ISSN 1827-9635 (print) © Firenze University Press ISSN 1827-9643 (online) www.fupress.com/ah Acta Herpetologica 6(1): 35-45, 2011 The phylogenetic position of Lygodactylus angularis and the utility of using the 16S rDNA gene for delimiting species in Lygodactylus (Squamata, Gekkonidae) Riccardo Castiglia*, Flavia Annesi Dipartimento di Biologia e Biotecnologie “Charles Darwin”, Università “La Sapienza” di Roma, Via Borelli 50, I-00161 Rome, Italy. *Corresponding author. E-mail: castiglia@uniroma1.it Submitted on: 2010, 27th December; revised on: 2011, 25th March ; accepted on: 2011, 27th April . Abstract. The African genus Lygodactylus Gray, is composed of roughly 60 spe- cies of diurnal geckos that inhabit tropical and temperate Africa, Madagascar, and South America. In this study, we assessed the phylogenetic position of L. angularis, for which molecular data were so far lacking, by means of sequence analysis of the mitochondrial 16S rDNA gene. We also compared intraspecific vs. interspecific genet- ic divergences using an extended data set (34 species, 153 sequences), to determine whether a fragment of this gene can be useful for species identification and to reveal the possible existence of new cryptic species in the genus. The analysis placed L. angu- laris in a monophyletic group together with members of “fischeri” and “picturatus” groups. Nevertheless, the independence of the “angularis” lineage is supported by the high genetic divergence. Comparison of intraspecific vs. interspecific genetic dis- tances highlights that, assuming an equal molecular rate of evolution among the stud- ied species for the used gene, the threshold value useful for recognising a candidate new species can be tentatively placed at 7%. We identified four species that showed an intraspecific divergence higher than, or close to, the 7% threshold: L. capensis (8.7%), L. gutturalis (9.3%), L. madagascariensis (6.5%) and L. picturatus (8.1%). Moreover, two species, L. mombasicus and L. verticillatus, are paraphyletic in terms of gene gene- alogy. Thus, the study shows that a short fragment of the 16S rDNA gene can be an informative tool for species-level taxonomy in the genus Lygodactylus. Keyword. African reptiles, biodiversity, DNA barcoding, genetic divergence, sister species, taxonomy. INTRODUCTION The genus Lygodactylus Gray, 1864 encompasses about 60 species of diurnal geckos that inhabit tropical and temperate Africa, Madagascar, and South America (Spawls et al., 36 R. Castiglia and F. Annesi 2002). Lygodactylus has traditionally been divided into different “species groups” based on phenotypic characteristics (Loveridge, 1947; Pasteur, 1965; Jacobsen, 1992; Röll, 2000, 2004; Puente et al., 2009). The phylogenetic relationships among species from Madagascar have been assessed by Puente et al. (2005) using a mitochondrial gene. Recently, Röll et al. (2010) provided a robust phylogeny of 28 species by using both mitochondrial and nuclear genes. This phylog- eny mainly confirmed the monophyly of many of the species groups, previously identified on the base of morphology, as well as provided evidence for at least two additional lineages. Moreover, this study identified some divergent taxa which possibly belonged to additional undescribed species. In fact, the taxonomy of Lygodactylus at a species level is still uncertain, with some species known only from the type locality. Furthermore, the earlier identification of many species was based on scale features only and, consequently, due to their small sizes, the identification during field work is extremely difficult (Spawls et al., 2002). A recent approach to studying species limits involves the use of a single short frag- ment of DNA that can help in the identification of an organism by its assignment to a pre- viously described species (DNA barcoding). In the same way, it is possible to predict and describe new taxa using DNA (DNA taxonomy) (Blaxter, 2004). Moreover, this approach allows inclusion of additional information in a taxonomy based on morphological char- acters only (Padial et al., 2010). New candidate species can be identified on the basis of 1) elevated intraspecific genetic distances, similar to the ones found among different species and 2) the presence of paraphyletic species that suggests a need of taxonomic revision. Yet, this approach has some limitation, as in case of incomplete gene sorting, hybridiza- tion, different rates of molecular evolution at the same marker in a group of species or accidental amplification of nuclear mitochondrial pseudogenes (Frézal and Leblois, 2008). Different genes are currently used as DNA barcodes in animals (e.g., cytochrome oxi- dase subunit I – COI, 16S rDNA – 16S, cytochrome b - cytb). Evidence suggests that no “universal” gene exists. To be useful for delimiting species, the gene must have one prin- cipal characteristic that is to exhibit limited overlap between intraspecific and interspe- cific divergence. For example, an accurate comparison of performances of COI and 16S in amphibians showed that 16S is well suited to distinguish between intraspecific and inter- specific divergence, whereas, in birds, the best results were obtained using COI (Vences et al., 2005; Aliabadian et al., 2009). In this study, we first analyse the phylogenetic position of L. angularis, a member of the “angularis” species group together with L. grzimeki Bannikov, using a mitochondrial molec- ular marker. At present, there are no other molecular data on the members of this species group neither to confirm that they both belong to a separate lineage, nor to place them in a phylogenetic context. Besides, Lygodactylus seems to be an appropriate genus for testing the utility of a single DNA fragment as a DNA barcode, because of the problems in identifica- tion and unstable taxonomy. A recent paper by Chiari et al. (2009), concerning species from Madagascar, highlighted that two of the eight analysed species (L. madagascariensis and L. tolampyae) showed a very high intraspecific genetic divergence using a fragment of the 16S rDNA. Consequently, in this paper, we extend the record with a dataset formed by a higher number of species in order to determine whether the same genetic marker can be useful for species identification. Finally, we reveal the possible presence of new cryptic species in Lygo- dactylus by providing evidence of genetic divergence and species paraphyly. 37DNA taxonomy in Lygodactylus MATERIALS AND METHODS We utilised almost all the available 16S rDNA sequences from Lygodactylus downloaded from GenBank (updated on February 2011; 34 species, 153 sequences; Table 1). Sequences from specimens with uncertain taxonomical attribution were excluded. Sequences from additional speci- Table 1. Species, number of localities and number of individuals used for the comparative analysis done on the 16S rRNA gene fragment. Lygodactylus sp. 1 and sp.2 are undescribed species (Röll et al. 2010). Data from Chiari et al. (2009), Röll et al. (2010), Puente et al. (2005) and Rocha et al. (2009) and present data. Species Number of Localities Number of specimens L. angularis Günther (1893) 1 1 L. arnoulti Pasteur (1964) 1 5 L. bivittis (Peters, 1883) 1 4 L. blancae Pasteur (1995) 1 2 L. bradfieldi Hewitt (1932) 2 3 L. capensis (Smith, 1849) 6 9 L. chobiensis Fitzsimons 1932 1 2 L. conraui Tornier (1902) 1 2 L. gravis Pasteur (1964) 1 2 L. grotei Sternfeld (1911) 1 2 L. guibei Pasteur (1964) 2 3 L. gutturalis Bocage (1873) 2 5 L. keniensis Parker (1936) 1 3 L. kimhowelli Pasteur (1995) 1 4 L. laterimaculatus Pasteur 1964 1 3 L. lawrencei Hewitt (1926) 3 3 L. madagascariensis (Boettger, 1881) 3 6 L. miops Günther (1891) 1 4 L. mirabilis Pasteur (1962) 1 31 L. mombasicus Loveridge (1935) 2 4 L. montanus Pasteur (1964) 1 2 L. pauliani Pasteur and Blanc (1991) 1 1 L. picturatus (Peters, 1868) 2 4 L. pictus (Peters, 1883) 3 12 L. rarus Pasteur and Blanc (1973) 1 2 L. stevensoni Hewitt (1926) 1 2 L. thomensis (Peters, 1880) 1 1 L. tolampyae (Grandidier, 1872) 3 5 L. tuberosus Mertens (1965 1 11 L. verticillatus Mocquard (1895) 2 5 L. williamsi Loveridge (1952) 1 2 Lygodactylus sp. 1 1 4 Lygodactylus sp. 2 1 1 38 R. Castiglia and F. Annesi mens belonging to four species were added to the dataset (L. mombasicus, Nairobi, Kenya, 01°16’S - 36°49’E; L. picturatus, Morogoro, Tanzania, 06°49’S - 37°40’E ; L. capensis, Mutanda, Zambia, 12°22’S - 26°16’E; L. angularis, Mbeya, Tanzania, 12°33’S - 25°41’E; accession numbers HQ872459- 63). The new 16S rRNA gene sequences were obtained from tissues fixed in ethanol 80%. DNA was extracted using the QIAmp tissue extraction kit (Qiagen). The primers 16SA-L (light chain; 59-CGC CTG TTT ATC AAA AAC AT-39) and 16SB-H (heavy chain; 59-CCG GTC TGA ACT CAG ATC ACG T-39) were used to amplify a section of the mitochondrial 16S ribosomal RNA gene (Palumbi et al. 1991). The PCR cycling procedure was performed as follows: 34 cycles of denaturation for 90 sec at 95 °C, primer annealing for 60 sec at 50 °C, and extension for 90 sec at 72 °C. All sequences were aligned with Muscle, using default settings, and then adjusted manually. Sites including gaps and hypervariable regions, identified by visual inspection of the alignment, were removed. The final alignment was 432bp long. A preliminary NJ tree was built with this dataset, using Kimura 2-parameter distances (K2P) and 10000 bootstrap replicates generated by MEGA 4 (Kumar et al., 2008). We calculated the intraspecific distance for each species and for each species with 2 or more populations separately. Average K2P distances were computed based on pairwise comparisons of all sequences for each of these species. A correlation was made between number of localities and average intraspecific genetic distance. Information about the number of populations were obtained from the published articles. For interspecific distances, we calculated mean pairwise K2P distance among all pairs of species and, separately, between pairs of sister species. Sister species were identi- fied according to our phylogenetic tree (see below) and by referring to the tree proposed by Röll et al. (2010). To study the phylogenetic position of L. angularis, the dataset was pruned to reduce the computation time of analyses, and only two or three sequences were retained for each species (80 sequences, 432bp, from 35 species). When possible, we kept sequences from different populations for each species. Phylogenetic relationships were assessed by Bayesian inference (BI), unweighted maximum parsimony (MP) and maximum likelihood (ML). We used the same outgroups used by Röll et al. (2010), i.e., Phelsuma standingi and Rhoptropella ocellata, since they are the closest rela- tives of Lygodactylus (Austin et al., 2004). The appropriate model of substitution was chosen using the Model Test 3.7 program (Posada and Crandall, 1998). Models of evolution, which provide the best approximation to the data, were chosen for subsequent analysis according to the Akaike information criterion (AIC). The chosen model was the General Time Reversible (GTR) model with rate variation among sites (+G), a pro- portion of invariable sites I=0.2038 and a gamma distribution shape parameter of 0.3846. The mod- els and parameters were used for ML trees in phyML (Guindon and Gascuel, 2003). Maximum parsimony trees were obtained with PAUP 4.0b10 (Swofford, 2000) using a heuris- tic search and tree –bisection– reconnection and random addition of sequences. The robustness of the nodes was assessed using the bootstrap with 1000 replicates for MP and 500 replicates for ML. For the BI we constructed the phylogeny using the software MrBayes v. 3.1.2 (Huelsenbeck and Ronquist, 2001) using the same model as in the ML analysis. Two independent Markov chain Monte Carlo analyses were run. We used 1 million generations, four chains and a burn-in of 10% of the generated tree. Based on our phylogenetic results, the paraphyly of L. mombasicus and L. verticillatus was further investigated using a reduced dataset comprising sequences of species from the same species group. For L. mombasicus the dataset was composed of twenty-two sequences (482bp) belonging to seven species. For L. verticillatus the dataset was composed of twenty sequences (482bp) belonging to seven species. In this procedure, longer alignments could be used, since the internal hypervari- able region is less extended when close species are compared, which offers greater power for phylo- genetic resolution. Phylogenetic relationships were then assessed by BI, MP and ML. 39DNA taxonomy in Lygodactylus RESULTS Phylogenetic position of L. angularis. The phylogenetic analysis of the pruned dataset (432 bp, 80 sequences from 35 species) does not generally conflict with a previous phy- logeny obtained by multigene analysis (Röll et al., 2010) and exhibits deep, weakly sup- ported branches connecting well-supported species groups. The only species group that is not supported by present analysis is the “pictus-mirabilis” group, which is split into three lineages (not shown). The analysis places L. angularis in a moderately supported clade, which includes spe- cies of the “picturatus” and “fischeri” groups. (Fig. 1A). Within this clade, the species of the “fischeri” group and of the “picturatus” group are well supported. However, basal rela- tionships within this clade are not resolved. Instances of paraphyletic species. Our phylogenetic analysis identified two putative instances of paraphyletic species (L. mombasicus and L. verticillatus). The paraphyly was fur- ther analysed with a dataset including species related to either of them. For the “picturatus” group, the dataset retrieved a topology (Fig. 1A) similar to the one obtained by Röll et al. (2010) that also included nuclear data. However, haplotype LYG1 from L. mombasicus has a basal position with respect to a cluster formed by the other sequences belonging to L. mom- basicus and L. kimhowelli. This topology is well supported by all the methods used (Fig. 1A). For L. verticillatus, the paraphyly is straightforward, since two distinct clusters within L. verticillatus have been found (min-max, 8.8-9.1% sequence divergence). One of the two clusters has high sequence similarity with the specimens identified as L. heterurus (2.3% sequence divergence) (Fig. 1B). Intraspecific vs. interspecific genetic distance. Figure 2A shows the distribution of interspecific and intraspecific average pairwise genetic divergence. The two distributions overlap. The amount of overlap is between 4.9% and 9%. For intraspecific divergence, the range is from 0 to 9% (mean 1.8%; s.e. 2.6; N = 30), and four out of 30 values lie in the overlapping interval. Interspecific distances range from 4.9 to 37% (mean 24.2%; s.e. 6.0%), and eight out of 595 values lie in the overlapping interval. For interspecific pairwise divergence, we eliminated two pairs of close species that showed paraphyly (see above). The intraspecific genetic distance is strongly correlated with the number of localities (R = 0.78; P<<0.0001). Accordingly, the level of overlap between intra- and interspecific diver- gences can be inflated by the fact that we underestimated the intraspecific divergence: in many species, only one or a few populations were available for the analysis. For this reason we compared (Fig. 2B) the distribution of intraspecific genetic distanc- es for species including more than two populations with the distribution of genetic diver- gences between pairs of sister species. We used pairs of sister species in order to capture the divergence of recently emerged species. The intraspecific distances for species with more than 2 populations range between 0.6% to 9% (mean 4.4%, s. d. 2.8, N=11). The distribution of genetic divergences between pairs of sister species ranges between 4.9 and 15.7% (mean 10.5%, s.e. 4.0% N = 8). The two distributions overlap between 5% and 9%. The species with high average intraspecific genetic divergences are L. capensis (8.7%), L. gutturalis (9.3%), L. madagascariensis (6.5%), L. tolampyae (5.3%) and L. picturatus (8.1%). The sister species with the lowest intraspecific divergence are L. arnoulti and L. pauliani (4.9%). 40 R. Castiglia and F. Annesi  Fig. 1. Bayesian trees of A) the clade including the “picturatus” group, the “fischeri” group and L. angularis and B) the haplotypes belonging to L. verticillatus and L. heterurus. The haplotypes analysed are in bold. Numbers at the node correspond to the Bayesian posterior probabilities, the percentage of 1000 bootstrap replicates for maximum parsimony and over 500 replicates for Maximum Likelihood respectively. 41DNA taxonomy in Lygodactylus DISCUSSION Phylogenetic position of L. angularis. Even if the basal nodes of the tree are not sup- ported, the phylogenetic position of L. angularis is well-supported in our tree, which was built with a single mtDNA gene. This species resulted in a monophyletic group togeth-  Fig. 2. Pairwise K2P pairwise distances in the 16S gene in Lygodactylus. A) Black bars are comparisons among conspecific sequences (left axis); grey bars represent comparisons among different species (right axis). In (B), black bars refer to comparisons of conspecific specimens for species with two or more popu- lations. Gray bars refer to pairs of sister species only. 42 R. Castiglia and F. Annesi er with members of “fischeri” and “picturatus” groups. The clustering of the members of these two species groups was previously supported by Roll et al. (2010). All these species share some scale characters, such as an undivided mental scale in most species. This char- acter is also present in L. angularis (Loveridge, 1947), in accordance with the presently determined phylogenetic position of this species. The “fischeri” group contains West Afri- can species, and the “picturatus” group consists of mainly East African species. L. angula- ris shares a similar geographic distribution in eastern Africa with the member of the “pic- turatus” group. Moreover, L. angularis is also large-sized, as the species of the “picturatus” group, whereas geckos of the “fischeri” group are small and slender. Another character that may be in common between members of the “picturatus” group and L. angularis, is the presence of sexual dichromatism with coloured males. In fact, L. angularis males have rose pink ventral surface, while females are entirely lemon yellow. Nevertheless, the inde- pendence of the “angularis” lineage from the species of “picturatus” and “fischeri” groups is supported by the high genetic divergence (14-20%). Inter- and intraspecific genetic divergence. The distribution of intraspecific and inter- specific genetic divergence shows an overlap between 5% and 9%. The distribution of the interspecific divergence represents probably a good approximation of the real differences between species. However, we found that intraspecific divergence depends strongly on the number of localities sampled. For this reason, we suggest that the intraspecific diversity might be underestimated. Despite this limitation, the threshold values useful for recognising a “good” species, according to our data, can be tentatively placed at 7% (K2P distance). In fact, only a cou- ple of species shows an interspecific genetic divergence below this value (L. arnoulti and L. pauliani, 4.9% genetic divergence). Owing to scarcity of data on intraspecific variation, we cannot say whether the use of a lower threshold (5%) would produce an excessive number of false positives. Additional intraspecific data are needed to resolve this problem and ulti- mately to support a threshold shift from 7% to 5%. We identified four species that showed an intraspecific divergence higher than, or close to, the 7% threshold: L. capensis (8.7%), L. gutturalis (9.3%), L. madagascariensis (6.5%) and L. picturatus (8.1%). Present data confirm and extend previous results obtained with the same mtDNA marker on eight species of Lygodactylus, among which L. madagascariensis and L. tolampyae showed high intraspecific divergence (4.3% and 9.1%) (Chiari et al., 2009). Comparing the present and the previously published data, one can see that the differences in values of genetic divergences are most likely due to the different number of individuals (for in L. tolampyae) and the different length of the sequence used. In fact, the 16S rDNA gene has conserved regions, but also hypervariable ones. Consequently, the length of the fragment used for analysis affects easily the calculation of divergence. For a barcoding approach, obviously, the inter- and intraspe- cific divergence should be only compared by means of the same dataset. The species with high intraspecific divergences, L. gutturalis and L. capensis, are wide- ly distributed. The divergent sequences belong to distant populations (Guinea Bissau and Uganda for L. gutturalis; Zambia, Namibia, and South Africa for L. capensis). Additional nuclear markers and morphological characteristics should be used to infer the taxonomic status in these cases. The same concern arises for L. picturatus, a species in south-eastern Kenya and eastern Tanzania. The specimen of from central Tanzania studied here shows a 43DNA taxonomy in Lygodactylus high divergence and results basal to the other sequences from the populations of the Kenyan and Tanzanian coasts. Current information on L. madagascariensis, distributed in the north- western part of Madagascar, is too scant to allow any tentative discussion of species limits. Two instances of paraphyletic species. We identified two paraphyletic species in terms of gene genealogy, namely L. mombasicus and L. verticillatus. In fact, the haplotype of L. mombasicus from Kenya, here analysed, is basal to other sequences from L. mombasicus and L. kimhowelli. According to Röll et al. (2010), L. kimhowelli and L. mombasicus are very similar morphologically. Their colour patterns differ, but they share all scale charac- ters and the pattern of conspicuous black markings on head and neck (Röll, 2003). Genet- ic divergence between haplotypes of L. mombasicus and L. kimhowelli is low (2.6-3.1%), whereas the haplotype from Kenya diverges by 6% with respect to both of them. This value is only slightly lower than the threshold values identified above. It is evident that an accurate morphological and molecular study of specimens belonging to L. mombasicus and L. kimhowelli is needed before any conclusions regarding species status can be drawn. Another species found to be paraphyletic is L. verticillatus. Röll et al. (2010) conclud- ed that L. verticillatus and L. heterurus, small geckos that are very similar in morphol- ogy, are also quite similar genetically. For this reason, the authors proposed conspecificity of the two taxa. Our analysis includes two additional sequences (Puente et al., 2005) and reveals a more complex pattern. In fact, these sequences are the sister group of the other sequences belonging to L. verticillatus and L. heterurus, with a very high genetic diver- gence of 9%. This value is higher than the 7% threshold proposed above and suggests that L. verticillatus and L. heterurus are two different sister species. CONCLUSION This study shows that a short fragment of the 16S rDNA gene may be an informative tool for species-level taxonomy in the genus Lygodactylus. However, other genes should be tested, since no comparative results with other molecular markers are reported in this paper. Future studies should be planned to obtain a more accurate picture of intraspecific divergence for the studied gene, as well as for other molecular markers, and to include samples from underrepresented African regions. Finally, additional nuclear markers and other kind of data (for example, behavioural) should be analysed in those instance, where high intraspecific divergences were confirmed. ACKNOWLEDGEMENTS We wish to express our gratitude to Georges Pasteur, who kindly identified the specimens studied in this paper and Ekaterina Gornung for critical reading of the manuscript. We also thank Lucia Salis for help in laboratory. Two anonymous referees provided very helpful comments to the manuscript. This work was supported by funds “Università” (grants to R.C.). 44 R. Castiglia and F. Annesi REFERENCES Aliabadian, M., Kaboli, M., Nijman, V., Vences, M. (2009): Molecular identification of birds: performance of Distance-Based DNA barcoding in three genes to delimit parapatric species. PLoS ONE 4: e4119. doi:10.1371/journal.pone.0004119 Austin, J.J., Arnold, E.N., Jones, C.G. (2004): Reconstructing an island radiation using ancient and recent DNA: the extinct and living day geckos (Phelsuma) of the Mas- carene islands. Mol. Phylogenet. Evol. 31: 109-122. Blaxter, M., L. (2004): The promise of a DNA taxonomy. Phil. Trans. R. Soc. Lond. B 359: 669-679. Chiari, Y., Vieites, D.R., Guo, J., Bora, P., Vences, M. (2009): High haplotype diversity in a microendemic Malagasy gecko species, Lygodactylus mirabilis (Pasteur, 1962). Zootaxa, 2269: 43-52. Frézal, L., Leblois, R. (2008): Four years of DNA barcoding: current advances and pros- pects. Infect. Genet. Evol. 8: 727-736. Guindon, S., Gascuel, O. (2003): A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 52: 696-704. Huelsenbeck, J.P., Ronquist, F. (2001): MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics. 17:754-755. Jacobsen, N. (1992): New Lygodactylus taxa (Reptilia: Gekkonidae) from the Transvaal. Bonn. Zool. Beitr. 43: 527–542. Kumar, S., Dudley, J., Nei, M., Tamura, K. (2008): MEGA: A biologist-centric software for evolutionary analysis of DNA and protein sequences. Brief. Bioinform. 9: 299-306. Loveridge, A. (1947): Revision of the African lizards of the family Gekkonidae. Bull. Mus. Comp. Zool., Cambridge (Mass) 98: 1–469. Padial, J.M., Miralles, A., De la Riva, I., Vences, M. (2010): The integrative future of tax- onomy. Front. Zool. 7: 16 doi:10.1186/1742-9994-7-1 Palumbi, S.R., Martin, A.P., Romano, S.L., McMillan, W.O., Stice, L., Grabowski G. (1991): The Simple Fool’s Guide to PCR. Dept. of Zoology, University of Hawaii, Honolulu. Pasteur, G. (1965): Recherches sur l’évolution des lygodactyles, lézards afromalgaches actuels. Trav. Inst. Scient. Chérif., Sér. Zool. 29: 1-132. Posada, D., Crandall, A.K. (1998): Modeltest: testing the model of DNA substition. Bioin- formatics 14: 817-818. Puente, M., Thomas, M., Vences, M. (2005): Phylogeny and biogeography of Malagasy dwarf geckos, Lygodactylus Gray, 1864: preliminary data from mitochondrial DNA sequences (Squamata: Gekkonidae). In: African Biodiversity – Molecules, Organ- isms, Ecosystems. p. 229-235. Huber, B.A., Lampe, K.H., Eds, Springer, Berlin. Puente, M., Glaw, F., Vieites, D.R., Vences, M. (2009): Review of the systematics, morphol- ogy and distribution of Malagasy dwarf geckos, genera Lygodactylus and Micros- calabotes (Squamata: Gekkonidae). Zootaxa 2103: 1-76. Rocha, S., Vences, M., Glaw, F., Posada, D. Harris, D.J. (2009): Multigene phylogeny of Malagasy day geckos of the genus Phelsuma. Mol. Phylogenet. Evol. 52: 530-537. Röll, B. (2000): Two sympatric Lygodactylus species in coastal areas of Eastern Africa (Reptilia, Gekkonidae). Bonn. Zool. Monogr. 46: 189-198. Röll, B. (2003): Delightful dwarfs: diurnal geckos of the genus Lygodactylus. Gekko 3: 21-33. 45DNA taxonomy in Lygodactylus Röll, B. (2004): Lygodactylus luteopicturatus Pasteur, 1965 [1964]: ein Synonym von Lygo- dactylus picturatus (Peters, 1870) (Sauria, Gekkonidae). Sauria, Berlin, 25: 33-37. Röll, B., Pröhl, H., Hoffmann, K.P. (2010): Multigene phylogenetic analysis of Lygodacty- lus dwarf geckos (Squamata: Gekkonidae). Mol. Phylogenet. Evol. 56: 327-335. Spawls, S., Howell, K., Drewes, R., Ashe, J. (2002): A Field Guide to Reptiles of East Afri- ca. Academic Press, San Diego. 543pp. Swofford, D. L. (2000): PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Vences, M., Thomas, M., Bonett R.M., Vieites D.R. (2005): Deciphering amphibian diver- sity through DNA barcoding: chances and challenges. Phil. Trans. R. Soc. B 360: 1859-1868. bbib67 OLE_LINK1 OLE_LINK2 bbib28 OLE_LINK5 OLE_LINK6 OLE_LINK7 OLE_LINK8 _GoBack OLE_LINK1 OLE_LINK2 OLE_LINK3 OLE_LINK4 OLE_LINK1 OLE_LINK2 OLE_LINK19 OLE_LINK20 OLE_LINK21 OLE_LINK29 OLE_LINK3 OLE_LINK4 OLE_LINK5 OLE_LINK31 OLE_LINK14 OLE_LINK15 OLE_LINK12 OLE_LINK13 OLE_LINK16 OLE_LINK17 OLE_LINK22 OLE_LINK23 OLE_LINK24 OLE_LINK8 OLE_LINK9 OLE_LINK10 OLE_LINK11 OLE_LINK18 OLE_LINK27 OLE_LINK28 OLE_LINK25 OLE_LINK26 OLE_LINK6 OLE_LINK7 OLE_LINK34 OLE_LINK37 OLE_LINK38 Acta Herpetologica Vol. 6, n. 1 - June 2011 Firenze University Press Widespread bacterial infection affecting Rana temporaria tadpoles in mountain areas Rocco Tiberti Extreme feeding behaviours in the Italian wall lizard, Podarcis siculus Massimo Capula1, Gaetano Aloise2 Lissotriton vulgaris paedomorphs in south-western Romania: a consequence of a human modified habitat? Severus D. Covaciu-Marcov*, Istvan Sas, Alfred Ş. Cicort-Lucaciu, Horia V. Bogdan Body size and reproductive characteristics of paedomorphic and metamorphic individuals of the northern banded newt (Ommatotriton ophryticus) Eyup Başkale1, Ferah Sayım2 , Uğur Kaya2 Genetic characterization of over hundred years old Caretta caretta specimens from Italian and Maltese museums Luisa Garofalo1, John J. Borg2, Rossella Carlini3, Luca Mizzan4, Nicola Novarini4, Giovanni Scillitani5, Andrea Novelletto1 The phylogenetic position of Lygodactylus angularis and the utility of using the 16S rDNA gene for delimiting species in Lygodactylus (Squamata, Gekkonidae) Riccardo Castiglia*, Flavia Annesi Localization of glucagon and insulin cells and its variation with respect to physiological events in Eutropis carinata Vidya. R. Chandavar1, Prakash. R. Naik2* The Balearic herpetofauna: a species update and a review on the evidence Samuel Pinya1, Miguel A. Carretero2 Effects of mosquitofish (Gambusia affinis) cues on wood frog (Lithobates sylvaticus) tadpole activity Katherine F. Buttermore, Paige N. Litkenhaus, Danielle C. Torpey, Geoffrey R. Smith*, Jessica E. Rettig Food composition of Uludağ frog, Rana macrocnemis Boulenger, 1885 in Uludağ (Bursa, Turkey) Kerim Çiçek Preliminary results on tail energetics in the Moorish gecko, Tarentola mauritanica Tommaso Cencetti1,2, Piera Poli3, Marcello Mele3, Marco A.L. Zuffi1 Climate change and peripheral populations: predictions for a relict Mediterranean viper José C. Brito1, Soumia Fahd 2, Fernando Martínez-Freiría1, Pedro Tarroso1, Said Larbes3, Juan M. Pleguezuelos4, Xavier Santos5 Assessing the status of amphibian breeding sites in Italy: a national survey Societas Herpetologica Italica* Osservatorio Erpetologico Italiano ACTA HERPETOLOGICA Journal of the Societas Herpetologica Italica ACTA HERPETOLOGICA Rivista della Societas Herpetologica Italica