Peninsula was added to All of these tadpoles were further raised until metamor-
the sampling to allow us to understand the context of phosis or death. Toadlets were fed with Drosophila and
the B. balearicus invasion in Sicily and to compare Sici- juvenile crickets, and mealworms with weekly additions
lian B. balearicus genotypes with those of mainland ori- of calcium and vitamins. Illumination of terrariums
gin. A few samples came from scientific collections included sun-light spectrum fluorescent lamps including
(MVZ, NME, ZFMK) or were collected throughout the a natural-like UV fraction. Clutch,embryonic,tadpole
years (e.g. road-kills). Tissue samples consisted of finger and juvenile development were documented photogra-
tips and muscles (road kill) from sub adult and adult phically (Figures 5 to 7).