Zootaxa 3745 (4): 449–468 www.mapress.com /zootaxa / Copyright © 2013 Magnolia Press
Article
ISSN 1175-5326 (print edition)
ZOOTAXA
ISSN 1175-5334 (online edition)
http://dx.doi.org/10.11646/zootaxa.3745.4.3 http://zoobank.org/urn:lsid:zoobank.org:pub:C3C076EF-1E00-49CF-B9DA-4BACAF670016
New species of Stenodactylus (Squamata: Gekkonidae) from the Sharqiyah Sands in northeastern Oman MARGARITA METALLINOU & SALVADOR CARRANZA* Institute of Evolutionary Biology (CSIC-Universitat Pompeu Fabra). Passeig Marítim de la Barceloneta 37-49, E-08003 Barcelona, Spain. * Corresponding author:
[email protected]
ABSTRACT A new species of gecko of the genus Stenodactylus (Squamata: Gekkonidae) is described from the dune desert of Al Sharqiyah Sands in northeastern Oman. Stenodactylus sharqiyahensis sp. nov. is characterized morphologically by its small size, snout shape, webbing between fingers not very extended, relatively short limbs, and scalation. It is genetically distinct in the mitochondrial DNA and the nuclear MC1R gene from Stenodactylus arabicus to which it has previously been referred. The new species seems to have a restricted distribution confined to the Sharqiyah Sands, which remain isolated from other sand deserts in Arabia. In addition, the data presented herein confirm new locality records for Stenodactylus arabicus in the easternmost limit of its distribution range in western central Oman. Key words: gecko, Arabia, deserts, phylogeny, taxonomy, systematics, evolution, mitochondrial DNA, MC1R.
INTRODUCTION As a result of its complex geological history, geographical position and the great diversity of landscapes, Oman harbors a high diversity of plant and animal species (Arnold 1977, 1980a; and other articles in the same volumes). Ever since the first biodiversity studies carried out in the country in the 1970’s, reptiles have attracted considerable attention and interest, which was soon reflected in numerous scientific publications, including the description of several new species (e.g. Arnold 1972, 1977, 1980a, 1986a; Arnold & Gardner 1994; Carranza & Arnold 2012; Gardner 1994; among others). However, with few exceptions (Gallagher & Arnold 1988), these studies were mainly centered in the two main mountain ranges of Oman: the Hajar Mountains in the North, and the Dhofar Mountains in the South (Arnold 1972, 1980a, 1986b; Arnold & Gallagher 1977; Arnold & Gardner 1994; Gardner 1994), leaving areas in between largely unexplored. Yet, it is especially in these desert and arid areas where several reptile species thrive (Arnold 1986b). Gekkonid lizards of the genus Stenodactylus Fitzinger, 1826 are among the most conspicuous faunal elements of the arid areas of Arabia and North Africa (Arnold 1980b). The genus comprises twelve strictly nocturnal species, highly specialized to arid and hyper-arid habitats, eight of which inhabit the Arabian Peninsula and areas immediately adjacent to the North. This important diversity contrasts with the only four Stenodactylus species found in the much larger area of North Africa (Baha El Din 2006; Metallinou et al. 2012). A recent phylogenetic study with complete taxonomic sampling indicated that the genus most probably originated in Arabia (Metallinou et al. 2012). Although most of the Stenodactylus species were described before the 1960’s and a morphological revision of the group was carried out in 1980 (Arnold 1980b), in recent years there have been numerous studies based both on morphological and molecular data, some of which have also proposed taxonomical changes (Baha El Din 2006; Fujita & Papenfuss 2011; Kratochvil et al. 2001; Metallinou et al. 2012; Metallinou & Crochet 2013). The most relevant of these, for the Arabian species, has been the allocation of Stenodactylus khobarensis (Haas, 1957) to the resurrected genus Pseudoceramodactylus Haas, 1957. Among the Arabian species, Stenodactylus grandiceps Haas, 1952 is found in the northern parts of the Arabian Peninsula, S. leptocosymbotes Leviton and Anderson, 1967 is Accepted by U. Fritz: 4 Dec. 2013; published: 6 Dec. 2013
449
restricted to the southeastern areas in Arabia and S. pulcher Anderson, 1896 and S. yemenensis Arnold, 1980b are endemic to Yemen and bordering parts of Saudi Arabia. Three species, S. doriae (Blanford, 1874), S. slevini Haas, 1957 and S. arabicus (Haas, 1957), have extensive distribution ranges across large regions of the Arabian Peninsula (Arnold 1980b; Sindaco & Jeremcenko 2008). Stenodactylus species that share great similarities in size and niche preference, like for example S. grandiceps, S. slevini, S. leptocosymbotes and S. yemenensis, all found on hard sand ground, are roughly parapatric and do not coexist (Arnold, 1980b). In other cases, however, up to three different species can be observed in the same locality, where they occupy different microhabitats: S. leptocosymbotes on hard sand, and S. doriae, S. arabicus on soft sand, with the latter being much smaller in size (Arnold 1980b; Fujita & Papenfuss 2011; pers. obs.). Stenodactylus arabicus is the smallest species of the genus (Arnold 1980b). It can be found on sandy substrate ranging from the northern edge of the Arabian Gulf, to central Arabian Peninsula to the west and all the way down to the Arabian Sea. The vast area of Rub al Khali, the world’s largest sand desert (Vincent 2008), provides an excellent terrain for S. arabicus. However, due to its extremely harsh conditions, very little data on species distributions is available from this region, creating large discontinuities in the ranges of several widely distributed Arabian species. Nevertheless, recent genetic analyses showed little divergence between S. arabicus samples from almost the two extremes of its known distribution, Kuwait and southern Oman (Metallinou et al. 2012). On the other hand, a specimen from northeastern Oman analyzed by Fujita and Papenfuss (2011) was found to be genetically very distinct from another from northern United Arab Emirates (UAE). Metallinou et al. (2012) used samples from six different localities from across the species’ range to show that S. arabicus harbors two deep reciprocally monophyletic lineages, one of them being restricted to the Sharqiyah Sands in northeastern Oman, referred to as S. cf. arabicus. In this study, we use morphological as well as mitochondrial and nuclear DNA data from a comprehensive collection of tissue samples and specimens from all over Oman (where the new lineage was found) and other parts of Arabia, to further explore the geographic distribution and systematics of S. arabicus and S. cf. arabicus. The results of these studies confirm the distinctiveness of this latter form, which is herein described as a new Stenodactylus species, endemic to the Sharqiyah Sands in northeastern Oman.
MATERIAL AND METHODS Molecular analyses Molecular samples, DNA extraction and amplification A total of 24 individuals of Stenodactylus arabicus and 27 of the new species described herein were included in the molecular study. Stenodactylus pulcher was used to root the tree based on published evidence (Metallinou et al. 2012). A list of all individuals included in the molecular analyses with their taxonomic identifications, DNA, sample, voucher codes, voucher references, corresponding geographical distribution data and GenBank accession numbers for all sequenced genes is presented in Appendix I. A map indicating the geographical distribution of all samples used in the genetic study is shown in Fig. 1. Genomic DNA was extracted from ethanol-preserved tissue samples using the Qiagen DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA, USA) or the SpeedTools Tissue DNA Extraction kit (Biotools, Madrid, Spain). Three genetic markers were PCR-amplified and sequenced: two mitochondrial fragments of the genes encoding the ribosomal 12S rRNA (12S) and 16S rRNA (16S), respectively, and part of the nuclear melano-cortin 1 receptor (MC1R) gene. Primers, PCR conditions and source references for the amplification of the two mitochondrial fragments are listed in Metallinou et al. (2012), and for the MC1R in Smid et al. (2013). Sequence analysis Geneious v. R6 (Biomatters Ltd.) was used for assembling and editing the chromatographs. Heterozygous positions for the MC1R nuclear coding gene fragment were identified based on the presence of two peaks of approximately equal height at a single nucleotide site in both strands. The same fragment was translated into amino acids and no stop codons were observed. DNA sequences were aligned using MAFFT v.6 (Katoh & Toh, 2008) applying parameters by default (Auto strategy, Gap opening penalty: 1.53, Offset value: 0.0). Phased sequences of the MC1R
450 · Zootaxa 3745 (4) © 2013 Magnolia Press
METALLINOU & CARRANZA
fragment were used for the network analyses. SEQPHASE (Flot 2010) was used to convert the input files, and the software PHASE v. 2.1.1 to resolve phased haplotypes (Stephens et al. 2001). Default settings of PHASE were used except for phase probabilities that were set as ≥ 0.7 (see Harrigan et al. 2008). Uncorrected mean genetic distances between and within groups for the mitochondrial gene fragments were calculated using MEGA 5 (Tamura et al. 2011), using the p-distance method.
FIGURE 1. Map of localities of examined material in this study. Circles indicate samples used in the genetic analyses, squares indicate samples used in morphological analyses and diamonds indicate specimens examined through photographic material from museum collections. Arrows indicate type localities of S. arabicus and S. sharqiyahensis sp. nov. Exact localities of all specimens can be found in Appendices I and II.
Phylogenetic and network analyses Best-fitting models were inferred for each gene independently using jModeltest v.0.1.1 (Posada 2008) under the Akaike information criterion (AIC) (Akaike 1973). These were GTR+G for the 12S, GTR+I+G for the 16S and HKY for the unphased MC1R. Phylogenetic analyses were performed using Maximum Likelihood (ML) and Bayesian (BI) methods. Maximum Likelihood analyses were performed in RAxML v.7.0.3 (Stamatakis 2006). A GTR+G model was used with parameters estimated independently for each partition. Reliability of the ML tree was assessed by bootstrap analysis (Felsenstein 1985) including 1000 replications. Bayesian analyses were performed with BEAST v.1.6.1 (Drummond & Rambaut 2007) using the same dataset used in the ML analysis but without the outgroup. Analyses were run three times for 5x107 generations with a sampling frequency of 5000. Models and prior specifications applied were as follows (otherwise by default): model of sequence evolution for each independent partition as above; coalescent constant size process of speciation; random starting tree; alpha Uniform (0, 10); fix mean rate of molecular clock model to 1. The xml file was manually modified to use Ambiguities=“TRUE” for the MC1R partition to account for variability in the heterozygous positions, instead of treating them as missing data. Nodes were considered strongly supported if they received ML bootstrap values ≥ 70% and posterior probability (pp) support values ≥ 0.95 (Huelsenbeck & Rannala 2004; Wilcox et al. 2002).
NEW STENODACTYLUS SPECIES FROM OMAN
Zootaxa 3745 (4) © 2013 Magnolia Press ·
451
Genealogical relationships between taxa in the MC1R nuclear gene were assessed with haplotype networks constructed using statistical parsimony as implemented in the program TCS v.1.21 (Clement et al. 2000). Phased sequences were used (see above), all 96 of which were full-length, and a connection limit of 95% was applied. Morphological analyses Morphological samples, variables and museums acronyms Fifty-three alcohol-preserved adult specimens were included in the morphological analyses. A list of all studied specimens with their Museum accession codes and corresponding sample code, country, and geographical distribution data is presented in Appendix I. In addition, 96 specimens from the Natural History Museum, London and the California Academy of Sciences, including the holotype of Stenodactylus arabicus (CAS84321), were photographed but not measured, and therefore were only used for clarifying the taxonomic identity of these museum specimens and also to have a better idea of the geographical distribution of the two species in Arabia (Appendix II). Specimens included in the morphological study (Appendixes I and II) are mapped in Fig. 1. The material studied was obtained from the following institutions: Natural History Museum, London, UK (BMNH), Museo Civico di Storia Naturale di Carmagnola, Turin, Italy (MCCI), National Museum in Prague, Prague, Czech Republic (NMP), Oman Natural History Museum, Muscat, Oman (ONHM), California Academy of Sciences, California, San Francisco, USA (CAS), and S. Carranza’s field series temporarily housed at the Institute of Evolutionary Biology (IBE), Barcelona, Spain. Variables for the morphological analyses were selected based on previous taxonomic studies of Stenodactylus (Arnold 1980b; Kratochvil et al. 2001), and on our personal observations. The following measurements were taken by the same person (M.M.) on the right side of each specimen using a digital caliper with accuracy to the nearest 0.01 mm and were expressed in millimeters: snoutvent length (SVL), measured from tip of snout to vent; head-neck length (HNL), measured from tip of snout to posterior insertion of forelimb; head length (HL), measured from tip of snout to base of the anterior ear border; snout length (SL), measured from tip of snout to anteriormost point of eye; snout height (SH), measured at anteriormost point of eye; distance between eyes (EW), measured at occipital region, excluding palpebral fold; maximum head width (HW), measured at its widest part, usually at the level of temporal region; transverse eye diameter (ED); forearm length (FL), from wrist to elbow; arm length (AL), from elbow to the insertion of the forelimb on the posterior side; tibia length (BL), measured from ankle to knee; femur length (ML), measured from knee to the insertion of the hind limb on the posterior side; tail length (TL), from vent to tip of tail; tail width (TW), measured at half tail length. In addition to the morphometric continuous variables, the following meristic (pholidotic) and categorical variables were collected by the same person (M.M.) using a dissecting microscope. Meristic variables: number of upper labial scales (ULS); number of lower labial scales (LLS); number of scales between supranasals (SS); number of tubercles on first (ventral) row of cloacal tubercles (CT1); number of tubercles on second row of cloacal tubercles (CT2). Categorical variables: protrusion of the perinarial region (PP) 1: nostrils not protuberant, 2: weakly protuberant, 3: moderately, 4: strongly; snout shape on lateral view (SP) - 1: convex, 2: straight, 3: concave; 4: mixed (convex from anteriormost point of eye to mid-snout switching to concave from mid-snout to tip of snout); presence of extended webbing between fingers (WB) - 1: absence, 2: presence. Multivariate analyses Statistical analyses were used to investigate if there were differences in size and shape between Stenodactylus arabicus and the new species described herein. The 14 morphometric continuous variables, the five meristic and the three categorical were analyzed independently. Specimens IBECN135, IBECN3490, IBECN7228, NMP6V 74953, MCCI-R1809, IBECN4366, IBECN2615, MCCI-R1793(1) were not included in the multivariate analyses of the continuous variables as a result of missing data for some of the characters. After removing these specimens, the dataset of continuous variables included 45 specimens in total, 23 of which (11 males and 12 females) corresponded to S. arabicus, and 22 (12 males and 10 females) to the new species described herein. For the analyses of the shape and meristic variables, all 53 specimens were included (Appendix I), of which 26 corresponded to S. arabicus (12 males and 14 females) and 27 (12 males and 15 females) to the new species. Morphometric and meristic variables were log 10-transformed before the analyses and tested for normality (Shaphiro-Wilk’s test) and homogeneity of variances (Levene’s test). Statistical analyses of the morphological data included descriptive statistics, Scheffe’s post-hoc tests and Fisher exact probability test. Taxonomic and sexual
452 · Zootaxa 3745 (4) © 2013 Magnolia Press
METALLINOU & CARRANZA
differences in the morphometric and meristic variables were tested with MANCOVA and MANOVA, respectively. All analyses were performed in Statistica v. 5.4 (Stat-Soft, Inc., Tulsa, OK, USA), except Fisher exact probability tests for 2x3 and 2x4 contingency tables that was performed using the web application VassarStats (www.vassarstats.net).
FIGURE 2. ML tree of S. arabicus and S. sharqiyahensis sp. nov. inferred using 12S, 16S mtDNA and MC1R nuclear gene fragments. Bootstrap values ≥70% are shown next to the nodes. Black circles on nodes indicate posterior probability values ≥0.95. The tree was rooted using S. pulcher. Information on the samples included is given in Appendix I.
NEW STENODACTYLUS SPECIES FROM OMAN
Zootaxa 3745 (4) © 2013 Magnolia Press ·
453
RESULTS Molecular analyses The dataset used for the phylogenetic analyses consisted of an alignment of 1592 base pairs (bp) of concatenated mitochondrial and unphased nuclear DNA for 52 Stenodactylus, including 24 Stenodactylus arabicus, 27 specimens of the new species described herein, and one S. pulcher that was used as outgroup. The concatenated alignment included 384 bp of 12S (48 variable), 540 bp of 16S (54 variable), and 668 bp of MC1R (4 variable and 20 heterozygous). The results of the ML and BI analyses are presented in Fig. 2. The two trees were almost identical, differing only in the position of the clade comprising samples M34 and M35 (within the main clade of S. arabicus in the ML tree, or sister to all the other S. arabicus in the BI tree). The two species are reciprocally monophyletic and well supported by high bootstrap and pp values (Fig. 2). The uncorrected genetic distances (pdistances) between them are 8.1 ±1.3% for the 12S and 4.6 ±0.8% for the 16S. The level of genetic variability within S. arabicus is 1.6 ±0.4% for the 12S and 1.65 ±0.31% for the 16S, and within the new species described herein 0.6 ±0.4% for the 12S and 0.4 ±0.13% for the 16S. The results of the haplotype network analysis are presented in Fig. 3 and clearly show that all 13 MC1R haplotypes of S. arabicus are private as well as all 11 haplotypes of the new species.
FIGURE 3. Haplotype network of the phased sequences of nuclear marker MC1R. Phase probabilities were set as ≥0.7. Information on the samples included is shown in Appendix I.
Multivariate analyses of the morphological data Body shape differences were tested using a two-way MANCOVA (sex/taxa interactions, SVL as a covariate for size correction) including all morphometric variables. The results of the MANCOVA showed that, while differences between sexes for the 13 variables were not significant (MANCOVA: Wilks Lambda=0.708; d.f.=13, 28; P=0.575), there were significant differences between the two taxa (MANCOVA: Wilks Lambda=0.166; d.f.=13, 28; P
