Abstract
The North African fire salamander, Salamandra algira, is distributed in Algeria, Morocco and Ceuta (Spanish territory located on the north coast of Africa), but until now rather limited information has been available on the populations across the Algerian part of its range. We here provide a first analysis of the phylogeography of this species in Algeria, based on 44 samples from populations distributed across 15 localities in Central Algeria. We sequenced three segments of mitochondrial DNA, covering 12S rRNA, cytochrome b (Cytb) and the D-loop. The mtDNA sequences of the Algerian samples were strongly different from the Moroccan populations occurring west of the Moulouya River (corresponding to the subspecies S. a. tingitana and S. a. splendens) but sister to the genetically rather similar population from the Beni Snassen Massif in eastern Morocco (subspecies S. algira spelaea). Among the Algerian specimens studied, those from the westernmost site, Chrea Massif, were the sister clade to the remaining populations, and the overall genetic divergence was low, with a maximum of five mutational steps in a 295 bp fragment of cytochrome b.
Introduction
Salamandra Garsault, 1764 is a genus of terrestrial salamanders widely distributed across the western Palearctic, extending from the Iberian Peninsula and North Africa to Asia Minor (Veith, 1994). The African Salamandra algira Bedriaga, 1883 is confined to the humid and sub-humid forests of Algeria, Morocco and Ceuta (Spain) (Bons and Geniez, 1996; Schleich, Kästle and Kabisch, 1996; Steinfartz, Veith and Tautz, 2000; Martínez-Medina, 2001; Donaire-Barroso and Bogaerts, 2003; Escoriza et al., 2006; Beukema et al., 2010; Velo-Antón et al., 2014). Its presence in Tunisia (Salvador, 1996) is currently not confirmed (Bogaerts et al., 2013).
Salamandra algira is the sister species of the widespread European S. salamandra (Vences et al., 2014) and was long seen as its subspecies. Numerous publications have mainly provided data on the distribution, behavior and reproduction of S. algira in Morocco, leading to taxonomic revisions based on genetic and morphological characters (Donaire-Barroso and Bogaerts, 2001, 2003; Donaire-Barroso, Bogaerts and Herbert, 2001; Martínez-Medina, 2001; Bogaerts and Donaire-Barroso, 2003; Bogaerts et al., 2007; Beukema et al., 2010; Ben Hassine and Escoriza, 2014; Velo-Antón et al., 2014). Currently from the Moroccan range of the species three subspecies are defined: Salamandra algira tingitana Donaire and Bogaerts, 2003 from northern Morocco; S. a. splendens Beukema et al., 2013 from central Morocco, i.e., central and western Rif Mountains and north-eastern part of the Middle Atlas; S. a. spelaea Escoriza and Comas, 2007 from the Beni Snassen Massif in easternmost Morocco. Since the type locality of the species is in eastern Algeria (Mont Edough near Bône = Annaba; based on neotype designation by Eiselt [1958]), all Algerian populations are usually considered to belong to the nominal subspecies, S. a. algira (Bedriaga, 1883).
However, S. algira remains poorly studied in Algeria where it is known from various mountainous locations not far from the Mediterranean coast (Annaba, Kabylia, Blida Atlas and Oranie; Boulenger, 1891; Doumergue, 1901; Bons, 1972; Veith, 1994). Although a few genetic studies included Algerian samples from Blida Atlas and Kabylia (Gasser, 1978) and Annaba (Joger and Steinfartz, 1994; Steinfartz, Veith and Tautz, 2000) almost no data on the phylogeography, ecology, reproduction, and morphological variation are known from the Algerian part of the species distribution range (Bogaerts and Donaire-Barroso, 2003; Escoriza et al., 2006). Herein, our objective is to provide a first exploration of phylogeographic relationships of S. algira in Algeria based on mitochondrial DNA sequences obtained from new samples collected at 15 distinct sites.
Material and methods
Fieldwork and sampling
From January 2013 until June 2014 we conducted fieldwork in Kabylia, northeastern Algeria (fig. 1; 37.18°N-36.13°N/5.38°E-3.22°E), mainly in the humid periods from early October to late March, given that S. algira reproduces in autumn-winter (October to March) in the Maghreb (Escoriza and Ben Hassine, 2014a). We searched during daytime for larvae in suitable water bodies. Coordinates were recorded with GPS (Global Positioning System with an accuracy of 10 m; table 1) and tissue samples (toe clips of adults or fin clips of larvae) were collected and preserved in 99% ethanol.
Our molecular analysis included 44 newly collected Algerian samples from 15 representative locations, mostly from Kabylia region covering a wide Algerian distribution range of the species except the easternmost (Edough Massif) and westernmost site (Remchi, Mont des Traras) in the country, and combined these with sequences of the Moroccan subspecies retrieved from Genbank (see below).
DNA extraction, PCR amplification and sequencing
Total genomic DNA was extracted using proteinase K (10 mg/ml) digestion followed by a standard salt extraction protocol (Bruford et al., 1992). Standard polymerase chain reactions were performed in a final volume of 12.5 μl and using 0.3 μl each of 10 μM primer, 0.25 μl of total dNTP 10 mM, 0.1 μl of 5 u/μl GoTaq and 2.5 μl 5× GoTaq Reaction Buffer and 8.05 μl of distilled water. We PCR-amplified three segments of mitochondrial DNA, i.e., representing parts of the 12S rRNA and Cytochrome b (Cytb) genes, and of the D-loop (supplementary table S3). The successfully amplified products were purified using Exonuclease I and Shrimp Alkaline Phosphatase (SAP) or Antarctic phosphatase (AP) according to the manufacturer’s instructions (NEB) in order to inactivate remaining primers and dNTPs. Purified PCR templates were cycle-sequenced using Big-Dye Ready Reaction with the amplification primers on an AB 3130 automated DNA sequencer (Applied Biosystems). 12S rRNA and Cytb were sequenced with the forward primers only, while the D-loop segment was sequenced in both directions. Chromatograms were checked and sequences were manually corrected where necessary, and assembled using CodonCode Aligner V.3.5.6 (Codoncode Corporation) computer software. The newly identified sequences were submitted to Genbank (accession numbers KT335613-KT335714; also newly submitted and partly used were sequences from Steinfartz et al. [2000], KT335852-KT335946; details in supplementary table S2).
Phylogenetic analysis
Sequences obtained in this study were combined with those obtained in previous works from Moroccan populations (Steinfartz, Veith and Tautz, 2000; Escoriza et al., 2006; Escoriza and Comas, 2007; Beukema et al., 2010; Vences et al., 2014), as retrieved from Genbank (supplementary table S2).
Map of sampling locations of Salamandra algira used in this study. Different symbols and colors denote the four subspecies currently recognized. The Moulouya River, coinciding with the deepest phylogeographic split within S. algira, is shown in light blue. The inset haplotype network is based on DNA sequences of a segment of the Cytb gene (295 bp) of Algerian samples of Salamandra algira, plus S. a. spelaea. H1-H8 are haplotype numbers. Size of circles is proportional to the number of samples found to be represented by a certain haplotype (indicated by scale). The map indicates for the Algerian populations the population number (in bold; corresponding to table 1) and the haplotype number present (in italics). This figure is published in colour in the online version.
Citation: Amphibia-Reptilia 37, 1 (2016) ; 10.1163/15685381-00003025
Because not all DNA segments could be amplified from all specimens, we followed the following strategy for analysis:
(i) To avoid possible artefacts in tree reconstruction as they can be caused by missing data, we restricted the phylogenetic analysis to those samples for which both 12S and Cytb sequences were available, resulting in a data matrix with missing data only for D-loop. The sole exception was the sample from Seraidi (Edough Massif – Annaba), the easternmost location and type locality of S. algira; for this sample, only D-loop was available and it was included despite missing 12S and Cytb due to its taxonomic and biogeographic importance. The phylogenetic dataset consequently included 30 samples from 15 Algerian populations, 36 Moroccan samples, and 10 hierarchical outgroups of other Salamandra species (supplementary table S2).
(ii) We also carried out an analysis including all S. algira samples for which at least one of the three DNA segments was sequenced; as expected, this tree showed rather disparate branch lengths for several samples, but overall resulted in a similar topology (supplementary fig. S2).
Geographical coordinates (Lat, Long) of sampling localities for Salamandra algira algira and S. a. spelaea for which DNA sequences were used in the phylogenetic analyses. See online supplementary materials for sampling localities of the remaining Moroccan subspecies, S. a. tingitana and S. a. splendens, and for DNA accession numbers. Samples from Algeria (except for the locality Annaba) were all newly collected and sequenced for this study; coordinates refer to own GPS readings. Moroccan sequences were all taken from Genbank. Cytb haplotypes (HT) are given according to their numbers in the inset haplotype network (fig. 1), with the numbers of occurrences in parentheses.
(iii) In order to understand in more detail the variation among Algerian populations of S. algira, we constructed a haplotype network based on sequences of Cytb. Because our focus was on the variation within and among the Algerian populations, we restricted the network to include only samples of Algerian populations, and the Moroccan subspecies spelaea which geographically and phylogenetically is closest to the Algerian populations.
Sequences were aligned with the MUSCLE algorithm under default settings implemented in MEGA5 software (Tamura et al., 2011) and alignments adjusted by eye where necessary. Alignment was unambiguous; only single indels were present to align outgroups, and no indels were present in the ingroup of Algerian samples, for 12S and D-loop. PartitionFinder 1.0.1 software (Lanfear et al., 2012) was used to infer the best-fitting models of molecular evolution and partition schemes by the AICc for each DNA segment (supplementary table S4). The best-fitting partition/substitution model scheme was implemented in a Bayesian Inference (BI) phylogenetic analysis on MrBayes 3.2 (Ronquist et al., 2012), for the concatenated data set. MrBayes was run for 20 million generations, with three heated and one cold Markov chains, sampling every 1000 generations, and checking for parameter convergence using Tracer v. 1.5 (http://beast.bio.ed.ac.uk/Tracer). The trees corresponding to 25% of the generations were discarded as burn-in and a majority-rule consensus tree with the corresponding posterior probabilities of nodes was calculated.
Median-joining haplotype networks were constructed for Cytb on Network 4.6.1.2 (Fluxus Technology Limited 2012; Bandelt et al., 1999). For this analysis all sequences were cut to remove sections of missing data at the end or beginning.
Majority-rule consensus tree from a Bayesian Inference analysis of Salamandra algira sequences of segments of the mitochondrial 12S and Cytb genes, and the D-loop (length of total concatenated alignment: 1683 bp). The tree was rooted with S. infraimmaculata, with various subspecies of S. salamandra as hierarchical outgroups (removed for graphical reasons in order to avoid excessively long branches; see supplementary documents for full tree including outgroups). Numbers at nodes are posterior probabilities (only values of 0.90 or higher are shown). Terminals were included according to strategy 1 as specified in Material and methods, i.e., for all samples, 12S and Cytb sequences were available, except for the sample from Annaba (D-loop only). This figure is published in colour in the online version.
Citation: Amphibia-Reptilia 37, 1 (2016) ; 10.1163/15685381-00003025
Results
The Bayesian Inference analysis based on the concatenated alignment of the three DNA segments produced a largely resolved tree (fig. 2), in which numerous nodes received high support from posterior probability values. Of the total 1683 characters of the concatenated alignment, 1462 were invariable and 132 were parsimony-informative. As in previous studies, Salamandra algira was resolved as monophyletic with maximum support (posterior probability PP = 1.0), and within the species, two main clades were apparent (both supported by PP = 1.0): a first clade containing the Moroccan populations west of the Moulouya River valley, corresponding to the reciprocally monophyletic subspecies tingitana and splendens; and a second clade containing all populations east of the Moulouya River valley.
Within this second clade, eastern Moroccan S. a. spelaea was the sister group of all the Algerian samples of S. a. algira (supported by PP = 1.0; support for the spelaea and algira subclades, PP = 1 and PP = 0.97 respectively). In the Algerian clade the samples from the westernmost sampled locality Chrea branched off first. The remaining samples formed a clade sister to the Chrea samples (PP = 0.99), and did not show a clear and strongly supported phylogeographic structure. The easternmost population Annaba, represented by only one D-loop sequence, stood out by a long branch but clustered in between the remaining Algerian samples.
A haplotype network based on 295 bp Cytb sequences confirmed a low variation among the Algerian samples included in this study. A total of seven haplotypes were identified in Algeria which differed from each other by one or two mutations (fig. 1). The haplotype found in S. a. spelaea from the Moroccan locality Beni Snassen (haplotype 8) differed by a minimum of four mutational steps from the Algerian haplotypes. It was connected to haplotypes occurring in Chrea (H7) and Tifiras Skikda, although the latter sample was not placed basal in the tree based on the combined data set (fig. 2). Several of the haplotypes were found co-occurring at the same locations (table 1).
Discussion
Our new data obtained mostly from central Algeria, combined with published information allow for several conclusions on the status of Algerian populations of S. algira.
Our field surveys confirm the persistence of the species in the Kabylia region (localities 2-14 in table 1) and in the Chrea Massif (Blida), previously reported by Salvador (1996), Schleich, Kästle and Kabisch (1996), Bogaerts and Donaire-Barroso (2003), and Escoriza and Comas (2007). Several other historical records remain to be revisited in the future, such as Constantine in eastern Algeria (Guichenot, 1850), Algiers (Boulenger, 1882), the most western populations reported from Rahr-el-Maden near Oran (Doumergue, 1901), and the type locality Mont Edough near Annaba. Our sample from the locality Tifiras (wilaya of Skikda) confirms the presence of the species beyond its currently known range, further extending reports by Escoriza and Ben Hassine (2014b). This finding indicates that the distribution of S. algira is wider than historically suggested and might include a continuous area of occurrence stretching from Kabylia to Annaba.
Our results confirm that Algerian populations of Salamandra algira are genetically distinct from the Moroccan ones, a fact previously assessed on the basis of only single sequences from the easternmost locality Annaba (Steinfartz, Veith and Tautz, 2000; Donaire-Barroso and Bogaerts, 2003; Escoriza et al., 2006; Beukema et al., 2013). The available data also suggest, at least from a mitochondrial perspective, that the Algerian subspecies (S. a. algira) is the sister clade of the Beni Snassen population on the east side of the Moulouya River valley in Morocco.
Samples from Kabylia neither show deep divergences nor a clear phylogeographic structure which could indicate (i) this region being a transition area between eastern and western populations during Pleistocene climatic oscillations (e.g., Fritz et al., 2009) or (ii) a continuous connectivity among populations in this region allowed by a potential climatic stability. On the contrary, the westernmost new sampled locality Chrea, showed a more distinct differentiation and represents the sister clade to all other Algerian populations studied here. Yet, considering the entire Algerian range sampled herein (ca. 350 km distance between populations, not counting Annaba) the phylogeographic structure is much less pronounced than it is in Morocco, where in a comparable geographic extension two subspecies (tingitana and splendens) occur of which especially splendens in addition shows two distinct subclades.
For the easternmost Algerian population Jebel Edough (Annaba) represented by only one D-loop sequence, the available data do not support a deep divergence, despite this region (Edough Peninsula) is considered as fossil island and should contain a high level of endemism (e.g. for the newt Pleurodeles poireti; Carranza and Wade, 2004).
For the populations east of the Moulouya river valley, in light of the available data, the low genetic diversity and nested phylogenetic position of Algerian samples, could be explained by two complementary hypotheses: (i) a relatively recent stepwise eastwards colonization of Algeria, from the area currently occupied by the Beni Snassen population (subspecies spelaea), and (ii) a regular and ongoing gene flow among all populations. However, we cannot exclude that the pattern observed, and the suggested explanations, might be too simplistic and partly caused by incomplete sampling in west Algeria rather than geographical differentiation as discussed by Fritz et al. (2009) for tortoises.
Acknowledgements
We are grateful to Meike Kondermann and Gaby Keunecke for help during labwork. Analysis of Moroccan samples partly was supported by a joint research project led by the Natural History Museum of Braunschweig and the University of Marrakech (BMBF: 01DH13015); we are grateful to Tahar Slimani and El Hassan El Mouden for their collaboration.
References
Bandelt H.J., Forster P., Röhl A. (1999): Median-joining networks for inferring intraspecific phylogenies. Mol. Biol. Evol. 16: 37-48.
Bedriaga J. (1883): Beiträge zur Kenntniss der Amphibien und Reptilien der Fauna von Corsika. Arch. Naturgesch. 49: 124-273.
Ben Hassine J., Escoriza D. (2014): New ecological data on the family Salamandridae in the Maghreb. Herpetol. Rev. 45 (2): 1-5.
Beukema W., de Pous P., Donaire D., Escoriza D., Bogaerts S., Toxopeus A.G., De Bie C.A.J.M., Roca J., Carranza S. (2010): Biogeography and contemporary climatic differentation among Moroccan Salamandra algira. Biol. J. Linnean. Soc. 101: 626-641.
Beukema W., de Pous P., Donaire-Barroso D., Bogaerts S., García-Porta J., Escoriza D., Arribas O.J., El Mouden E.H., Carranza S. (2013): Review of the systematics, distribution, biogeography and natural history of Moroccan amphibians. Zootaxa 3661: 1-60.
Bogaerts S., Donaire-Barroso D. (2003): Sobre el politipismo en Salamandra algira Bedriaga, 1883. Bol. Asoc. Herpetol. Esp. 14: 1-2.
Bogaerts S., Valkenburg K., Donaire-Barroso D., Espallargas G. (2007): New localities of the north African fire salamander (Salamandra algira) at the southern limit of its distribution in Morocco. Z. Feldherpetol. 14: 238-241.
Bogaerts S., Donaire-Barroso D., Pasmans F., Böhme W., Carranza S. (2013): Do north African fire salamanders, Salamandra algira, occur in Tunisia? Herpetol. Notes 6: 301-306.
Bons J. (1972): Herpétologie Marocaine 1. Liste commentée des amphibiens et reptiles du Maroc. Bull. Soc. Sci. Nat. Maroc 52: 107-126.
Bons J., Geniez P. (1996): Amphibiens et reptiles du Maroc (Sahara Occidental compris), Atlas Biogéographique. Asociación Herpetológica Española, Barcelona.
Boulenger G.A. (1882): Catalogue of Batrachia Grandientia s. Caudata and Batrachia Apoda in the Collection of the British Museum. British Museum of Natural History, London.
Boulenger G.A. (1891): Catalogue of the reptiles and batrachians of Barbary (Morocco, Algeria, Tunisia) based chiefly upon the notes and collections made in 1880-1884 by M. Fernand Lataste. Trans. Zool. Soc. London 13: 93-164.
Bruford M.W., Hanotte O., Brookfield J.F.Y., Burke T. (1992): Single-locus and multilocus DNA fingerprint. In: Molecular Genetic Analysis of Populations: a Practical Approach, p. 225-270. Hoelzel A.R., Ed., IRL Press, Oxford.
Carranza S., Arnold E.N. (2004): History of west Mediterranean newts, Pleurodeles (Amphibia: Salamandridae), inferred from old and recent DNA sequences. Syst. Biol. Divers 1: 327-337.
Carranza S., Wade E. (2004): Taxonomic revision of Algero-Tunisian Pleurodeles (Caudata: Salamandridae) using molecular and morphological data. Revalidation of the taxon Pleurodeles nebulosus (Guichenot, 1850). Zootaxa 488: 1-24.
Donaire-Barroso D., Bogaerts S. (2001): Observations on viviparity of Salamandra algira in north Morocco. In: Herpetologia Candiana, p. 147-151. Lymberakis P., Valakos E., Pafilis P., Mylonas M., Eds, Natural History Museum of Crete, Irakleo.
Donaire-Barroso D., Bogaerts S. (2003): A new subspecies of Salamandra algira Bedriaga, 1883 from northern Morocco. Podarcis 4: 84-100.
Donaire-Barroso D., Bogaerts S., Herbert D. (2001): Confirmación de desarrollo larvario completo intrauterino en Salamandra algira (Bedriaga, 1883) del Noroeste de Marruecos. Butll. Soc. Catalana Herpetol. 5: 107-110.
Doumergue F. (1901): Essai sur la Faune Erpétologique de l’Oranie: Avec des Tableaux Analytiques et des Notions pour la Détermination de tous les Reptiles & Batraciens du Maroc, de l’Algérie et de la Tunisie. L. Fouce, Oran.
Eiselt J. (1958): Der Feuersalamander Salamandra salamandra (L). Beiträge zu einer taxonomischen Synthese. Abhandlungen und Berichte für Naturkunde, Museum für Naturkunde, Magdeburg 10: 77-154.
Escoriza D., Ben Hassine J. (2014a): Microclimatic variation in multiple Salamandra algira populations along an altitudinal gradient: phenology and reproductive strategies. Acta Herpetol. 9: 33-41.
Escoriza D., Ben Hassine J. (2014b): Salamandra algira (north African fire salamander): new distribution area in Algeria. Herpetol. Bull. 128: 24-25.
Escoriza D., Comas M.M. (2007): Description of a new subspecies of Salamandra algira (Bedriaga, 1883) (Amphibia: Salamandridae) from the Beni Snassen massif (northeast Morocco). Salamandra 43: 77-90.
Escoriza D., Comas M.M., Donaire D., Carranza S. (2006): Rediscovery of Salamandra algira (Bedriaga, 1883) from the Beni Snassen massif (Morocco) and phylogenetic relationships of north African Salamandra. Amphibia-Reptilia 27: 448-455.
Fritz U., Harris D.J., Fahd S., Rouag R., Martínez E.G., Casalduero A.G., Siroky P., Kalboussi M., Jdeidi T.B., Hundsdörfer A.K. (2009): Mitochondrial phylogeography of Testudo graeca in the western Mediterranean: old complex divergence in north Africa and recent arrival in Europe. Amphibia-Reptilia 30: 63-80.
Gasser F. (1978): Le polytypisme de l’espèce paléarctique Salamandra salamandra (Amphibiens Urodèles). II. Systématique, relations génétiques et tendances évolutives dans l’aire de répartition. Arch. Zol. exp. Gén. 119: 635-668.
Guichenot A. (1850): Histoire naturelle des reptiles et des poissons. Exploration scientifique de l’Algérie pendant les années 1840, 1841, 1842. Imprimerie Nationale, Paris.
Joger U., Steinfartz S. (1994): Electrophoretic investigations in the evolutionary history of the west Mediterranean Salamandra. Mertensiella 4: 241-254.
Lanfear R., Calcott B., Ho S.Y.W., Guindon S. (2012): PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Mol. Biol. Evol. 29: 1695-1701.
Martinez-Medina F.J. (2001): Nuevos registros de anfibios y reptiles (Amphibia, Reptilia) en la Sierra del Haus (Noroeste de Marruecos). Bol. Asoc. Herpetol. Esp. 12: 2-5.
Ronquist F., Teslenko M., van der Mark P., Ayres D.L., Darling A., Höhna S., Larget B., Liu L., Suchard M.A., Huelsenbeck J.P. (2012): MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst. Biol. 61: 539-542.
Salvador A. (1996): Amphibians of northwest Africa. Smithson. Herpetol. Inf. Serv. 109: 1-43.
Schleich H.H., Kästle W., Kabisch K. (1996): Amphibians and Reptiles of North Africa. Koeltz Scientific Books, Koenigstein, Germany.
Steinfartz S., Veith M., Tautz D. (2000): Mitochondrial sequence analysis of Salamandra taxa suggests old splits of major lineages and postglacial recolonizations of central Europe from distinct source populations of Salamandra salamandra. Mol. Ecol. 9: 397-410.
Tamura K., Peterson D., Peterson N., Stecher G., Nei M., Kumar S. (2011): MEGA5, molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28: 2731-2739.
Veith M. (1994): Morphological, molecular and life history variation in Salamandra salamandra (L.). Mertensiella 4: 355-397.
Velo-Antón G., García-Cardenete L., Cazalla F.J., Martínez-Freiría F. (2014): New record of Salamandra algira isolated on the north-western Tingitana peninsula, with some notes on the reproductive modes within the species. Bol. Asoc. Herpetol. Esp. 25: 1-3.
Vences M., Sanchez E., Hauswaldt J.S., Eikelmann D., Rodríguez A., Carranza S., Donaire D., Gehara M., Helfer V., Lötters S., Werner P., Schulz S., Steinfartz S. (2014): Nuclear and mitochondrial multilocus phylogeny and survey of alkaloid content in true salamanders of the genus Salamandra (Salamandridae). Mol. Phylogenet. Evol. 73: 208-216.
Footnotes
Associate Editor: Judit Vörös.
