Cryptic Species in Iphisa Elegans Gray, 1851 (Squamata: Gymnophthalmidae) Revealed by Hemipenial Morphology and Molecular Data

bs_bs_banner

Zoological Journal of the Linnean Society, 2012, 166, 361–376. With 3 ﬁgures

Cryptic species in Iphisa elegans Gray, 1851 (Squamata: Gymnophthalmidae) revealed by hemipenial morphology and molecular data

PEDRO M. SALES NUNES1*, ANTOINE FOUQUET1, FELIPE F. CURCIO1, Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 PHILIPPE J. R. KOK2,3 and MIGUEL TREFAUT RODRIGUES1

1Departamento de Zoologia, Instituto de Biociências, Universidade de São Paulo, Caixa Postal 11.461, CEP 05422-970, São Paulo, SP, Brazil 2Biology Department, Unit of Ecology and Systematics, Vrije Universiteit Brussel, 2 Pleinlaan, B-1050 Brussels, Belgium 3Department of Vertebrates, Royal Belgian Institute of Natural Sciences, 29 rue Vautier, B-1000 Brussels, Belgium

Received 7 November 2011; revised 17 May 2012; accepted for publication 17 May 2012

Iphisa elegans Gray, 1851 is a ground-dwelling lizard widespread over Amazonia that displays a broadly conserved external morphology over its range. This wide geographical distribution and conservation of body form contrasts with the expected poor dispersal ability of the species, the tumultuous past of Amazonia, and the previously documented prevalence of cryptic species in widespread terrestrial organisms in this region. Here we investigate this homogeneity by examining hemipenial morphology and conducting phylogenetic analyses of mitochondrial (CYTB) and nuclear (C-MOS) DNA sequence data from 49 individuals sampled across Amazonia. We detected remarkable variation in hemipenial morphology within this species, with multiple cases of sympatric occurrence of distinct hemipenial morphotypes. Phylogenetic analyses revealed highly divergent lineages corroborating the patterns suggested by the hemipenial morphotypes, including co-occurrence of different lineages. The degrees of genetic and morphological distinctness, as well as instances of sympatry among mtDNA lineages/morphotypes without nuDNA allele sharing, suggest that I. elegans is a complex of cryptic species. An extensive and integrative taxonomic revision of the I. elegans complex throughout its wide geographical range is needed.

ADDITIONAL KEYWORDS: C-MOS – cryptic diversity – CYTB – hemipenis – prezygotic isolation.

INTRODUCTION Many small terrestrial vertebrates apparently similar in body form and external characters have surprising Biodiversity is unevenly distributed on Earth with widespread distributions if we consider the extent of tropical forests sheltering more than 50% of the living their range, their putative low vagility, and the highly species known to science (Wilson, 1992; Gaston & complex climatic and geological history of the region Williams, 1996; Myers et al., 2000). Despite its recog- (Antonelli et al., 2010; Hoorn et al., 2010). Studies nition as a megadiverse biome, the species richness of addressing patterns of genetic diversity within such Amazonia remains superﬁcially known and patterns species are scarce but most have revealed ancient and of diversiﬁcation throughout its area are poorly well-geographically structured lineages, suggesting understood (Bush, 1994; Noonan & Wray, 2006). complexes of cryptic species (Chek et al., 2001; Noonan & Wray, 2006; Fouquet et al., 2007; *Corresponding author. E-mail: [email protected] Amézquita et al., 2009; Mott & Vieites, 2009; Geurgas

& Rodrigues, 2010; Funk, Caminer & Ron, 2012). hemipenial polymorphism (McDowell, 1979; Cole & DNA sequences have indeed been decisive in the Hardy, 1981; Zaher & Prudente, 1999; Inger & Marx, recognition of hidden diversity in morphologically 1962) apparently contradicts such a hypothesis, sug- similar forms, from the African elephants (Vogel, gesting that hemipenial characters are not distinc- 2001) to the tiny Amazonian sphaerodactylid geckos tively more conservative, accurate, or informative of the genus Chatogekko (Geurgas & Rodrigues, 2010; than any other morphological features (Myers, 1974; Gamble et al., 2011). Moreover, the adoption of a more Zaher & Prudente, 1999). In fact, the importance of integrative taxonomy as a framework to integrate variation in traits that mediate pre-zygotic isolation, diverse lines of evidence such as molecular and mor- such as call advertisement in frogs, is clear and often phological data has greatly improved the reliability represents a crucially relevant source of characters Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 and efficiency of species delineation (Padial et al., for taxonomic approaches (Padial et al., 2010). 2010; Miralles et al., 2010). The rare reports of discrete populational variations The monotypic genus Iphisa Gray, 1851 (Gymnoph- in hemipenial morphology refer to shape, ornamenta- thalmidae) is one such widespread species that is tion, and/or size (Myers, 1974; Cole & Hardy, 1981). present throughout Amazonia (Dixon, 1974; Peters & Inger & Marx (1962) detected four remarkably Donoso-Barros, 1986). It is a small (maximum snout– distinct hemipenial morphotypes in the snake vent length around 62 mm) leaf litter-dwelling species Calamaria lumbricoidea; nonetheless, despite the showing little morphological variation along this area. geographical proximity among the populations inves- It was erected to allocate I. elegans, after a specimen tigated, none of the patterns was found in sympatry. collected by Wallace and Bates with an imprecise type Zaher & Prudente (1999) also reported remarkably locality comprising a vast area around the region of different hemipenial morphotypes, including sympat- Belém, Pará State, northern Brazil (Gray, 1851; Dixon, ric occurrence, in the widespread Neotropical forest 1974). The only revision of the genus supported Iphisa snake Siphlophis compressus (Pseudoboini). Although as comprising a single species and recognized the the studies above have interpreted cases of hemipe- populations from Peru and Bolivia as a distinct sub- nial polymorphism as intraspecific variation, hemipe- species (I. e. soinii) diagnosed by lack of prefrontals nial differences between populations that are and presence of a higher average number of femoral- otherwise largely homogeneous may also indicate the preanal pores (25.0 vs 19.3 in the typical form) (Dixon, existence of cryptic species, as demonstrated by Pru- 1974). However, as Dixon’s (1974) scheme was based dente & Passos (2010). Moreover, there have as yet on a restricted sample (50 specimens, mainly from been no attempts to explore genetic information in northern and western Amazonia) it was never followed such cases of morphological variation. Interpreting (Hoogmoed, 1973; Ávila-Pires, 1995). Moreover, the the evolutionary significance of these particular status of I. e. soinii with respect to the nominal form examples of morphological plasticity is far from easy, has never been properly examined on the basis of a and molecular data may greatly contribute to the more representative sampling. clarification of such complex patterns. In the course of a broad study of gymnophthalmid Oriented by the astonishing hemipenial variation hemipenial morphology (Nunes, 2011) we detected a observed in I. elegans, a lizard that is otherwise mor- surprising hemipenial variation in the organs of phologically extremely conservative, we (1) investi- I. elegans. The hemipenis of squamates is directly gate its morphological and genetic homogeneity involved in copulation and differences in hemipenial examining hemipenis (49 specimens from 32 different morphology can be considered a physical mechanism localities) and DNA sequences (22 samples from 16 of reproductive isolation (Pope, 1941; Arnold, localities; nuclear and mitochondrial genes) across 1986a). Therefore, characteristics of this organ are Amazonia in order to (2) explore hemipenial variabil- expected to be highly informative for phylogenetic ity among populations and (3) discuss the morpho- studies and species characterization. In fact, for logical diversity in the light of DNA sequences of more than a century hemipenial morphology has nuclear and mitochondrial genes, suggesting hypoth- been efficiently used to understand the relationships eses for the origin of the hemipenial variation found among squamates. in I. elegans throughout Amazonia. The relevance of hemipenial data for phylogenetic studies has been widely debated in the literature. According to some authors, the evolution of this organ MATERIAL AND METHODS would be less subjected to ecological and environmen- tal constraints than features of external morphology, TAXON SAMPLING and thus would be phylogenetically more informative We examined the hemipenis of 49 specimens of I. el- than other structures (Dowling, 1967; Arnold, 1986a; egans from 32 different localities in Brazil, Ecuador, Keogh, 1999). However, evidence of intraspecific French Guiana, Guyana, and Peru (Appendix 1).

Vouchers are deposited in the following institutions sequences of gymnophtalmid species used as outgroup (acronyms in parentheses): American Museum of used remained slightly shorter (~150 bp). Natural History, New York, USA (AMNH); Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil (INPA); Museu de Zoologia, Universidade de São ALIGNMENT AND SEQUENCE DATA ANALYSIS Paulo, São Paulo, Brazil (MZUSP); National Museum Alignments were verified by eye and trimmed to of Natural History, Smithsonian Institution, Wash- remove the most incomplete data, leading to 752 ington, DC, USA (USNM); Natural History Museum, aligned base pairs (bp) of the CYTB gene and 532 bp University of Kansas, Lawrence, USA (KU); Institut of the C-MOS gene. Royal des Sciences Naturelles de Belgique, Brussels, We used the software MrModeltest version 2.3 Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 Belgium (IRSNB); Museo de Historia Natural, Uni- (Nylander, 2004) to select the substitution models that versidad Nacional Mayor San Marcos, Lima, Peru best fit each codon position of CYTB and the fragment (MHNSM); and Centro de Ornitología y Biodiver- as a whole according to the Akaike Information Cri- sidad, Lima, Peru (CORBIDI). terion (Akaike, 1974). The resulting three models for For the DNA-based phylogenetic analyses, we each codon position were used in a partitioned Baye- selected tissue samples (liver) of 22 individuals of sian analysis (Appendix 3) performed with MrBayes I. elegans from 16 localities; 12 of these samples cor- 3.1 (Huelsenbeck & Ronquist, 2001). Bayesian analy- responded to males for which we had hemipenial sis consisted of two independent runs of 1.0 ¥ 107 preparations. To use the closest taxa as outgroups, we generations with random starting trees and ten also analysed DNA sequences of one individual of Markov chains (one cold) sampled every 1000 genera- each genus within Heterodactylini and Iphisiini tions. Adequate burn-in was determined by examining according to Rodrigues et al. (2009). a plot of the likelihood scores of the heated chains for convergence on stationarity as well as the effective HEMIPENIAL PREPARATIONS sample size of values in Tracer 1.5 (Rambaut & Drum- mond, 2003). We also performed maximum-likelihood Hemipenis preparation followed the procedures (ML) and maximum-parsimony (MP) analyses with described by Manzani & Abe (1988), as modified by PAUP 4.0b10 (Swofford, 1993). The ML analysis was Pesantes (1994) and Zaher (1999) for snake organs. In conducted using the best fitting model estimated for addition, we used an alcohol solution of Alizarin Red the entire CYTB fragment. We computed 100 non- to stain ornamenting calcareous structures in an parametric bootstrap pseudoreplicates (Efron, 1979; adaptation of the procedures used by Uzzell (1973) Felsenstein, 1985) with the heuristic search option, and Harvey & Embert (2008). Terminology for hemi- tree bisection-reconnection (TBR) branch swapping penial characters follows Dowling & Savage (1960), and ten random taxon addition replicates per pseu- Uzzell (1973), Zaher (1999), and Myers, Rivas Fuen- doreplicate. Support for proposed clades using MP was mayor & Jadin (2009). assessed via 10 000 non-parametric bootstrap pseudoreplicates (Efron, 1979; Felsenstein, 1985) with the MOLECULAR METHODS heuristic search option, TBR branch swapping and ten Genomic DNA was extracted using the Promega DNA random taxon addition replicates per pseudoreplicate. extraction kit. One fragment of the mitochondrial In total, 275 characters were parsimony informative, gene (mtDNA) Cytochrome b (CYTB) and one frag- and 55 variable characters were not parsimony infor- ment of the nuclear gene (nuDNA) oocyte maturation mative. We considered relationships with posterior factor Mos (C-MOS) were amplified by standard PCR probabilities Ն0.95 and/or bootstrap percentages techniques. Primers and PCR conditions used for Ն70% (Hillis & Bull, 1993) to be strongly supported. amplification were as described by Bickham, Wood & Trees were rooted on Heterodactylini (Caparaonia + Patton (1995) and Kocher et al. (1989) for CYTB and Heterodactylus + Colobodactylus) according to Rod- by Godinho et al. (2006) and Saint et al. (1998) for rigues et al. (2009). C-MOS (Appendix 2). PCR products were purified To determine the most probable alleles for individu- using EXOI (Exonuclease I) and SAP (Shrimp Alka- als recovered as heterozygous on the C-MOS frag- line Phosphatase) techniques. Sequencing was per- ment we used PHASE (Stephens, Smith & Donnelly, formed using ABI Big Dye v3.1 (ABI, Foster City, CA, 2001; Stephens & Donnelly, 2003) implemented in USA) and resolved on an automated sequencer at DnaSP 5 (Librado & Rozas, 2009). We used default Instituto de Química da Universidade de São Paulo conditions, including 500 iterations (which were suf- (IQUSP – São Paulo, Brazil) and Genomic Engen- ficient to reach stationarity), a burn-in of 100, and a haria (São Paulo). Sequences were edited and aligned thinning interval of 1. To improve reliability, we ran with Codon Code Aligner v.3.5.2. New sequences were the algorithm multiple times with a different random deposited in GenBank (Appendix 1). Some CYTB number of seeds. We chose the run with the highest

Figure 1. Hemipenial morphotypes. A, morphotype 1 (MZUSP 82662: Aripuanã, Mato Grosso, Brazil); B, morphotype 2 (MZUSP 82428: Juruena, Mato Grosso, Brazil); C, morphotype 3 (MZUSP 100257: Igarapé-Açu, Amazo- nas, Brazil); D, morphotype 4 (IRSNB 17069: Kaieteur National Park, Guyana); E, morphotype 5 (MHNSM 16718: Distrito Genaro Herrera, Provincia Requena, Peru). Scale bars = 3 mm. ᭤ average value for the goodness of ﬁt. One individual Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 remained with ambiguous phasing (MHNC 10082). Statistical parsimony network was calculated for the phased C-MOS alignment using TCS 1.21 (Clement, Posada & Crandall, 2000), with a 95% connection limit.

RESULTS HEMIPENIAL MORPHOLOGY Analysis of hemipenial morphology revealed a remarkable variation among I. elegans populations; in one case, variations were also detected among specimens of the same population (i.e. specimens belonging to the same locality). We recognized ﬁve distinct hemipenial morphotypes (Fig. 1) based on hemipenial body shape, position and number of calcareous spicules, and size and form of lobes. We describe the ﬁve hemipenial morphotypes, explicitly associating each one with their respective localities of occurrence.

Morphotype 1 (Figs 1A, 2A: Candidate species 1) Localities of occurrence: BRAZIL: Mato Grosso: Apiacás; Aripuanã; PERU: Loreto: Rio Ampiyacu.

General description: Hemipenes slender and cylindrical; hemipenial body covered by numerous calciﬁed spicules; lobes short, apexes ornamented by small papillate folds.

Variation: Populations from Aripuanã and Rio Ampiy- acu (the latter not sampled in the molecular analyses) have a narrow bare area on the central region of the asulcate face, whereas the hemipenial bodies of the remaining specimens are entirely covered by spicules.

Morphotype 2 (Figs 1B, 2A: Candidate species 2) Localities of occurrence: BRAZIL: Amazonas: Alto Rio Aripuanã; Interfluve of Rivers Madeira–Purus; Campo Catuquira; Campo Tupana; São Sebastião (Rio Abacaxis); Igarapé-Açu (Rio Abacaxis); Mato Grosso: Juruena; Rondônia: Rio Machado. penial body and on the asulcate face; spines of asul- General description: Base of hemipenial body dis- cate face organized in two rows, converging towards tinctly wider than apex (‘pear-shaped’ organ); large apex into a single row attaining level of lobular calcified spines present on lateral surface of hemi- crotch; flounces lacking calcified spicules; long and

© 2012 The Linnean Society of London, Zoological Journal of the Linnean Society, 2012, 166, 361–376 CRYPTIC SPECIES IN IPHISA ELEGANS GRAY, 1851 365 Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021

Figure 2. A, phylogram hypothesized from Bayesian analysis using 752 bp of mtDNA (CYTB). For Bayesian analysis we used a partitioned model of evolution combining one model, estimated using MrModeltest 2, for each codon position. Node supports are indicated with 1, posterior probability*100 (10 000 000 generations sampled every 1000; 10 chains); 2, maximum likelihood bootstrap support (n = 100); maximum parsimony bootstrap support (n = 10 000). Posterior probabilities equal to 1 and 0.99 and bootstrap values equal to 100 and 99 are indicated with ‘*’ while ‘–’ indicates bootstrap support <50%. Relationships that remained poorly supported are indicated in red. Hemipenial morphotypes are illustrated beside each of their corresponding clades. Localities for which molecular and hemipenial data are available are indicated by ‘°’. B, statistical parsimony network based on phased C-MOS alleles (n = 44). Colours correspond to the major clades illustrated in the mtDNA-based phylogenic tree reconstruction with the size of the circles being proportional to the frequency of the allele as also indicated in the circles. Each C-MOS allele has been coded with a letter (a–l).

© 2012 The Linnean Society of London, Zoological Journal of the Linnean Society, 2012, 166, 361–376 366 P. M. S. NUNES ET AL. slender finger-shaped appendices present on apexes General description: Hemipenial body cylindrical; of lobes; sulcus spermaticus ending on apex of lateral surfaces of body and lateral margins of asul- lobular appendix. cate face covered by rows of flounces adorned by calcified spines; central area of asulcate face orna- Variation: Specimens from Campo Catuquira and mented by a wide stripe of scattered spines that Campo Tupana (both not sampled in the molecular become increasingly numerous towards apex; lobes analyses) lack spines on lateral surface of hemipenial small, lobular appendices absent. body, and the spines of asulcate face are not arranged in convergent rows, but scattered throughout its Variation: The number and size of spines on the central region. central area of the asulcate face vary among the Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 localities sampled. Morphotype 3 (Figs 1C, 2A: Candidate species 3) Localities of occurrence: BRAZIL: Amazonas: Manaus; São Sebastião (Rio Abacaxis); Igarapé-Açu MOLECULAR PHYLOGENETIC ANALYSES (Rio Abacaxis). The topology recovered from mtDNA is relatively well resolved, with the ingroup (genus Iphisa) having 12 of General description: Hemipenes cylindrical; lobes 21 nodes strongly supported by posterior probability small, apices ornamented by papillate flounces; (PP Ն 0.95) and eight of 21 nodes strongly supported lobular appendices absent; lateral surfaces covered by by maximum parsimony bootstrap (bootstrap support flounces with calcified spicules; central area of asul- Ն 95%) (Fig. 2A). Although levels of support are het- cate face ornamented by two simple rows of spines erogeneous among the methods used, there are no originating at base, converging until midpoint of body conflicts among topologies. Iphisa elegans was unam- and then assuming centrifugal orientation towards biguously recovered as monophyletic and, although lobe base. poorly resolved, the relationships among gymnoph- talmids match the topology of Rodrigues et al. (2009). Variation: In contrast to the pattern described above, The genetic diversity recovered within Iphisa is the spine rows present in the central area of the striking and is broadly congruent with the hemipenial asulcate face are not simple in one specimen from patterns (Fig. 2A). Each population harbours a highly Manaus (MZUSP 8354), but composed of irregularly divergent lineage, the only exception being the three aligned sets of spines. southern populations of clade 1A from the localities of UHE Guaporé, Apiacás, and Montenegro- Morphotype 4 (Figs 1D, 2A: Candidate species 4) Cacaulândia. As a matter of comparison, the degree of Localities of occurrence: BRAZIL: Amapá: Igarapé divergence (maximum p distance = 0.152) within Camaipi, Rio Maracá; Rondônia: UHE Jirau; Iphisa is comparable with the divergence among FRENCH GUIANA: Paracou, Sinnamary; GUYANA: related Iphisiini genera, which have remarkably dis- Kaieteur National Park. tinct general morphology (Appendix 4). The analyses recovered two major clades within General description: Hemipenial body cylindrical; Iphisa: Clade 1 occurs in southern Amazonia (from lobes small, lobular appendices lacking; lateral sur- Rio Abacaxis to Guaporé), and Clade 2 occurs over the faces covered by flounces with calcified spicules; remaining territory of Amazonia (from the Guiana central area of asulcate face mostly bare, ornamented Shield to Peru and Rondônia). Each of these clades by two parallel spine rows running in roughly sagital includes highly divergent and strongly supported sub- position. clades. Clade 1 displays a clear structure, although its two subclades show wide geographical overlap over Variation: One specimen from Igarapé Camaipi has a the Abacaxis and the Aripuanã basins. Clade 2 and its stripe of scattered spines along the midline of asul- subdivisions are not well supported. Nevertheless, cate face reaching proximal two-thirds of hemipenial given the geographical distribution of the subclades body; distally from this point, only two parallel rows (Peru vs. Guiana Shield and Rondônia) we considered remain towards apex with a bare area between them. these two groups in latter analyses. The genetic subdivisions detected are so pronounced that each of Morphotype 5 (Figs 1E, 2A: Candidate species 5) these subclades (1A, 1B, 2A, and 2B) is itself divided Localities of occurrence: ECUADOR: Morona- into multiple highly divergent lineages. Such genetic Santiago: Rampon; Napo: Puerto Libre, Rio Aguarico; structure reveals that our sampling remains insuffi- PERU: Amazonas: Rio Cenepa, Rio Marañon valley; cient to estimate the actual distribution of these lin- Cuzco: Pagoreni; San Martin; Distrito Genaro eages so that the degree of geographical overlap Herrera, Provincia Requena; Loreto: Rio Corrientes. among them may in fact be far more important.

Nevertheless, it is noteworthy that, in one case, two in delimiting cryptic species lies in distinguishing lineages (1Ba and 1Bb) within subclade 1B are syn- between broad-admixture, narrow contact zones with topic and that at least subclades 1A and 1B appear restricted hybridization and complete isolation (Wake largely sympatric and do not share any C-MOS allele & Jockusch, 2000), allowing assessment of the inde- (Fig. 2B). Clade 2A appears to cover a wide area from pendence of evolutionary trajectories of the entities. the Guiana Shield to Rondônia, but again the differ- Thus, delineation of species not only requires the use ent lineages involved show high degrees of genetic of a combination of multiple lines of evidence, but also divergence. Even the two populations sampled at very a thorough sampling to provide an accurate descrip- close localities in Peru (EEBB Pithecia and Ham- tion of the biodiversity. burgo, in Loreto) are highly divergent (CYTB p dis- The striking congruence between different types of Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 tance = 0.11; Appendix 4). characters (hemipenial, mitochondrial, and nuclear Interestingly, the nuDNA data provide a structure data) in I. elegans provides strong evidence support- that is concordant with the most basal mtDNA splits ing the existence of distinct species (Dayrat, 2005; within Iphisa. Subclade 1A displays a fixed substitution DeSalle, Egan & Siddal, 2005; Padial et al., 2010) and within subclade 1B the two groups do not share any that remained masked by overall homogeneous exter- allele. The only instances of allele sharing are among nal morphology. In addition, some of these candidate geographically distant populations and correspond to species distributions overlap geographically, with the central haplotype, which probably represents an syntopy observed in at least one locality (São ancestral state. Therefore, our data provide no evidence Sebastião, Rio Abacaxis). Broader sampling through- of gene flow among genetically differentiated popula- out the range of the genus will probably reveal more tions occurring in geographical proximity. instances of co-occurrence of distinct entities. The geographically overlapping mtDNA-based groups do not share any nuclear alleles, suggesting that these lineages are reproductively isolated. On the other DISCUSSION hand, the candidate species that are undistinguish- CRYPTIC SPECIES able on the basis of nuDNA and have more similar Inaccuracy in the evaluation of diversity has impor- hemipenial morphology were found in distant loca- tant ramifications, and thus a precise delimitation of tions and are probably geographically isolated. There- species is essential to many disciplines such as bio- fore, concordance between the criteria of coalescence geography, ecology, macroevolution, biodiversity and isolation, as coined by de Queiroz (1998), as well assessment, and conservation given that species are as two independent lines of evidence (DNA and mor- basic units of analysis. However, delineating species phology) suggest that several cryptic species exist remains a challenge. Cryptic species are detected under the name I. elegans and indicate the need of when one recognizes two or more distinct species reformulations towards a more informative taxonomic previously classified as a single species due to overall system for the genus Iphisa. morphological similarity that prevents immediate Our results indicate the existence of five major obvious distinction (Bickford et al., 2007; Pfenninger groups supported by molecular and hemipenial data, & Schwenk, 2007; Trontelj & Fišer, 2009). Recent contrasting with Dixon’s (1974) conclusion that the DNA-based studies (e.g. Fouquet et al., 2007; Koch genus is monotypic with no more than two distin- et al., 2009; Mott & Vieites, 2009; Geurgas & Rod- guishable varieties (i.e. subspecies). However, Dixon’s rigues, 2010; Oliver, Adams & Doughty, 2010; (1974) conclusions were based on a rather limited Hekkala et al., 2011; Morin et al., 2011; Wu et al., sampling and strictly on morphology. 2011) have demonstrated the existence of consider- Within the clades recovered here, the Peruvian ably divergent lineages ignored by taxonomic systems lineage designated as clade 2B corresponds to speci- solely based on morphological grounds, revealing mens that morphologically fit Dixon’s (1974) descrip- degrees of genetic divergence that reflect millions of tion of I. e. soinii. Unfortunately, specimens from the years of evolutionary history. Although morphology type locality of I. e. elegans [300-mile radius of Pará, may be of little use in revealing important historical Brazil, sensu Gray (1851); by Pará, Dixon (1974) divergences among cryptic species (Elmer, Dávila & probably refers to the municipality of Belém, Pará Lougheed, 2007; Koch et al., 2009; Geurgas & Rod- State] were not available for our analyses. However, rigues, 2010), morphological evidence remains crucial in our sample, the specimens from the closest locali- in their description and precise diagnoses (Hillis & ties to the Belém region (Pará state) are from the Wiens, 2000). Therefore, species limits are undoubt- left bank of the Amazon River and appear nested in edly better understood through the combination of clade 2A. The imprecise locality of the holotype and different kinds of information (Padial et al., 2010) the limited number of samples available for this such as DNA and morphology. Another crucial point study prevent any conclusion regarding the actual

© 2012 The Linnean Society of London, Zoological Journal of the Linnean Society, 2012, 166, 361–376 368 P. M. S. NUNES ET AL. correspondence between the nominal form of I. e. other squamate groups. Molecular investigations on elegans and any of the lineages detected herein. previously documented cases of intraspeciﬁc hemipe- Therefore, we believe that any nomenclatural actions nial morphological variation [e.g. Calamaria lumbri- regarding the I. elegans complex depend on a com- coidea (Inger & Marx, 1962)] and Siphlophis prehensive taxonomic revision with more extensive compressus (Zaher & Prudente, 1999)] may also sampling, as well as more detailed morphological reveal cryptic species and thus provide additional and molecular analyses. support for our interpretation regarding the presence of a complex of species under the name I. elegans.

PREMATING REPRODUCTIVE ISOLATION Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 The hemipenial variation within I. elegans is unusual ACKNOWLEDGMENTS for squamate species showing otherwise homogeneous We are grateful to D. Frost and D. Kizirian (AMNH), overall morphology (e.g. Uzzell, 1966, 1973; Arnold, K. de Queiroz, R. McDiarmid, R. Heyer and G. Zug 1986a, b; Zaher, 1999). Regarding gymnophthalmids, (USNM), W. Duellman and L. Trueb (KU), J. C. Chap- such a level of variation could be compared with the arro (MHNC), P. Venegas (CORBIDI), J. H. C. Santa- variation observed between different related genera Gádea and J. Suárez (MHNSM), S. M. Souza and R. (Presch, 1978; Nunes, 2011). Thus, it is striking that Vogt (INPA), and H. Zaher and C. Castro-Mello these hemipenial morphotypes are sympatric, with at (MZUSP) for providing access to specimens and tissue least four lineages occurring in very close geographi- samples under their care. We are also grateful to P. cal proximity, or even in syntopy [specimens belong- Hayward and two anonymous reviewers for the sug- ing to clades 1Ba and 1Bb are syntopic on the same gestions and criticism of the manuscript. P.J.R.K.’s bank of Rio Abacaxis, but are remarkably distinct fieldwork in Guyana was made possible with the with respect to hemipenial structure (see Figs 2A, 3)]. financial support of the Belgian Directorate-General In contrast, the allopatric lineages constituting clade of Development Cooperation and the help and support 2 display more similar hemipenes (Fig. 2). of the Prime Minister of Guyana, the Honorable Such results suggest that hemipenial morphology Samuel Hinds. P.J.R.K. thanks G. Seegobin, P. Ben- may be directly linked to the origin of this pattern. A jamin, H. Sambhu, F. Marco, R. Williams, and I. reasonable preliminary hypothesis for the origin of Roopsind for field assistance and M. Kalamandeen, K. such pattern could rely in a speciation process occur- Holder, and C. Bernard (University of Guyana) for ring through reinforcement of premating isolation as a help in obtaining export permits. P.J.R.K.’s research consequence of secondary contact between lineages permits in Guyana (180604BR011, 030605BR006) (Dobzhansky, 1940, 1951; Butlin, 1987; Liou & Price, and export permits (031204SP017, 191205SP011 and 1994; Hoskin et al., 2005). Although the basis of this 040406SP014) were issued by the Guyana Environ- process has been seriously questioned in recent mental Protection Agency. We are grateful to M. decades (Butlin, 1987, 1995, 2004), there remains Antunes, M. Concistré, M. Sena, and S. Baroni for strong support (Hoskin et al., 2005; Lemmon, 2009). support during laboratory procedures and D. Pavan, Speciation by reinforcement is based on prezygotic G. Skuk, J. Cassimiro, J. M. Ghellere, M. A. Sena, M. isolation between two or more populations previously Teixeira Jr, R. Recoder, S. M. Souza, V. Verdade, V. hybridizing, enhancing characters that decrease gene Xavier, and E. M. Freire for help in collecting speci- flow between them. According to Servedio & Noor mens. J. Cassimiro kindly revised an early version of (2003), if two populations have diverged to such an the manuscript. A.F., F.F.C., M.T.R. and P.M.S.N. extent that they produce unfit hybrids, one must were supported by Fundação de Amparo à Pesquisa expect that more successful offspring will result from do Estado de São Paulo (FAPESP) and Conselho individuals belonging to the same population; there- Nacional de Desenvolvimento Científico e Tecnológico fore, those characters increasing assortative mating (CNPq) (FAPESP: A.F. – grant 2009/51931–9, M.T.R. will be favoured until full speciation eventually takes – grant 2003/10335–8, P.M.S.N. – grant 2007/ place (Butlin, 1987). Nevertheless, in the case of Iphisa 00811–8; CNPq: F.F.C. – grant 2009.1.812.41.2). the ecological differences or characteristics favouring assortative mating remain uncertain, as habitat is apparently similar among groups. Therefore, this REFERENCES hypothesis implies that the differences in hemipenial Akaike H. 1974. A new look at the statistical model identi- general conformation may prevent hybridization fication. IEEE Transactions on Automatic Control 19: 716– among cryptic species exposed to secondary contact 723. and ultimately favoured speciation and range overlap. Amézquita A, Lima AP, Jehle R, Castellanos L, Ramos Similar patterns of hemipenial variation concor- Ó, Crawford AJ, Gasser H, Hödl W. 2009. Calls, colours, dant with genetic variability may be expected for shapes, and genes: a multi-trait approach to the study of

© 2012 The Linnean Society of London, Zoological Journal of the Linnean Society, 2012, 166, 361–376 CRYPTIC SPECIES IN IPHISA ELEGANS GRAY, 1851 369 Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021

Figure 3. A, localities of occurrence of the ﬁve hemipenial morphotypes of some Iphisa elegans specimens sampled herein (asterisks, morphotype 1; black squares, morphotype 2; white triangle, morphotype 3; white squares, morphotype 4; white circle, morphotype 5; white star, sympatric morphotypes 2 and 3); B, localities of the I. elegans specimens submitted to molecular analyses with clades illustrated on Figure 2 emphasized by polygons (1A, 1B and 2A) and an ellipse (2B). Common darkened area represents an approximation of the Amazonia geographical coverage; black numbers correspond to speciﬁc localities listed in Appendix 1.

geographic variation in the Amazonian frog, Allobates femo- their relation to geological and climatic models. In: Hoorn C, ralis. Biological Journal of the Linnean Society 98: 826–838. Wesselingh F, eds. Amazonia, landscape and species evolu- Antonelli A, Quijada-Mascareñas A, Crawford AJ, tion. Oxford: Blackwell Publishing, 386–404. Bates AJ, Velazco JM, Wüster W. 2010. Molecular Arnold EN. 1986a. Why copulatory organs provide so many studies and phylogeography of Amazonian tetrapods and useful taxonomic characters: the origin and maintenance of

hemipenial differences in lacertid lizards (Reptilia: Lac- frog, Eleutherodactylus ockendeni. BMC Evolutionary ertidae). Biological Journal of the Linnean Society 29: 263– Biology 7: 1–14. 281. Felsenstein J. 1985. Conﬁdence limits on phylogenies — an Arnold EN. 1986b. The hemipenis of lacertid lizards approach using the bootstrap. Evolution 39: 783–791. (Reptilia: Lacertidae): structure, variation and systematic Fouquet A, Gilles A, Vences M, Marty C, Blanc M, implications. Journal of Natural History 20: 1221–1257. Gemmell NJ. 2007. Underestimation of species richness in Avila-Pires TCS. 1995. Lizards of Brazilian Amazonia (Rep- Neotropical frogs revealed by mtDNA analyses. Plos ONE 2: tilia: Squamata). Zoologische Verhandelingen (Leiden) 299: 1–10. 1–706. Funk WC, Caminer M, Ron SR. 2012. High levels of cryptic Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier R, species diversity uncovered in Amazonian frogs. Proceedings

Winker K, Ingram KK, Das I. 2007. Cryptic species as a of the Royal Society B: Biological Sciences 279: 1806–1814. Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 window on diversity and conservation. Trends in Ecology & Gamble T, Daza JD, Colli GR, Vitt LJ, Bauer AM. 2011. Evolution (Personal Edition) 22: 148–155. A new genus of miniaturized and pug-nosed gecko from Bickham JW, Wood CC, Patton JC. 1995. Biogeographic South America (Sphaerodactylidae: Gekkota). Zoological implications of cytochrome b sequences and allozymes in Journal of the Linnean Society 163: 1244–1266. sockeye (Oncorhynchus nerka). The Journal of Heredity 86: Gaston KJ, Williams PH. 1996. Spatial patterns in taxo- 140–144. nomic diversity. In: Gaston KJ, ed. Biodiversity: a biology of Bush MB. 1994. Amazonian speciation — a necessarily numbers and difference. Oxford: Blackwell Science, 202– complex model. Journal of Biogeography 21: 5–17. 229. Butlin RK. 1987. Speciation by reinforcement. Trends in Geurgas SR, Rodrigues MT. 2010. The hidden diversity of Ecology & Evolution (Personal Edition) 2: 8–13. Coleodactylus amazonicus (Sphaerodactylinae, Gekkota) Butlin RK. 1995. Reinforcement: an idea evolving. Trends in revealed by molecular data. Molecular Phylogenetics and Ecology & Evolution (Personal Edition) 10: 432–434. Evolution 54: 583–593. Butlin RK. 2004. Mystery of mysteries no longer? Evolution Godinho R, Domingues V, Crespo EG, Ferrand N. 2006. 58: 243–245. Extensive intraspecific polymorphism detected by SSCP at Chek AA, Lougheed SC, Bogart JP, Boag PT. 2001. the nuclear Cmos gene in the endemic Iberian lizard Perception and history: molecular phylogeny of a diverse Lacerta schreiberi. Molecular Ecology 15: 731–738. group of neotropical frogs, the 30-chromosome Hyla (Anura: Gray JE. 1851. Description of a new genus and family of Hylidae). Molecular Phylogenetics and Evolution 18: 370– cyclosaurian lizards, from Para. Proceedings of Zoological 385. Society of London 19: 38–39. Clement M, Posada D, Crandall KA. 2000. TCS: a com- Harvey MB, Embert D. 2008. Review of Bolivian Dipsas puter program to estimate gene genealogies. Molecular (Serpentes: Colubridae), with comments on other South Ecology 9: 1657–1659. American species. Herpetological Monographs 22: 54–105. Cole CJ, Hardy LM. 1981. Systematics of North American Hekkala E, Shirley MH, Amato G, Austin JD, Charter S, colubrid snakes related to Tantilla planiceps (Blainville). Thorbjanarson J, Vliet KA, Houck ML, DeSalle R, Bulletin of the American Museum of Natural History 171: Blum MJ. 2011. An ancient icon reveals new mysteries: 201–284. mummy DNA resurrects a cryptic species within the Nile Dayrat B. 2005. Toward integrative taxonomy. Biological crocodile. Molecular Ecology 20: 4199–4215. Journal of the Linnean Society 85: 407–415. Hillis DM, Bull JJ. 1993. An empirical-test of bootstrapping DeSalle R, Egan MG, Siddal M. 2005. The unholy trinity: as a method for assessing confidence in phylogenetic analy- taxonomy, species delimitation and DNA barcoding. Philo- sis. Systematic Biology 42: 182–192. sophical Transactions of the Royal Society B 360: 1905– Hillis DM, Wiens JJ. 2000. Molecules versus morphology in 1916. systematics: conflicts, artifacts and misconceptions. In: Dixon JR. 1974. Systematic review of the microteiid genus Wiens JJ, ed. Phylogenetic analysis of morphological data. Iphisa. Herpetologica 30: 133–139. Washington, DC: Smithsonian Institution Press, 1–20. Dobzhansky T. 1940. Speciation as a stage in evolutionary Hoogmoed MS. 1973. Notes on the herpetofauna of Surinam divergence. American Naturalist 74: 312–321. IV. The lizards and amphisbaenians of Surinam. Biogeo- Dobzhansky T. 1951. Genetics and the origin of species, 3rd graphica 4: i–ix + 1–419. edn. New York: Columbia University Press. Hoorn C, Wesselingh FP, ter Steege H, Bermudez MA, Dowling HG. 1967. Hemipenes and other characters in colu- Mora A, Sevink J, Sanmatín I, Sanchez-Meseguer A, brid classification. Herpetologica 23: 138–142. Anderson CL, Figueiredo JP, Jaramillo C, Riff D, Dowling HG, Savage JM. 1960. A guide to the snake hemi- Negri FR, Hooghiemstra H, Lundberg J, Stadler T, penis: a survey of basic structure and systematic character- Särkinken T, Antonelli A. 2010. Amazonia through time: istics. Zoologica 45: 17–28. Andean uplift, climate change, landscape evolution, and Efron B. 1979. 1977 Rietz Lecture—bootstrap methods— biodiversity. Science 330: 927–931. another look at the jackknife. Annals of Statistics 7: 1–26. Hoskin CJ, Higgie M, Mcdonald KR, Moritz C. 2005. Elmer KR, Dávila JA, Lougheed SC. 2007. Cryptic diver- Reinforcement drives rapid allopatric speciation. Nature sity and deep divergence in an upper Amazonian leaflitter 437: 1353–1356.

Huelsenbeck JP, Ronquist F. 2001. MRBAYES: Bayesian Venezuelan Guayana, with notes on ‘microteiid’ hemipenes inference of phylogenetic trees. Bioinformatics 17: 754–755. (Squamata: Gymnophthalmidae). American Museum Novi- Inger RF, Marx H. 1962. Variation of hemipenis and clo- tates 3660: 1–31. acain the colubrid snake Calamaria lumbricoidea. System- Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca atic Zoology 11: 32–38. GAB, Kent J. 2000. Biodiversity hotspots for conservation Keogh S. 1999. Evolutionary implications of hemipenial mor- priorities. Nature 403: 853–858. phology in the terrestrial Australian elapid snakes. Zoologi- Noonan BP, Wray KP. 2006. Neotropical diversiﬁcation: the cal Journal of the Linnean Society 125: 239–278. effects of a complex history on diversity within the poison frog Koch A, Arida E, Schmitz A, Böhme W, Ziegler T. 2009. genus Dendrobates. Journal of Biogeography 33: 1007–1020. Reﬁning the polytypic species concept of mangrove monitors Nunes PMS. 2011. Morfologia hemipeniana dos lagartos

(Squamata: Varanus indicus group): a new cryptic species microteídeos e suas implicações nas relações filogenéticas da Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 from the Talaud Islands, Indonesia, reveals the underesti- família Gymnophthalmidae (Squamata: Teiioidea) – Volume mated diversity of Indo-Australian monitor lizards. Austra- 1 (Text). Unpublished D. Phil. Thesis, Universidade de São lian Journal of Zoology 57: 29. Paulo. Kocher TD, Thomas WK, Meyer A, Edwards SV, Pääbo S, Nylander JA. 2004. MrModeltest v2. Program distributed Villablanca FX, Wilson AC. 1989. Dynamics of mitochon- by the author. Evolutionary Biology Centre, Uppsala drial DNA evolution in animals – amplification and sequenc- University. ing with conserved primers. Proceedings of the National Oliver PM, Adams M, Doughty P. 2010. Molecular evidence Academy of Sciences of the United States of America 86: for ten species and Oligo-Miocene vicariance within a 6196–6200. nominal Australian gecko species (Crenadactylus ocellatus, Lemmon EM. 2009. Diversification of conspecific signals in Diplodactylidae). BMC Evolutionary Biology 10: 1–11. sympatry: geographic overlap drives multidimensional Padial JM, Miralles A, de la Riva I, Vences M. 2010. The reproductive character displacement in frogs. Evolution 63: integrative future of taxonomy. Frontiers in Zoology 7: 1–14. 1155–1170. Pesantes OS. 1994. A method for preparing the hemipenis of Librado P, Rozas J. 2009. DnaSP v5: a software for com- preserved snakes. Journal of Herpetology 28: 93–95. prehensive analysis of DNA polymorphism data. Bioinfor- Peters JA, Donoso-Barros R. 1986. Catalogue of the Neotro- matics 25: 1451–1452. pical Squamata: Part II. Lizards and amphisbaenians. Revi- Liou LW, Price TD. 1994. Speciation by reinforcement of sed edition. Washington, DC: Smithsonian Institution Press. premating isolation. Evolution 48: 1451–1459. Pfenninger M, Schwenk K. 2007. Cryptic animal species Manzani PR, Abe AS. 1988. Sobre dois novos métodos de are homogeneously distributed among taxa and biogeo- preparo do hemipenis de serpentes. Memorias do Instituto graphical regions. BMC Evolutionary Biology 7: 21. Butantan (São Paulo) 50: 15–20. Pope CH. 1941. Copulatory adjustment in snakes. Field McDowell SB. 1979. A catalogue of the snakes of New Guinea Museum of Natural History, Zoology. Series 24: 249–252. and the Solomons, with special reference to those in the Presch W. 1978. Descriptions of the hemipenial morphology Bernice P. Bishop Museum. III. Boinae and Acrochordoidea in eight species of microteiid lizards (Family Teiidae, Sub- (Reptilia; Serpentes). Journal of Herpetology 13: 1–92. family Gymnophthalminae). Herpetologica 34: 108–112. Meyer A. 1993. Phylogenetic relationships and evolutionary Prudente ALC, Passos P. 2010. New cryptic species of processes in East African cichlid fishes. Trends in Ecology Atractus (Serpentes: Dipsadidae) from Brazilian Amazonia. and Evolution 8: 279–284. Copeia 2010: 397–404. Miralles A, Vasconcelos R, Pereira A, Harris DJ, Car- de Queiroz K. 1998. The general lineage concept of species, ranza S. 2010. An integrative taxonomic revision of the species criteria, and the process of speciation. In: Howard Cape Verdean skinks (Squamata, Scincidae). Zoologica DJ, Berlocher SH, eds. Endless forms: species and specia- Scripta 40: 16–44. tion. Oxford: Oxford University Press, 57–75. Morin PA, Archer FI, Foote AD, Vilstrup J, Allen EE, Rambaut A, Drummond A. 2003. Tracer, version 1.2. Avail- Wade P, Durban J, Parsons K, Pitman R, Li L, Bouffard able at: http://tree.bio.ed.ac.uk/software/tracer/ P, Nielsen SCA, Rasmussen M, Willerslev E, Gilbert Rodrigues MT, Cassimiro J, Pavan D, Curcio FF, MTP, Harkins T. 2011. Complete mitochondrial genome Verdade VK, Pellegrino KC. 2009. A new genus of phylogeographic analysis of killer whales (Orcinus orca) microteiid lizard from the Caparaó Mountains, southeastern indicates multiple species. Genome Research 20: 908–916. Brazil, with a discussion of relationships among Gymnoph- Mott T, Vieites DR. 2009. Molecular phylogenetics reveals thalminae (Squamata). American Museum Novitates 3673: extreme morphological homoplasy in Brazilian worm lizards 1–27. challenging current taxonomy. Molecular Phylogenetics and Saint KM, Austin CC, Donnellan SC, Hutchinson MN. Evolution 51: 190–200. 1998. C-mos, a nuclear marker useful for squamate phylo- Myers CW. 1974. The systematics of Rhadinaea (Colubridae), genetic analysis. Molecular Phylogenetics and Evolution 10: a genus of New World snakes. Bulletin of the American 259–263. Museum of Natural History 153: 1–262. Servedio MR, Noor MAF. 2003. The role of reinforcement in Myers CW, Rivas Fuenmayor G, Jadin RC. 2009. New speciation: theory and data. Annual Review of Ecology, species of lizards from Auyantepui and La Escalera in the Evolution, and Systematics 34: 339–364.

Stephens M, Donnelly P. 2003. A comparison of Bayesian Wake DB, Jockusch EL. 2000. Detecting species borders methods for haplotype reconstruction. American Journal of using diverse data sets. In: Bruce RC, Jaeger RG, Houck Human Genetics 73: 1162–1169. LD, eds. The biology of plethodontid salamanders. Amster- Stephens M, Smith NJ, Donnelly P. 2001. A new statistical dam: Kluwer Academic/Plenum Publishers, 95–119. method for haplotype reconstruction from population data. Wilson EO. 1992. The diversity of life. Cambridge, MA: American Journal of Human Genetics 68: 978–989. Belknap Press of Harvard University Press. Swofford DL. 1993. PAUP – a computer-program for phylo- Wu H-C, Lin R-C, Hung H-Y, Yeh C-F, Chu J-H, genetic inference using maximum parsimony. The Journal Yang X-J, Yao C-J, Zou F-S, Yao C-T, Li S-H, Lei F-M. of General Physiology 102: 9A. 2011. Molecular and morphological evidences reveal a Trontelj P, Fišer C. 2009. Cryptic species diversity should cryptic species in the vinaceous roseﬁnch Carpodacus vina-

not be trivialised. Systematics and Biodiversity 7: 1–3. ceus (Fringillidae; Aves). Zoologica Scripta 40: 468–478. Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 Uzzell T. 1966. Teiid lizards of the genus Neusticurus (Rep- Zaher H. 1999. Hemipenial morphology of the South Ameri- tilia, Sauria). Bulletin of the American Museum of Natural can xenodontine snakes, with a proposal for a monophyletic History 132: 277–328. Xenodontinae and a reappraisal of colubroid hemipenes. Uzzell T. 1973. A revision of lizards of the genus Prionodac- Bulletin of the American Museum of Natural History 240: tylus, with a new genus for P. leucostictus and notes on the 1–168. genus Euspondylus (Sauria, Teiidae). Postilla 159: 1–67. Zaher H, Prudente ALC. 1999. Intraspeciﬁc variation of the Vogel G. 2001. African elephant species splits in two. Science hemipenis in Siphlophis and Tripanurgos. Journal of Her- 293: 1414. petology 33: 698–702.

APPENDIX 1 Locality information and specimens of Iphisa elegans and outgroups sampled in this study. Locality numbers correspond to Fig. 3 and clade numbers correspond to Figs 2 and 3B. Type of data is represented by: G (genetic) and M# (morphological and the respective hemipenial morphotype). Voucher numbers are represented by field and/or institutional identification numbers. Accession numbers of the previously published sequences refer to specific individuals and are denoted by an asterisk, and the accession numbers of specimens analysed in this study refer to the alleles. Brazilian state abbreviations are as follows: AM, Amazonas; AP, Amapá; BA, Bahia; CE, Ceará; GO, Goiás; MT, Mato Grosso; PA. Pará; RO, Rondônia; SP, São Paulo.

Voucher Type Species Locality/Coordinates of data Field no. Institutional no. Clade Cytb Cmos

Iphisa e. elegans (1) Apiacás, MT, Brazil G/M1 968293 MZUSP 81634 1A JX079892 a/a JX079870 (09°34′10″S, 57°21′04″W) Iphisa e. elegans (2) Aripuanã, MT, Brazil M1 – MZUSP 82662 – (10°19′00″S, 59°27′34″W) M1 – MZUSP 82666 – G/M1 977413 MZUSP 82655 1A JX079894 a/a JX079872 G 977426 MZUSP 82656 1A JX079895 a/a JX079873 G/M1 977669 MZUSP 82669 1A JX079896 a/a JX079874 (3) Cachoeira das Pombas, G MTR 10210 1Ba JX079902 JX079880 AM, Brazil (06°24′00″S, 60°21′00″W) (4) Cachoeirinha, Rio G RCV 2225 – 1A JX079907 a/a JX079885 Madeira, AM, Brazil (05°29′16″S, 68°48′23″W) (5) Campo Catuquira, AM, M2 SMS 197 INPA 20313 – Brazil (04°54′16″S, M2 SMS 208 INPA 20326 – 61°06′46″W) M2 SMS 209 INPA 20298 – Iphisa e. elegans (6) Campo Tupana, AM, M2 SMS 092 INPA 20302 – Brazil (04°09′37″S, 60°07′51″W) Iphisa elegans (7) Centro de M5 – MHNSM 16718 – soinii Investigaciones del IIAP, Distrito Genaro Herrera, Provincia Requena, Peru (04°54′17″S, 73°39′11″W) Iphisa e. elegans (8) Comunidade Projó, Alto M2 SMS 028 INPA 18445 – Rio Aripuanã, AM, Brazil Iphisa e. elegans M2 SMS 022 INPA 18432 – (07°37′01″S, 60°40′54″W)

APPENDIX 1 Continued

Voucher Type Species Locality/Coordinates of data Field no. Institutional no. Clade Cytb Cmos

Iphisa e. elegans (9) G Uniban – 2B JX079910 a/a JX079888 Montenegro-Cacaulândia 1704 border, RO, Brazil (10°19′53″S, 63°05′24″W) Iphisa elegans (10) EEBB Pithecia, Rio G MHNC – 2B JX079901 b/d Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 soinii Samiria, Dist. Parinari, 10115 JX079879 Loreto, Peru (05°13′51″S, 74°37′40″W) Iphisa e. elegans (11) Essequibo Co., M4 – AMNH 21294 – Kartabo, Kuyuwini Landing, Guyana (06°21′00″ N 58°41′00″ W) Iphisa elegans (12) Hamburgo, Rio G MHNC – 2B JX079900 e/f JX079878 soinii Samiria, Dist. Parinari, 10082 Loreto, Peru (05°14′26″S, 75°07′09″W) Iphisa e. elegans (13) Igarapé Camaipi, AP, G/M4 LG 1786 MZUSP 88464 2A JX079898 b/b JX079876 Brazil (00°01′00″S, 51°42′00″W) Iphisa e. elegans (14) Igarapé-Açu, Rio M2 MTR 12758 MZUSP 100251 – Abacaxis, AM, Brazil G/M3 MTR 12772 MZUSP 100256 1Bb JX079911 j/k JX079889 (04°20′39″S, 58°38′06″W) M3 MTR 12867 MZUSP 100257 – G/M3 MTR 12892 MZUSP 100258 1Bb JX079913 i/j JX079891 M3 MTR 12896 MZUSP 100259 – M2 MTR 12913 MZUSP 100260 – Iphisa e. elegans (15) Interﬂuve of Rios M2 SMS 201 INPA 20331 – Madeira and Purus, AM, Brazil (04°59′30″S, 61°07′09″W) Iphisa e. elegans (16) Itapinima, AM, Brazil G RCV 2247 – 1A JX079908 a/a JX079886 (05°24′37″S, 60°43′16″W) Iphisa e. elegans (17) Juruena, MT, Brazil G/M2 976915 MZUSP 82428 1Ba JX079893 JX079871 (10°19′05″S, 58°21′32″W) Iphisa e. elegans (18) Kaieteur National M4 – IRSNB 17069 – Iphisa e. elegans Park, Guyana G/M4 PK 1412 – 2A JX079905 b/b JX079883 Iphisa e. elegans (05°11′00″S, 59°28′00″W) G/M4 PK 1413 – 2A JX079906 b/b JX079884 Iphisa e. elegans M4 PK 1564 – – Iphisa e. elegans (19) Left margin Rio M2 SMS 008 INPA 18458 – Machado, RO, Brazil (08°10′37″S, 62°48′56″W) Iphisa e. elegans (20) Manaus, AM, Brazil M3 – MZUSP 8354 – (03°06′24″S, 60°01′32″W) Iphisa e. elegans (21) mouth Rio Cenepa, M5 – AMNH 56223 – Peru (04°35′00″S, 72°12′00″W) Iphisa elegans (22) Pagoreni, Distrito M5 – MHNSM 29541 – soinii Echarate, Provincia La Convencion, Cuzco, Peru (11°47′09″S, 72°42′05″W) Iphisa e. elegans (23) Paracou, French M4 – AMNH 139958 – Guiana (05°16′31″N, 52°55′25″W) Iphisa e. elegans (24) Puerto Libre, Rio M5 – KU 122173 – Aguarico, Ecuador (00°04′50″N, 76°47′30″W)

APPENDIX 1 Continued

Voucher Type Species Locality/Coordinates of data Field no. Institutional no. Clade Cytb Cmos

Iphisa e. elegans (25) Rampon, near M5 – USNM 196107 – Chiguaza, Ecuador (02°01′00″S, 77°58′00″W) Iphisa e. elegans (26) Rio Ampiyacu, M1 – MZUSP 13964 – Estirón, Peru Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 (03°19′03″S, 71°51′00″W) Iphisa e. elegans (27) Rio Corrientes , M5 – CORBIDI 2684 – Loreto, Peru (03°03′54″S, 75°49′35″W) Iphisa e. elegans (28) San Martin, Cuzco, M5 – USNM 538401 – Peru (11°47′08″S, 72°41′57″W) Iphisa e. elegans (29) São Sebastião, AM, G/M2 MTR 12750 MZUSP 100249 1Ba JX079903 g/g JX079881 Brazil (04°18′32″S, G MTR 12785 MZUSP 100252 1Bb JX079912 i/l JX079890 58°38′11″W) G/M2 MTR 12821 MZUSP 100253 1Ba JX079904 g/h JX079882 Iphisa e. elegans (30) UHE Guaporé, MT, G RGL 1781 – 1A JX079909 a/a JX079887 Brazil (15°07′00″S, 58°58′00″W) Iphisa e. elegans (31) UHE Jirau, RO, G H503 MZUSP 100247 2A JX079897 b/b JX079875 Brazil (09°37′38″S, 65° 26′46″W) Iphisa e. elegans (32) Vai-Quem-Quer, PA, G LG 744 – 2A JX079899 b/c JX079877 Brazil (01°30′00″S, 55°50′00″W) Acratosaura Morro do Chapéu, BA, G MTR – OG JX079915 mentalis Brazil (11°33′00″S, 906448 41°09′22″W) Alexandresaurus Ilhéus, BA, Brazil G MD304 – OG JX079917 camacan (14°47′20″S, 39°02′58″W) Caparaonia Parque Nacional Caparaó, G MTR 10852 – OG JX079916 itaiquara MG, Brazil (20°28′00″S, 41°49′00″W) Colobodactylus Campos do Jordão, SP, G LG761 – OG JX079918 dalcyanus Brazil (22°44′22″S, 45°35′29″W) Colobodactylus PE Ilha do Cardoso, SP, G 3444 – OG JX079919 taunayi Brazil (25°07′50″S, 48°58′12″W) Colobosaura Serra da Mesa, GO, Brazil G LG1145 – OG JX079920 modesta (13°50′02″S, 48°18′10″W) Heterodactylus Serra da Cantareira, SP, G 1504 – OG JX079921 imbricatus Brazil (23°26′00″S, 46°38′47″W) Stenolepis ridleyi Ibiapaba, CE, Brazil G LG2123 – OG JX079914 (05°02′58″S, 40°55′20″W)

APPENDIX 2 Details of the primers used with names, sequence and original reference.

Sequence Reference

Cytb H15149 TGCAGCCCCTCAGAATGATATTTGTCCTCA Kocher et al. (1989) CYB-02H AAACTGCAGCCCCTCAGAATGATATTTGTCCTCA Kocher et al. (1989) CYB-05L GCCAACGGCGCATCCTTCTTCTT Meyer (1993) LGL765 GAAAAACCAYCGTTGTWATTCAACT Bickham et al., 1995 Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 Cmos LSCH1 CTCTGGKGGCTTTGGKKCTGTSTACAAGG Godinho et al. (2006) LSCH2 GGTGATGGCAAARGAGTAGATGTCTGC Godinho et al. (2006) G73 GCGGTAAAGCAGGTGAAGAAA Saint et al. (1998) G74 TGAGCATCCAAAGTCTCCAATC Saint et al. (1998)

APPENDIX 3 Models used for each partition in the Bayesian analysis.

Partition Model Base frequencies Nst Rmat Rates Shape Pinvar

1 Cytb GTR + I + G 0.2912 0.2506 0.2084 6 3.5044 33.3230 3.5775 Gamma 0.4434 0.4986 2.2249 74.1059 2 Cytb HKY + I 0.1970 0.2623 0.1474 2 TRatio = 2.9587 Equal NA 0.8113 3 Cytb GTR + I + G 0.4140 0.3034 0.0539 6 0.2352 9.3226 0.1672 Gamma 2.1566 0.0071 0.3785 4.7455 Total HKY + I + G 0.3408 0.2957 0.0937 2 TRatio = 6.7942 Gamma 1.3596 0.5238

© 2012 The Linnean Society of London, Zoological Journal of the Linnean Society, 2012, 166, 361–376 376 .M .NUNES S. M. P. APPENDIX 4 Pairwise p distance among samples using pairwise deletion. The eight highest distances (>15%) within Iphisa are indicated in bold and in filled grey cells. The portion of the table that includes comparisons among outgroups is also indicated with filled grey cells. Above the diagonal matrix another table includes mean p distance among major clades as defined in the text. OUT (for outgroup) only comprises 02TeLnenSceyo London, of Society Linnean The 2012 © Iphisiini. TAL ET 968293 976975 0.137 977413 0.085 0.130 Clade . 977426 0.086 0.134 0.005 1A 977669 0.084 0.132 0.003 0.003 1B 0.130 IE503 0.124 0.136 0.133 0.134 0.132 2A 0.135 0.143 LG1786 0.138 0.144 0.133 0.133 0.133 0.093 2B 0.135 0.138 0.119 LG744 0.142 0.152 0.137 0.141 0.138 0.109 0.097 OUT 0.183 0.191 0.180 0.184 MNHC10082 0.132 0.136 0.128 0.129 0.126 0.113 0.113 0.134 MNHC10115 0.137 0.137 0.137 0.138 0.136 0.109 0.125 0.112 0.110 MTR10210 0.140 0.056 0.133 0.134 0.132 0.137 0.149 0.150 0.146 0.150 MTR12750 0.145 0.094 0.130 0.133 0.130 0.133 0.144 0.149 0.141 0.142 0.098 MTR12772 0.126 0.113 0.110 0.114 0.112 0.138 0.140 0.132 0.130 0.129 0.110 0.113 MTR12892 0.122 0.112 0.106 0.110 0.108 0.138 0.137 0.129 0.129 0.129 0.109 0.112 0.004 MTR12785 0.120 0.117 0.125 0.128 0.125 0.140 0.144 0.144 0.142 0.144 0.125 0.120 0.088 0.089 olgclJunlo h ina Society Linnean the of Journal Zoological MTR12821 0.145 0.094 0.130 0.133 0.130 0.133 0.144 0.149 0.141 0.142 0.098 0.000 0.113 0.112 0.120 PK1412 0.134 0.136 0.132 0.136 0.133 0.105 0.097 0.113 0.117 0.125 0.145 0.152 0.146 0.144 0.149 0.152 PK1413 0.134 0.136 0.132 0.136 0.133 0.105 0.097 0.113 0.117 0.125 0.145 0.152 0.146 0.144 0.149 0.152 0.000 RCV2225 0.092 0.152 0.108 0.109 0.106 0.140 0.152 0.141 0.142 0.145 0.149 0.146 0.126 0.125 0.132 0.146 0.144 0.144 RCV2247 0.064 0.134 0.093 0.094 0.092 0.120 0.126 0.120 0.132 0.134 0.129 0.145 0.108 0.109 0.106 0.145 0.129 0.129 0.080 RGL1781 0.025 0.133 0.085 0.086 0.084 0.120 0.142 0.146 0.133 0.136 0.136 0.145 0.124 0.125 0.118 0.145 0.138 0.138 0.093 0.060 UNIBAN1704 0.027 0.133 0.085 0.086 0.084 0.120 0.142 0.145 0.133 0.136 0.136 0.144 0.124 0.125 0.118 0.144 0.138 0.138 0.092 0.061 0.004 Acratosaura 0.193 0.194 0.171 0.172 0.169 0.185 0.190 0.204 0.193 0.197 0.207 0.206 0.197 0.194 0.209 0.206 0.184 0.184 0.197 0.185 0.187 0.187 mentalis Alexandresaurus 0.171 0.183 0.179 0.183 0.179 0.177 0.171 0.186 0.175 0.175 0.203 0.186 0.184 0.183 0.190 0.186 0.179 0.179 0.173 0.175 0.190 0.190 0.181 Colobosaura 0.192 0.185 0.165 0.165 0.162 0.159 0.177 0.192 0.182 0.196 0.180 0.176 0.177 0.176 0.189 0.176 0.177 0.177 0.177 0.188 0.195 0.193 0.171 0.194 modesta Caparaonia 0.198 0.200 0.191 0.194 0.191 0.197 0.201 0.194 0.196 0.190 0.200 0.196 0.186 0.187 0.190 0.196 0.194 0.194 0.180 0.191 0.196 0.194 0.188 0.169 0.183 itaquaira Stenolepis 0.191 0.206 0.190 0.193 0.190 0.159 0.178 0.193 0.175 0.182 0.203 0.190 0.194 0.193 0.194 0.190 0.170 0.170 0.178 0.194 0.186 0.186 0.154 0.176 0.151 0.190 ridleyi Colobodactylus 0.196 0.203 0.199 0.200 0.197 0.204 0.211 0.206 0.196 0.196 0.201 0.204 0.204 0.203 0.200 0.204 0.191 0.191 0.201 0.200 0.203 0.201 0.201 0.184 0.191 0.180 0.190

2012, , dalcyanus Colobodactylus 0.218 0.211 0.209 0.210 0.207 0.206 0.217 0.223 0.203 0.207 0.210 0.191 0.215 0.214 0.199 0.191 0.223 0.223 0.227 0.213 0.218 0.217 0.206 0.179 0.198 0.186 0.186 0.129 taunayi

166 Heterodactylus 0.197 0.209 0.183 0.186 0.183 0.206 0.195 0.198 0.191 0.192 0.200 0.195 0.183 0.181 0.202 0.195 0.208 0.208 0.192 0.192 0.205 0.203 0.192 0.174 0.184 0.142 0.178 0.189 0.181

imbricatus

361–376 , Downloaded from https://academic.oup.com/zoolinnean/article/166/2/361/2629180 by guest on 23 September 2021 September 23 on guest by https://academic.oup.com/zoolinnean/article/166/2/361/2629180 from Downloaded