Weak genetic divergence between the two South American toad-headed turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae)

Transcription

1 Amphibia-Reptilia 33 (2012): Weak genetic divergence between the two South American toad-headed turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) Mario Vargas-Ramírez 1,3,, Jan Michels 2, Olga Victoria Castaño-Mora 3, Gladys Cárdenas-Arevalo 3, Natalia Gallego-García 3, Uwe Fritz 1 Abstract. Mesoclemmys dahli and M. zuliae are two endangered, little-known toad-headed turtles with small distribution ranges in Colombia and Venezuela, respectively. Using the mitochondrial cytochrome b gene as a marker, we investigate their phylogeographic differentiation. Furthermore, based on 2341 bp of mtdna and 2109 bp of ndna of M. dahli, M. zuliae and allied chelid turtles, we infer their divergence time using a fossil-calibrated relaxed molecular clock approach. Mesoclemmys dahli and M. zuliae are closely related species, with an estimated mean divergence time of 10.6 million years. This estimate correlates with the uplift of the Serranía de Perijá, an Andean mountain chain separating their distribution ranges, suggesting that this event could have caused the evolution of the two species. Haplotype and nucleotide diversities of M. dahli are markedly higher than in Podocnemis lewyana, another endemic turtle species of Colombia. This pronounced dissimilarity may reflect differences in the phylogeographies and demographic histories of the two species, but also different habitat preferences. Keywords: Colombia, molecular clock, mtdna, ndna, phylogeography, Venezuela. Introduction While the phylogeographies of many Western Palaearctic and Nearctic turtles and tortoises are well-known (e.g. Walker et al., 1998; Lenk et al., 1999; Weisrock and Janzen, 2000; Karl and Wilson, 2001; Fritz et al., 2006, 2007, 2008, 2009; Rosenbaum, Robertson and Zamudio, 2007; Amato, Brooks and Fu, 2008; McGaugh, Eckerman and Janzen, 2008; Butler et al., 2011; Pedall et al., 2011; Ureña-Aranda and Espinosa de los Monteros, 2012), relatively few studies have focused on species from Central and South America so far (Souza et al., 2003; Pearse et al., 2006; Vargas-Ramírez et 1 - Museum of Zoology (Museum für Tierkunde), Senckenberg Dresden, A. B. Meyer Building, D Dresden, Germany 2 - Department of Functional Morphology and Biomechanics, Institute of Zoology, Christian-Albrechts- Universität zu Kiel, Am Botanischen Garten 1-9, D Kiel, Germany 3 - Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, Bogotá, Colombia Corresponding author; mario.vargas@senckenberg.de al., 2007, 2010a, 2012; González-Porter et al., 2011; Fritz et al., 2012a, 2012b). Among the South American species, members of the family Chelidae are the least studied ones, with just one publication on the phylogeography of Hydromedusa maximiliani (Souza et al., 2003). Chelidae, with 23 species, is the most species-rich of the seven families of turtles and tortoises in continental South America, and the most speciose group within this family are the toad-headed turtles of the genus Mesoclemmys. Currently, ten Mesoclemmys species are recognized, most of which occur in central South America (Bour and Zaher, 2005; Fritz and Havaš, 2007; van Dijk et al., 2011). Mesoclemmys dahli (Zangerl and Medem, 1958) and M. zuliae (Pritchard and Trebbau, 1984) represent little-known, isolated northern species with relict character (Bour and Zaher, 2005). Dahl s toad-headed turtle (M. dahli), endemic to Colombia, is the only chelid species occurring west of the Andes mountain range (Medem, 1966). It is known from the western part of the Caribbean region of Colombia, including the departments of Córdoba, Bolivar, Su- Koninklijke Brill NV, Leiden, DOI: /

2 374 M. Vargas-Ramírez et al. cre and Atlántico (Zangerl and Medem, 1958; Medem, 1966; Rueda-Almonacid et al., 2007) and from a newly discovered, isolated population in the department of Cesar (Medina-Rangel and Forero-Medina, 2008; fig. 1). Due to its small distribution range and extreme habitat deterioration (Rueda-Almonacid et al., 2007), M. dahli is listed in the threat category Critically Endangered by the IUCN Red List of Threatened Species (IUCN, 2011). The closely related Zulia toad-headed turtle (M. zuliae), endemic to Venezuela, is confined to the Zulia river and its tributaries in the south-western part of the Maracaibo basin (Pritchard and Trebbau, 1984; Rueda-Almonacid et al., 2007; fig. 1). Acknowledging its small distribution range, M. zuliae is listed in the category Vulnerable by the IUCN Red List of Threatened Species (IUCN, 2011). To date, some studies on the ecology and natural history of the two species have been performed (M. dahli: Medem, 1966; de la Ossa- Velasquez, 1998; Castaño-Mora and Medem, 2002; Rueda-Almonacid et al., 2007; Forero- Medina, Cárdenas-Arevalo and Castaño-Mora, 2011; M. zuliae: Rueda-Almonacid et al., 2007; Rojas-Runjaic, 2009). Furthermore, the two species were included in a study that investigated the patterns of geographical distribution of South American chelid turtles (Souza, 2005). The two species have similar habitat preferences (Medem, 1966; Pritchard and Trebbau, 1984; Rueda-Almonacid et al., 2007) and are morphologically very similar (see images and descriptions in Pritchard and Trebbau, 1984; Rueda-Almonacid et al., 2007). Yet, they have never been studied and compared using molecular markers. The present study aims to fill this gap by analyzing their kinship at a molecular genetic level and by applying a relaxed molecular clock to elucidate their divergence and biogeography. Materials and methods Sampling and laboratory procedures Twenty-five saliva samples of Mesoclemmys dahli and four tissue samples of M. zuliae were studied (fig. 1; Appendix). The samples from M. dahli were collected in both regions where the species is known to occur; 18 samples came from the department of Córdoba and seven samples from the department of Cesar. The samples from M. zuliae were collected at the type locality and two additional sites, covering largely its distribution range in north-south direction. For inferring phylogeography, sequences of the mitochondrial cytochrome b gene (cyt b) were generated for all samples. For molecular clock calculations, the mitochondrial 12S rrna and NADH dehydrogenase subunit 4 (ND4) genes and three nuclear loci, the oocyte maturation factor Mos gene (C-mos), the intron 1 of the RNA fingerprint protein 35 gene (R35) and the recombination-activating gene 2 (Rag 2), of one sample each of M. dahli and M. zuliae were sequenced. Sequences of the same mitochondrial and nuclear DNA fragments were also produced for M. gibba, Phrynops geoffroanus and P. hilarii, as far as these were not available from GenBank (see Appendix). Genomic DNA was extracted using a Qiagen DNA blood extraction kit (Qiagen Benelux B.V., Venlo, The Netherlands), following the manufacturer s instructions. The 12S rrna gene was amplified and sequenced using the primers L1091 and H1478 (Kocher et al., 1989), and for the partial ND4 gene plus adjacent DNA coding for trnas, the primers L-ND4 and H-Leu (Stuart and Parham, 2004) were applied. The cyt b gene was amplified and sequenced in two fragments overlapping by approximately 300 bp using the primer pairs mt-a-neu3 + mt-e-rev2 and mt-c- For2 + mt-f-na (Fritz et al., 2006; Praschag et al., 2007). For the nuclear loci the following primer pairs were used C- mos: G136 + G137 (Georges et al., 1998), R35: R35Ex1 + R35Ex2 (Fujita et al., 2004), and Rag 2: F2-1 + R2-1 (Le et al., 2006). PCRs were carried out in a total volume of 50 μl containing 1 unit Taq polymerase (Bioron, Ludwigshafen, Germany), 1 buffer (as recommended by the supplier), 0.5 μm of each primer, and 0.2 mm of each dntp (Fermentas, St. Leon-Rot, Germany). PCR products were purified using the ExoSAP-IT enzymatic cleanup (USB Europe GmbH, Staufen, Germany; modified protocol: 30 min at 37 C, 15 min at 80 C) and sequenced on an ABI 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). Phylogeographic analyses DNA sequences were edited and aligned using CHROMAS 1.51 ( and BIOEDIT (Hall, 1999). For the 1067-bp-long cyt b sequences of Mesoclemmys dahli, M. gibba and M. zuliae, uncorrected p distances were calculated in MEGA 5.05 (Tamura et al., 2011). Evolutionary relationships between sequences of M. dahli and M. zuliae were inferred using TCS 1.21 (Clement et al., 2000). For M. dahli, haplotype

3 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 375 Figure 1. Distribution of Mesoclemmys dahli and M. zuliae (based on Rueda-Almonacid et al., 2007) and sampling sites (colour-coded). For M. dahli, the Córdoba population is indicated in green and the Cesar population in orange; sites of M. zuliae, blue. The Serranía de Perijá is highlighted in red. Inset: juvenile Mesoclemmys dahli from the Cesar population (photo: Guido Medina). diversity (h) and nucleotide diversity (π) were calculated using DNASP 5.0 (Librado and Rozas, 2009). The influence of geography on genetic divergence within M. dahli and between M. dahli and M. zuliae was examined using Mantel tests as implemented in the software IBD (Isolationby-Distance; Bohonak, 2002). A first Mantel test was based on uncorrected p distances and geographical distances (km) between the two populations of M. dahli and M. zuliae. In another Mantel test, the influence of the Serranía de Perijá, an Andean mountain chain separating the distribution ranges of M. dahli and M. zuliae, was examined by combining a categorical matrix with genetic distances. In this matrix, the distance between the two populations of M. dahli was coded with 0 (corresponding to the absence of a mountain barrier), while the distance between each population of M. dahli and M. zuliae was coded with 1 (acknowledging the presence of the mountain barrier). Relaxed molecular clock The phylogeny of Mesoclemmys dahli, M. zuliae and other chelid turtle species was inferred to obtain a backbone for molecular clock calculations. For this purpose, the 12S rrna, cyt b, ND4, C-mos, R35 and Rag 2 sequences of M. dahli and M. zuliae were aligned with homologous sequence data of M. gibba, Chelus fimbriatus, Chelodina rugosa, Phrynops geoffroanus, P. hilarii, Pelomedusa subrufa, Podocnemis expansa, andindotestudo elongata (for accession numbers, see Appendix). Sequences for each taxon were concatenated. The resulting supermatrix was of 4450 bp length (including gaps), corresponding to 2341 bp of mtdna and 2109 bp of ndna. The partial 12S rrna gene contributed 493 bp; the nearly complete cyt b gene, 1067 bp; the partial ND4 gene plus adjacent DNA coding for trnas, 781 bp; C-mos, 390 bp; R35, 1025 bp; and Rag 2, 694 bp. Based on this supermatrix, phylogenetic trees were calculated using Maximum Likelihood (ML) and Bayesian Inference (BI) analyses, applying the following partition schemes: (1) unpartitioned, (2) by gene, i.e., each gene corresponds to a distinct partition, and (3) maximum partitioning, i.e., using each codon of each protein-coding gene as distinct partition plus each non-protein-coding gene or DNA block as distinct partition. Indotestudo elongata (family Testudinidae, suborder Cryptodira) served for treerooting. ML analyses were run with RAxML (Stamatakis, 2006) using the graphical user interface raxml- GUI 0.93 (Silvestro and Michalak, 2011) and the GTR + G model across every partition. To explore the robustness of the branching patterns, five independent ML searches were performed using the fast bootstrap algorithm. Subsequently, 1000 thorough bootstrap replicates were calculated and plotted against the tree with the highest likelihood value. The BI analyses were run in MrBAYES 3.1

4 376 M. Vargas-Ramírez et al. (Ronquist and Huelsenbeck, 2003) using four incrementally heated Markov chains; posterior probabilities were obtained from the 50% majority rule consensus tree. For each independent run, the variation in likelihood scores was examined by plotting ln L scores against the number of generations, and the burn-in was set to sample only the plateau of the most likely trees. In a conservative approach, 40% of all sampled trees were discarded, although the plateau of likelihood values had been reached before. The best-fit model of nucleotide substitution was established for each partition using the Akaike information criterion of MrMODELTEST (Nylander, 2002) and incorporated into a single tree search (mixed model partition approach; Nylander et al., 2004). The following models were suggested for the partitioning by gene: 12S rrna GTR + G, cyt b TVM+ I + G, ND4 GTR+ G, t-rnas TVM + G, C-mos K80 + G, R35 TVM + G, and Rag 2 HKY; and for the codons of protein-coding genes in the maximum partitioning scheme: cyt b 1st codon position GTR + G, cyt b 2nd codon position GTR + I + G, cyt b 3rd codon position TIM + G, ND4 1st codon position TVM + G, ND4 2nd codon position TVM + G, ND4 3rd codon position HKY + I + G, C-mos 1st codon position HKY + I, C-mos 2nd codon position TrNef + G, C-mos 3rd codon position K81uf + G, Rag 2 1st codon position TrN + G, Rag 2 2nd codon position GTR + G, and Rag 2 3rd codon position HKY. The divergence time of M. dahli and M. zuliae was estimated by a Bayesian relaxed molecular clock approach (MULTIDISTRIBUTE package; Thorne, Kishino and Painter, 1998; Thorne and Kishino, 2002). Fossil evidence was used for constraining the minimum ages of two nodes within the obtained phylogeny. The split between Chelodina rugosa and the South American chelids was calibrated with the range of million years (ma), based on the Lower Cretaceous record of Prochelidella cerrobarcinae (Cerro Barcino formation, Aptian-Albian?), the oldest South American chelid (de la Fuente et al., 2011; node A in fig. 4). Furthermore, the split between Pelomedusidae (represented by Pelomedusa subrufa) and Podocnemididae (represented by Podocnemis expansa)wasset to the lower and upper boundaries of the Valanginian ( ma), the Early Cretaceous stage from which the earliest podocnemidoid turtle is known (Cadena, 2011; node B in fig. 4). Chelodina rugosa is a representative of Australasian chelids that constitute the sister group of South American chelids (Georges et al., 1998), and Prochelidella is thought to be allied to the extant genus Acanthochelys of South America (de la Fuente et al., 2011). Pelomedusa subrufa and Podocnemis expansa are representatives of the families Pelomedusidae and Podocnemididae, respectively. Among extant turtles, Pelomedusidae and Podocnemididae together are the sister group of Chelidae (Gaffney et al., 2006). In order to determine the appropriate nucleotide substitution model parameters, the data set was analyzed using the program PAML 3.13 (Yang, 1997). Subsequently, the branch lengths and their variance-covariance matrix were estimated with the program ESTBRANCHES. Using the application MULTIDIVTIME, Markov chains were run three times with the settings numsamps = , sampfreq = 100 and a burn-in of and compared to test the stability of the results. According to Cadena (2011), the prior for the mean of the ingroup root age (rttm) was set to a minimum of ma (with 2 ma SD), corresponding to the split between (Pelomedusidae + Podocnemididae) and Chelidae. Mean and standard deviation of the rate of molecular evolution at the ingroup root node (rtrate and rtratesd) were substitutions per site and million years with 1 time unit = 1 ma (calculated with the mean of the branch lengths from ESTBRANCHES). Mean and standard deviation of the Brownian motion constant (brownmean and brownsd) were set to and bigtime to 220 ma according to the age of the oldest known chelonian, Odontochelys semitestacea (Li et al., 2008). Results Phylogeography The25cytb sequences of Mesoclemmys dahli and the four cyt b sequences of M. zuliae were assigned to two clearly distinct haplotype clusters in the parsimony network (fig. 2), differing by a minimum of 18 mutational steps. Sequences of M. dahli corresponded to 15 haplotypes; sequences of M. zuliae, to two haplotypes. Among haplotypes of M. dahli, some loops occurred. The maximum number of mutational steps within M. dahli resembles the minimum divergence between M. dahli and M. zuliae. The two haplotypes of M. zuliae differed by only one mutation. For the two populations of M. dahli, no shared haplotypes were observed. However, this could be a bias due to small sample size because the haplotypes of each population did not form a distinct cluster, but were rather randomly distributed in the network (fig. 2). The mean uncorrected p distance between sequences of M. dahli and M. zuliae was 2.22%; mean divergences within each species were 0.58% and 0.04%, respectively. The two populations of M. dahli differed by 0.77%. Within the Córdoba population of M. dahli a sequence divergence of 0.41% occurred; within the Cesar population the divergence was 0.33%. The sequence of M. gibba differed by 14.16% from that of M. dahli and by 13.27% from that of M. zuliae. The Mantel tests revealed no signifi-

5 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 377 correlation between genetic distance and the Serranía de Perijá acting as a geographical barrier (Z = , r = , p = 0.001). Haplotype diversities (h) of the populations of M. dahli from Córdoba (n = 18) and Cesar (n = 7) were and 0.952, respectively, and their nucleotide diversities (π) were and When all 25 sequences were lumped together, h was and π was Divergence time of Mesoclemmys dahli and M. zuliae The topologies of the ML and BI trees were congruent among all partition schemes and support values were always maximal or very close to the maximum (fig. 3). Mesoclemmys gibba was the sister taxon of M. dahli + M. zuliae. A clade comprising Phrynops geoffroanus + P. hilarii was sister to Mesoclemmys, and the South American species Chelus fimbriatus and the Australian species Chelodina rugosa were the successive sister taxa. This phylogeny, consistent with expectations (Georges et al., 1998; Gaffney et al., 2006), served as a backbone for relaxed molecular clock calculations (fig. 4). Obtained node ages are presented in table 1. The mean divergence time of M. dahli and M. zuliae was dated to 10.6 ma, with a standard deviation of 2.8 ma and a 95% confidence interval of ma. Discussion Figure 2. Parsimony network of cyt b haplotypes of Mesoclemmys dahli and M. zuliae, based on an alignment of 1067 bp length. Circle size indicates haplotype frequency. Missing haplotypes are shown as small black circles. Each line connecting haplotypes corresponds to one mutational step. Colour code as in fig. 1; haplotype codes refer to the Appendix. cant correlation between genetic and geographical distances when cyt b sequences of M. dahli and M. zuliae were compared (Z = , r = , p = ), but a significant According to our analyses, Mesoclemmys dahli is closely related to M. zuliae, which is in line with their morphological similarity (Pritchard and Trebbau, 1984). However, a phylogenetic analysis using morphological characters could not resolve the relationships among the ten currently recognized Mesoclemmys species (Bour and Zaher, 2005). We included in our phylogenetic analyses three of these species and found a well-supported sister group relationship of M. dahli and M. zuliae, with M. gibba being the

6 378 M. Vargas-Ramírez et al. Figure 3. Bayesian tree based on 2341 bp of mitochondrial DNA (12 S rrna, ND4, cyt b) and 2109 bp of nuclear DNA sequences (C-mos, R35, Rag 2; partitioned by gene). Support values are Bayesian posterior probabilities (top) and ML bootstrap values (bottom); asterisks indicate maximum support under both methods. This tree was used for the Relaxed Molecular Clock calculations. successive sister taxon (fig. 3). Compared to M. gibba, M. dahli and M. zuliae seem to be weakly differentiated. The number of mutational steps among haplotypes of M. dahli resembles the observed divergence between M. dahli and M. zuliae (fig. 2), and this pattern is also reflected by low between-species divergences with respect to uncorrected p distances of the cyt b gene. The sequence divergence between M. dahli and M. zuliae amounts only to 2.22%, while these two species differ from M. gibba by 14.16% and 13.27%, respectively. With respect to cyt b sequences, the value of 2.22% is close to the lower divergence limit for any distinct congeneric turtle and tortoise species. Yet, for some emydid turtles (Emys, Graptemys, Trachemys) even lower species divergence values were reported (Lamb et al., 1994; Fritz et al., 2005, 2012b). In analogy to the widely used barcoding approach (Hebert, Ratnasingham and de Waard, 2003) relying on divergences of the COI gene as a yardstick, uncorrected p distances of the cyt b gene have been frequently used as a tool for species delineation of chelonians (e.g. Vargas-Ramírez et al., 2010b; Praschag et al., 2011; Stuckas and Fritz, 2011; Fritz et al., 2012a, 2012b; Kindler et al., 2012). However, these studies pointed out that no universal thresholds should be used because the critical divergence value differs among various chelonian groups. Therefore, additional Mesoclemmys species need to be studied before the status of M. dahli and M. zuliae as distinct species should be challenged.

7 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 379 Figure 4. Divergence time estimates using the MULTIDISTRIBUTE package. Dark grey bars at nodes represent 95% confidence intervals. The light grey column on the right shows the time of the main uplift of the Serranía de Perijá (Late Miocene-Pliocene; Kellogg, 1984). Letters indicate nodes calibrated with fossil evidence (see table 1). PP, Plio-Pleistocene. Table 1. Results of the relaxed molecular clock analysis using the MULTIDISTRIBUTE package. SD = Standard deviation, CI = 95% confidence interval. All dates are given in million years (ma). Node Mean SD CI Mesoclemmys dahli + M. zuliae Phrynops geoffroanus + P. hilarii (Mesoclemmys dahli + M. zuliae) + M. gibba Mesoclemmys + Phrynops (Mesoclemmys + Phrynops) + Chelus ((Mesoclemmys + Phrynops) + Chelus) + Chelodina Pelomedusa + Podocnemis Chelidae + (Pelomedusa + Podocnemis) Node A (fig. 4): Fossil constraint: ma (Prochelidella cerrobarcinae; de la Fuente et al., 2011). Node B (fig. 4): Fossil constraint: ma (earliest podocnemidoid turtle; Cadena, 2011). Haplotype and nucleotide diversities of M. dahli are much higher than in another endemic turtle species of Colombia, Podocnemis lewyana. Based on the rapidly evolving D-loop, haplotype diversity (h) and nucleotide diversity (π) of119samplesofp. lewyana were and , respectively (Vargas-Ramírez et al., 2012), contrary to h = and π = in the 25 samples of M. dahli analyzed in the present study for the more slowly evolving cyt b gene. This pronounced dissimilarity may reflect differences in the phylogeographies and demographic histories of the two species, but also different habitat preferences. While P. lewyana is a true river turtle, M. dahli is typically found in shallow and quiet ponds or small

8 380 M. Vargas-Ramírez et al. brooks in seasonally dry tropical forest (Medem, 1966; Rueda-Almonacid et al., 2007), suggesting that the patchy structures in such habitat may favour more genetic diversity than the occurrence along a river course. Souza (2005) proposed that the distribution of South American chelids is mainly correlated with major river or drainage basins. According to this hypothesis, the distribution of M. dahli and M. zuliae is associated with the Magdalena and Orinoco basins, respectively. Notwithstanding that M. dahli occurs close to the Magdalena river, its typical habitats are neither directly linked to this river nor to its major tributaries. Mesoclemmys zuliae lives in similar habitats as M. dahli. Moreover, the range of M. zuliae is far away from the Orinoco (Rueda-Almonacid et al., 2007), and the rivers and streams of the Maracaibo basin are not connected to the Orinoco basin. Unlike true river turtles, as Podocnemis species, M. dahli and M. zuliae seem therefore to be not tied to the ecosystems of the Magdalena and Orinoco rivers. Consequently, it is unlikely that the origin of M. dahli and M. zuliae is directly related to the hydrographic history of these rivers. However, the estimated mean divergence time of M. dahli and M. zuliae correlates quite well with the uplift of the Serranía de Perijá (see above) in the Late Miocene and Pliocene (Kellogg, 1984; fig. 4), suggesting that this orogenetic process could be directly responsible for the evolution of the two species. Also in another South American chelid species, Hydromedusa maximiliani from eastern Brazil, phylogeographic structure is shaped by mountain chains, and the estimated divergence time of the major phylogeographic groups within H. maximiliani fits the uplift of the respective mountains during the Pliocene and Miocene (Souza et al., 2003). Acknowledgements. We thank the Grupo de Conservación y Biodiversidad of the Instituto de Ciencias Naturales de la Universidad Nacional de Colombia, Fundación Biodiversa Colombia and Instituto de Biología Tropical Roberto Franco (IBTRF) de la Universidad Nacional de Colombia for institutional and logistical support. Carl J. Franklin (Amphibian and Reptile Research Center, University of Texas at Arlington) and Ingo Pauler provided samples of Mesoclemmys zuliae. Christian Kehlmaier, Edgar Lehr, Anke Müller, and Anja Rauh assisted in the lab. Ylenia Chiari and João Lourenço shared sequences of Phrynops hilarii with us. Massimo Delfino, Marcelo Sánchez-Villagra, Markus Wilmsen and Juliana Sterli helped with literature of fossil chelids. Thanks to Guido Medina for the photograph of M. dahli. Mario Vargas-Ramírez research in Germany is funded by the Humboldt Foundation (Georg Forster fellowship). References Amato, M.L., Brooks, R.J., Fu, J. (2008): A phylogeographic analysis of populations of the wood turtle (Glyptemys insculpta) throughout its range. Mol. Ecol. 17: Bohonak, A.J. (2002): IBD (Isolation-by-Distance): a program for analyses of isolation by distance. J. Hered. 93: Bour, R., Zaher, H. (2005): A new species of Mesoclemmys, from the open formations of Northeastern Brazil (Chelonii, Chelidae). Pap. Avul. Zool. 45: Butler, J.M., Dodd, C.K., Aresco, M., Austin, J.D. (2011): Morphological and molecular evidence indicates that the Gulf Coast box turtle (Terrapene carolina major) is not a distinct evolutionary lineage in the Florida Panhandle. Biol. J. Linn. Soc. 102: Cadena, E.A. (2011): Potential earliest record of podocnemidoid turtles from the Early Cretaceous (Valanginian) of Colombia. J. Paleontol. 85: Castaño-Mora, O.V., Medem, F. (2002): Batrachemys dahli. In: Libro rojo de reptiles de Colombia. Libros rojos de especies amenazadas de Colombia, p Castaño- Mora, O.V., Ed., Instituto de Ciencias Naturales (Universidad Nacional de Colombia), Ministerio del Medio Ambiente and Conservación Internacional, Bogotá, Colombia. Clement, M., Posada, D., Crandall, K.A. (2000): TCS: a computer program to estimate gene genealogies. Mol. Ecol. 9: de la Fuente, M., Umazano, A.M., Sterli, J., Carballido, J.L. (2011): New chelid turtles of the lower section of the Cerro Barcino formation (Aptian-Albian?), Patagonia, Argentina. Cretaceous Res. 32: de la Ossa-Velasquez, J.L. (1998): Phrynops dahli: a little known turtle endemic to the Caribbean coast of Colombia. Reptilia 3: Forero-Medina, G., Cárdenas-Arevalo, G., Castaño-Mora, O.V. (2011): Abundance, home range, and movement patterns of the endemic species Dahl s toad-headed turtle (Mesoclemmys dahli) in Cesar, Colombia. Chelon. Conserv. Biol. 10: Fritz, U., Havaš, P. (2007): Checklist of chelonians of the world. Vertebr. Zool. 57:

9 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 381 Fritz, U., Fattizzo, T., Guicking, D., Tripepi, S., Pennisi, M.G., Lenk, P., Joger, U., Wink, M. (2005): A new cryptic species of pond turtle from southern Italy, the hottest spot in the range of the genus Emys. Zool. Scr. 34: Fritz, U., Auer, M., Bertolero, A., Cheylan, M., Fattizzo, T., Hundsdörfer, A.K., Martín Sampayo, M., Pretus, J.L., Široký, P., Wink, M. (2006): A rangewide phylogeography of Hermann s tortoise, Testudo hermanni (Reptilia: Testudines: Testudinidae): implications for taxonomy. Zool. Scr. 35: Fritz, U., Guicking, D., Kami, H., Arakelyan, M., Auer, M., Ayaz, D., Ayres Fernández, C., Bakiev, A.G., Celani, A., Džukić, G., Fahd, S., Havaš, P., Joger, U., Khabibullin, V.F., Mazanaeva, L.F., Široký, P., Tripepi, S., Valdeón Vélez, A., Velo Antón, G., Wink, M. (2007): Mitochondrial phylogeography of European pond turtles (Emys orbicularis, Emys trinacris) an update. Amphib.- Reptil. 28: Fritz, U., Ayaz, D., Buschbom, J., Kami, H.G., Mazanaeva, L.F., Aloufi, A.A., Auer, M., Rifai, L., Šilić, T., Hundsdörfer, A.K. (2008): Go east: phylogeographies of Mauremys caspica and M. rivulata discordance of morphology, mitochondrial and nuclear genomic markers and rare hybridization. J. Evolution. Biol. 21: Fritz, U., Ayaz, D., Hundsdörfer, A.K., Kotenko, T., Guicking, D., Wink, M., Tok, C.V., Çiçek, K., Buschbom, J. (2009): Mitochondrial diversity of European pond turtles (Emys orbicularis) in Anatolia and the Ponto- Caspian Region: multiple old refuges, hotspot of extant diversification and critically endangered endemics. Org. Divers. Evol. 9: Fritz, U., Alcalde, L., Vargas-Ramírez, M., Goode, E.V., Fabius-Turoblin, D.U., Praschag, P. (2012a): Northern genetic richness and southern purity, but just one species in the Chelonoidis chilensis complex. Zool. Scr. 41: Fritz, U., Stuckas, H., Vargas-Ramírez, M., Hundsdörfer, A.K., Maran, J., Päckert, M. (2012b): Molecular phylogeny of Central and South American slider turtles: implications for biogeography and systematics (Testudines: Emydidae: Trachemys). J. Zool. Syst. Evol. Res. 50: Fujita, M.F., Engstrom, T.N., Starkey, D.E., Shaffer, H.B. (2004): Turtle phylogeny: insights from a novel nuclear intron. Mol. Phylogenet. Evol. 31: Gaffney, E.S., Tong, H., Meylan, P.A. (2006): Evolution of the side-necked turtles: the families Bothremydidae, Euraxemydidae, and Araripemydidae. Bull. Am. Mus. Nat. Hist. 300: Georges, A., Birrel, J., Saint, K.M., McCord, W.P., Donnellan, S.C. (1998): A phylogeny for side-necked turtles (Chelonia: Pleurodira) based on mitochondrial and nuclear gene sequence variation. Biol. J. Linn. Soc. 67: González-Porter, G.P., Hailer, F., Flores-Villela, O., García- Anleu, R., Maldonado, J.E. (2011): Patterns of genetic diversity in the critically endangered Central American river turtle: human influence since the Mayan age? Conserv. Genet. 12: Hall, T.A. (1999): BIOEDIT: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids Symp. Ser. 4: Hebert, P.D.N., Ratnasingham, S., de Waard, J.R. (2003): Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc. R. Soc. Lond. B 270: S96-S99. IUCN [International Union for the Conservation of Nature and Natural Resources] (2011): IUCN Red List of Threatened Species [accessed 22 February 2012]. Karl, S.A., Wilson, D.S. (2001): Phylogeography and systematics of the mud turtle, Kinosternon baurii. Copeia 2001: Kellogg, J.N. (1984): Cenozoic tectonic history of the Sierra de Perijá, Venezuela-Colombia, and adjacent basins. Geol. Soc. Am. Mem. 162: Kindler, C., Branch, W.R., Hofmeyr, M.D., Maran, J., Široký, P., Vences, M., Harvey, J., Hauswaldt, J.S., Schleicher, A., Stuckas, H., Fritz, U. (2012): Molecular phylogeny of African hinge-back tortoises (Kinixys): implications for phylogeography and taxonomy (Testudines: Testudinidae). J. Zool. Syst. Evol. Res. 50: Kocher, T.D., Thomas, W.K., Meyer, A., Edwards, S.V., Pääbo, S., Villablanca, F.X., Wilson, A.C. (1989): Dynamics of mitochondrial DNA evolution in mammals: amplification and sequencing with conserved primers. PNAS 86: Lamb, T., Lydeard, C., Walker, R.B., Gibbons, J.W. (1994): Molecular systematics of map turtles (Graptemys): a comparison of mitochondrial restriction site versus sequence data. Syst. Biol. 43: Le, M., Raxworthy, C.J., McCord, W.P., Mertz, L. (2006): A molecular phylogeny of tortoises (Testudines: Testudinidae) based on mitochondrial and nuclear genes. Mol. Phylogenet. Evol. 40: Lenk, P., Fritz, U., Joger, U., Wink, M. (1999): Mitochondrial phylogeography of the European pond turtle, Emys orbicularis (Linnaeus 1758). Mol. Ecol. 8: Li, C., Wu, X.-C., Rieppel, O., Wang, L.-T., Zhao, L.-J. (2008): An ancestral turtle from the Late Triassic of southwestern China. Nature 456: Librado, P., Rozas, J. (2009): DNASP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: McGaugh, S.E., Eckerman, C.M., Janzen, F.J. (2008): Molecular phylogeography of Apalone spinifera. Zool. Scr. 37: Medem, F. (1966): Contribuciones al conocimiento sobre la ecología y distribución geográfica de Phrynops (Batrachemys) dahli (Testudinata, Pleurodira, Chelidae). Caldasia 9: Medina-Rangel, G.F., Forero-Medina, G.A. (2008): Geographical distribution. Mesoclemmys dahli. Herpetol. Rev. 39: 366. Nylander, J.A.A. (2002): MrMODELTEST v1.0b. Program distributed by the author. Available from: ebc.uu.se/systzoo/staff/nylander.html [accessed 22 February 2012].

10 382 M. Vargas-Ramírez et al. Nylander, J.A.A., Ronquist, F., Huelsenbeck, J.P., Nieves- Aldrey, J.L. (2004): Bayesian phylogenetic analysis of combined data. Syst. Biol. 53: Pearse, D.E., Arndt, A.D., Valenzuela, N., Miller, B.A., Cantarelli, V., Sites, J.R. (2006): Estimating population structure under nonequilibrium conditions in a conservation context: continent-wide population genetics of the giant Amazon River turtle, Podocnemis expansa (Chelonia; Podocnemididae). Mol. Ecol. 15: Pedall, I., Fritz, U., Stuckas, H., Valdéon, A., Wink, M. (2011): Gene flow across secondary contact zones of the Emys orbicularis complex in the Western Mediterranean and evidence for extinction and re-introduction of pond turtles on Corsica and Sardinia (Testudines: Emydidae). J. Zool. Syst. Evol. Res. 49: Praschag, P., Hundsdörfer, A.K., Fritz, U. (2007): Phylogeny and taxonomy of endangered South and Southeast Asian freshwater turtles elucidated by mtdna sequence variation (Testudines: Geoemydidae: Batagur, Callagur, Hardella, Kachuga, Pangshura). Zool. Scr. 36: Praschag, P., Stuckas, H., Päckert, M., Maran, J., Fritz, U. (2011): Mitochondrial DNA sequences suggest a revised taxonomy of Asian flapshell turtles (Lissemys Smith, 1931) and the validity of previously unrecognized taxa (Testudines: Trionychidae). Vertebr. Zool. 61: Pritchard, P.C.H., Trebbau, P. (1984): The Turtles of Venezuela. Contributions to Herpetology, 2. Society for the Study of Amphibians and Reptiles, Oxford, Ohio. Rojas-Runjaic, F.J.M. (2009): Mesoclemmys zuliae, el enigmático galápago cabezón del sur del lago Maracaibo. Reptilia 80: Ronquist, F., Huelsenbeck, J.P. (2003): MrBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: Rosenbaum, P.A., Robertson, J.A., Zamudio, K.R. (2007): Unexpectedly low genetic divergences among populations of the threatened bog turtle (Glyptemys muhlenbergii). Conserv. Genet. 8: Rueda-Almonacid, J.V., Carr, J.L., Mittermeier, R.A., Rodríguez-Mahecha, J.V., Mast, R.B., Vogt, R.C., Rhodin, A.G.J., de la Ossa-Velásquez, J., Rueda, J.N., Goettsch-Mittermeier, C. (2007): Las Tortugas y los cocodrilianos de los países andinos del trópico. Bogotá, Conservación Internacional (Serie de guías tropicales de campo). Silvestro, D., Michalak, I. (2011): raxmlgui: a graphical frontend for RAxML. Org. Divers. Evol., DOI: / s Souza, F.L. (2005): Geographical distribution patterns of South American side-necked turtles (Chelidae), with emphasis on Brazilian species. Rev. Esp. Herpetol. 19: Souza, F.L., Cunha, A.F., Oliveira, M.A., Pereira, G.A.G., dos Reis, S.F. (2003): Preliminary phylogeographic analysis of the Neotropical freshwater turtle Hydromedusa maximiliani (Chelidae). J. Herpetol. 37: Stamatakis, A. (2006): RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: Stuart, B.L., Parham, J.F. (2004): Molecular phylogeny of the critically endangered Indochinese box turtle (Cuora galbinifrons). Mol. Phylogenet. Evol. 31: Stuckas, H., Fritz, U. (2011): Identity of Pelodiscus sinensis revealed by DNA sequences of an approximately 180- year-old type specimen and a taxonomic reappraisal of Pelodiscus species (Testudines: Trionychidae). J. Zool. Syst. Evol. Res. 49: Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S. (2011): MEGA 5: Molecular Evolutionary Genetics Analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28: Thorne, J.L., Kishino, H. (2002): Divergence time and evolutionary rate estimation with multilocus data. Syst. Biol. 51: Thorne, J.L., Kishino, H., Painter, I.S. (1998): Estimating the rate of evolution of the rate of molecular evolution. Mol. Biol. Evol. 15: Ureña-Aranda, C.A., Espinosa de los Monteros, A. (2012): The genetic crisis of the Mexican Bolson tortoise (Gopherus flavomarginatus: Testudinidae). Amphib.-Reptil. 33: van Dijk, P.P., Iverson, J.B., Shaffer, H.B., Bour, R., Rhodin, A.G.J. (2011): Turtles of the world, 2011 update: annotated checklist of taxonomy, synonymy, distribution, and conservation status. Chelon. Res. Monogr. 5: Vargas-Ramírez, M., Chiari, Y., Castaño-Mora, O.V., Menken, S.B.J. (2007): Low genetic variability in the endangered Colombian endemic freshwater turtle Podocnemis lewyana (Testudines, Podocnemididae). Contr. Zool. 76: 1-7. Vargas-Ramírez, M., Maran, J., Fritz, U. (2010a): Redand yellow-footed tortoises (Chelonoidis carbonaria, C. denticulata) in South American savannahs and forests: do their phylogeographies reflect distinct habitats? Org. Divers. Evol. 10: Vargas-Ramírez, M., Vences, M., Branch, W.R., Daniels, S.R., Glaw, F., Hofmeyr, M.D., Kuchling, G., Maran, J., Papenfuss, T.J., Široký, P., Vieites, D.R., Fritz, U. (2010b): Deep genealogical lineages in the widely distributed African helmeted terrapin: evidence from mitochondrial and nuclear DNA (Testudines: Pelomedusidae: Pelomedusa subrufa). Mol. Phylogenet. Evol. 56: Vargas-Ramírez, M., Stuckas, H., Castaño-Mora, O.V., Fritz, U. (2012): Extremely low genetic diversity and weak population differentiation in the endangered Colombian river turtle Podocnemis lewyana (Testudines: Podocnemididae). Conserv. Genet. 13: Walker, D., Moler, P.E., Buhlmann, K.A., Avise, J.C. (1998): Phylogeographic uniformity in mitochondrial DNA of the snapping turtle (Chelydra serpentina). Anim. Conserv. 1: Weisrock, D.W., Janzen, F.J. (2000): Comparative molecular phylogeography of North American softshell turtles (Apalone): implications for regional and wide-scale historical evolutionary forces. Mol. Phylogenet. Evol. 14:

11 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 383 Yang, Z. (1997): PAML: a program package for phylogenetic analysis by maximum likelihood. Comput. Appl. Biosci. 13: Zangerl, R., Medem, F. (1958): A new species of chelid turtle, Phrynops (Batrachemys) dahli, from Colombia. Bull. Mus. Comp. Zool. 119: Submitted: February 28, Final revision received: July 25, Accepted: July 30, Associated Editor: Sylvain Ursenbacher.

12 384 M. Vargas-Ramírez et al. Appendix Table A.1. Mesoclemmys samples and outgroups used in the present study. MTD T numbers refer to saliva or tissue samples in the collection of the Museum of Zoology, Senckenberg Dresden. Species Locality mtdna haplotype GenBank accession numbers MTD T cyt b 12S rrna ND4 C-mos R35 Rag 2 Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) A JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) A JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) A JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) A JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) A JX Mesoclemmys dahli Momíl, Córdoba, Colombia ( N W) B JX Mesoclemmys dahli Serradero, Purisima, Córdoba, B JX Colombia ( N W) Mesoclemmys dahli La Confianza, Lorica, Córdoba, C JX Colombia ( N W) Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) D JX JX JX JX JX JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) E JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) E JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) E JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) F JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) G JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) H JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) H JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) H JX Mesoclemmys dahli Arache, Chimá, Córdoba, Colombia ( N W) I JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, J JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, J JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, K JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, L JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, M JX

13 Weak genetic divergence between turtles Mesoclemmys dahli and M. zuliae (Testudines: Pleurodira: Chelidae) 385 Table A.1. (Continued.) Species Locality mtdna haplotype GenBank accession numbers MTD T cyt b 12S rrna ND4 C-mos R35 Rag 2 Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, N JX Mesoclemmys dahli Chimichagua, Cienaga de Zapatoza, Cesar, O JX Mesoclemmys zuliae 10 km W of border of Cienagas de Juan Manuel National Park, JX Zulia, Venezuela ( N W) a Mesoclemmys zuliae 50 km N of border of Cienagas de Juan Manuel National Park, b JX JX JX JX JX JX Zulia, Venezuela ( N W) Mesoclemmys zuliae Caño Madre Vieja, Colon, Zulia, b JX Venezuela type locality (8 53 N W) Mesoclemmys zuliae Caño Madre Vieja, Colon, Zulia, b JX Venezuela type locality (8 53 N W) Mesoclemmys gibba Bolognesi, Ucayali, Peru JX JX JX JX Mesoclemmys gibba EF AF Phrynops geoffroanus JX JX JX JX JX JX Phrynops hilarii JN JN JN JX JX Chelus fimbriatus HQ HQ HQ AF AY Chelodina rugosa HQ HQ HQ AF AY Pelomedusa subrufa AF AF FN AF FR FN Podocnemis expansa AM AM FM AF AM AM Indotestudo elongata DQ DQ DQ AY HQ HQ260657