For Review Only

Journal of Zoological Systematics and Evolutionary Research

Systematic review of the polychromatic ground snakes Atractus snethlageae complex reveals four new species from threatened environments

Journal: Journal of Zoological Systematics and Evolutionary Research ManuscriptFor ID JZS.202000046.R3 Review Only Wiley - Manuscript type: Original Article

Date Submitted by the 16-Nov-2020 Author:

Complete List of Authors: Melo-Sampaio, Paulo R.; Universidade Federal do Rio de Janeiro Museu Nacional, Departamento de Vertebrados Passos, Paulo; Universidade Federal do Rio de Janeiro Museu Nacional, Vertebrados Prudente, Ana Lúcia; Museu Paraense Emilio Goeldi, Coordenação de Zoologia Venegas, Pablo; Centro de Ornitología y Biodiversidad (CORBIDI), División de Herpetología Torres-Carvajal , Omar; Pontificia Universidad Catolica del Ecuador, Biology

Keywords: Amazonia, Andes, endemism, hemipenis, molecular phylogeny

We review the Atractus snethlageae species complex based on the examination of 330 specimens throughout its entire distribution, including its type-series. We redefine A. snethlageae and recognize four new species previously assigned to it in the literature and natural history collections. Two of them are diagnosed through both phenotypic (meristic, morphometric, colour pattern and male genital structure) and molecular (phylogeny) evidence, while the other two are recognized on Abstract: the basis of morphological characters only. We show that some Amazonian lowland species have more restricted ranges. The area covering the eastern portion of the state of Pará and western portion of the state of Maranhão in Brazil harbours restricted endemism for Atractus. This region is the most threatened within Amazonia and it supports the biogeographic importance of that area. Finally, we discuss the expected changes in the taxonomy of these ground snakes with more robust hypotheses based on well-sampled phylogenies.

Page 1 of 81 Journal of Zoological Systematics and Evolutionary Research

1 Systematic review of the polychromatic ground snakes Atractus snethlageae 2 complex reveals four new species from threatened environments 3 4 Short running title: Systematics of Atractus snethlageae 5 6 Authors: Paulo R. Melo-Sampaio1*, Paulo Passos1, Ana L.C. Prudente2, Pablo J. 7 Venegas3 and Omar Torres-Carvajal4 8 9 1 Departamento de Vertebrados, Museu Nacional, Universidade Federal do Rio de 10 Janeiro, São Cristóvão, 20940-040, Rio de Janeiro, Rio de Janeiro, Brazil. 11 2 Laboratório de Herpetologia,For CoordenaçãoReview de Zoologia, Only Museu Paraense Emílio 12 Goeldi, CP 399, 66040-170, Belém, Pará, Brazil. 13 3 División de Herpetología-Centro de Ornitología y Biodiversidad (CORBIDI), Santa 14 Rita 105, of. 102, Huertos de San Antonio, Surco, Lima, Peru. 15 4 Museo de Zoología, Escuela de Ciencias Biológicas, Pontificia Universidad Católica 16 del Ecuador, Avenida 12 de Octubre y Roca, Apartado 17-01-2184, Quito, Ecuador. 17 18 * Corresponding author: Paulo Roberto Melo-Sampaio, e-mail: 19 [email protected] 20 21 Key words: Amazonia, Andes, endemism, hemipenis, molecular phylogeny. 22 Journal of Zoological Systematics and Evolutionary Research Page 2 of 81

23 Abstract: We review the Atractus snethlageae species complex based on the 24 examination of 330 specimens throughout its entire distribution, including its type- 25 series. We redefine A. snethlageae and recognize four new species previously assigned 26 to it in the literature and natural history collections. Two of them are diagnosed through 27 both phenotypic (meristic, morphometric, colour pattern and male genital structure) and 28 molecular (phylogeny) evidence, while the other two are recognized on the basis of 29 morphological characters only. We show that some Amazonian lowland species have 30 more restricted ranges. The area covering the eastern portion of the state of Pará and 31 western portion of the state of Maranhão in Brazil harbours restricted endemism for 32 Atractus. This biogeographically important region is also the most threatened within 33 Amazonia. Finally, weFor discuss Reviewthe expected changes Only in the taxonomy of ground snakes 34 with more robust hypotheses based on well-sampled phylogenies.

36 1 INTRODUCTION 37 The tree of life is being pruned by human activities at an unprecedented rate (Pimm et 38 al., 2014). Habitat loss and species extinction are alarming issues worldwide, especially 39 in the Neotropical region, where the largest continuous tropical forests and one of the 40 most diverse biotas occur (Wilson, 1988). High rates of extinction (hundreds to millions 41 of species) are expected in forthcoming years being surpassed only by catastrophic 42 events (Pimm et al., 2014). In recent years, many species have been described from the 43 Amazonian lowlands to the Tropical Andes (Blackburn, Giribet, Soltis, & Stanley, 44 2019). These intrinsically connected ecosystems are ranked among the likely to lose 45 most vertebrates as a result of current deforestation rates (Scheffers, Oliveira, Lamb, & 46 Edwards, 2019). To make this bleak scenario even worse, many species including 47 reptiles have suffered with unprecedent extinction rates in the last century (Ceballos et 48 al., 2015), mainly on the tropical ecosystems due to a combination of complex factors 49 (Ferrer-Paris et al., 2019).

50 Neotropical groundsnakes of the genus Atractus Wagler 1830 are the most 51 diversified, species-rich and taxonomically complex radiation of the suborder Serpentes 52 (Passos, Kok, Albuquerque, & Rivas, 2013). The genus comprises 143 currently 53 recognized species (Uetz, Freed, & Hosek, 2020). However, many species remain 54 known only from their type-series or from geographically restricted regions (Passos, 55 Scanferla, Melo-Sampaio, Brito, & Almendáriz, 2019). Consequently, some important Page 3 of 81 Journal of Zoological Systematics and Evolutionary Research

56 aspects related to geographic, ontogenetic, sexual dimorphism and population variation 57 have not yet been properly documented (Passos, Cisneros-Heredia, Rivera, Aguilar, & 58 Schargel, 2012). For that reason, the taxonomic status or species boundaries in some 59 taxa remains unsatisfactorily evaluated (Passos, Chiesse, Torres-Carvajal, & Savage, 60 2010), which is of special concern considering the high level of restricted endemism 61 and polychromatism reported in Atractus. 62 Systematic and taxonomic studies on Neotropical reptiles are of great

63 importance for their conservation. Twenty percent of Neotropical reptile species are 64 threatened, with the same proportion catalogued as Data Deficient (Böhm et al., 2013). 65 Moreover, species with restricted ranges (geographical or altitudinal), highly specialized 66 diet (earthworms), andFor limited Reviewactivity patterns (concentrated Only on daytime and twilight) 67 are sensitive to local extinctions (Ceballos et al., 2015). A global evaluation of trade risk 68 across the vertebrate tree of life pointed out the genus Atractus as potentially under 69 highest-risk for trade in the near future to pet activity (Scheffers et al., 2019). Therefore, 70 we must settle the Linnaean shortfall of this emblematic group of snakes in order to get 71 a better picture of its diversity and species-level conservation status. 72 Although some species groups were proposed for the genus over time (Passos, 73 Echevarría, & Venegas, 2013; Passos, Fernandes, Bérnils, & Moura-Leite, 2010c; 74 Passos, Mueses-Cisneros, Lynch, & Fernandes, 2009; Passos et al., 2016; Savage, 75 1960), only the A. elaps and A. roulei groups have been, consistently, recovered as 76 monophyletic through all phylogenetic hypotheses published to date (see Passos et al., 77 in press). Taxonomic history of banded Atractus began with the description of 78 Brachyorrhos badius, B. flammigerus and B. schach from Guianan region by Boie 79 (1827). Duméril, Bibron, & Duméril (1854) synonymized B. flammigerus and B. schach 80 with B. badius until Hoogmoed (1980) rediscovered the types of Atractus badius, A. 81 flammigerus and A. schach, resurrecting the latter two from the synonymy with A. 82 badius. Later, Hoogmoed (1982) conceived a widespread distribution for A. 83 flammigerus, including portions outside the Guiana Shield. Cunha & Nascimento 84 (1983) described Atractus flammigerus snethlageae based on Brazilian specimens from 85 north and south of the Amazon River. Cunha & Nascimento (1984) also reported 86 additional specimens of Atractus schach from eastern Pará and western Maranhão 87 states, Brazil. Subsequently, Vanzolini (1986a) listed A. flammigerus snethlageae in his 88 ‘Addenda and Corrigenda’ in the Catalogue of Neotropical Squamata. However, 89 Vanzolini (1986b) considered A. snethlageae as a full species without further Journal of Zoological Systematics and Evolutionary Research Page 4 of 81

90 comments. Dixon & Soini (1986) reported on two specimens of Atractus flammigerus 91 from the Iquitos region, and Duellman & Salas (1991) reported A. flammigerus from 92 Cuzco Amazónico (northern and southern Peru, respectively). Atractus snethlageae has 93 been reported from Rondônia State, Brazil (Ávila-Pires, Vitt, Sartorius, & Zani, 2009; 94 Bernarde & Abe, 2006; Bernarde, Albuquerque, Barros, & Turci, 2012; Jorge-da-Silva, 95 1993; Marçal, Gomes, & Coragem, 2011; Nascimento, Ávila-Pires, & Cunha, 1988). 96 Giraudo & Scrocchi (2000) re-identified as A. snethlageae a specimen previously 97 identified as A. badius (Serié, 1915) and discussed its presence in Argentina. Silva-Haad 98 (2004) provided the first record of A. snethlageae from Colombia, and Gonzales & 99 Embert (2008) reported it from Manupiri, Pando, Bolivia. The name Atractus badius 100 has been often used inFor reference Review to A. snethlageae Only throughout Amazonia (but see Passos 101 et al., in press). Similarly, A. snethlageae has been repeatedly mistaken with A. major 102 (Savage, 1960; Pérez-Santos & Moreno, 1988) or A. flammigerus (Dixon & Soini, 103 1986; Duellman, 2005). The difficulty in the correct assignment of the name A. 104 snethlageae reflects the remarkable polychromatism of this species throughout its 105 widespread distribution and the possible existence of a species complex (Schargel et al., 106 2013). Passos, Ramos, Fouquet, & Prudente (2017) restricted A. flammigerus to the 107 Guiana Shield. More recently, Melo-Sampaio, Passos, Fouquet, Prudente, & Torres- 108 Carvajal (2019) revised the taxonomic status of populations previously associated with 109 A. schach and A. snethlageae, restricting the distribution of the former to the Guiana 110 Shield and recognizing three new species, of which two—Atractus dapsilis Melo- 111 Sampaio, Passos, Fouquet, Prudente, & Torres-Carvajal and A. trefauti Melo-Sampaio, 112 Passos, Fouquet, Prudente, & Torres-Carvajal—resemble A. snethlageae in general 113 colouration. 114 Herein, we combine morphological analyses and molecular phylogenetics to test 115 species boundaries within the Atractus snethlageae complex throughout its distribution. 116 We analyse the largest dataset to date, including 330 specimens and 30 new tissue 117 samples of Atractus snethlageae (sensu lato) from Bolivia, Brazil, Colombia, Ecuador 118 and Peru, comprising most of its range of distribution as reported in the literature. 119 120 2 MATERIAL AND METHODS

121 2.1 MOLECULAR SAMPLING, TECHNIQUES, AND SELECTION OF SEQUENCES 122 We gathered liver tissue samples from 30 specimens representing six species within 123 Atractus snethlageae complex. These samples were obtained through loans or donations Page 5 of 81 Journal of Zoological Systematics and Evolutionary Research

124 from the following institutions: Universidade Federal do Acre (UFAC-RB), Rio Branco, 125 Brazil; Museu Nacional, Universidade Federal do Rio de Janeiro (MNRJ), Rio de 126 Janeiro, Brazil; Museu Paraense Emílio Goeldi (MPEG), Belém, Brazil; Universidade 127 Federal de Mato Grosso (UFMT-R), Cuiabá, Brazil; Laboratory of Herpetology (MTR) 128 from the Universidade de São Paulo (USP), São Paulo, Brazil; Museo de Zoología de la 129 Pontificia Universidad Católica del Ecuador (QCAZ), Quito, Ecuador; Museo de 130 Zoología de la Universidad Tecnológica Indoamérica (MZUTI), Quito, Ecuador; Centro 131 de Ornitología y Biodiversidad (CORBIDI), Lima, Peru; Museo de la Universidad de 132 San Marcos (MUSM), Lima Peru; and Centre de Recherche de Montabo (CNRS), 133 Cayenne, French Guiana, France. 134 We obtained nucleotideFor Review sequences from three Only mitochondrial genes, ribosomal 135 large subunit gene (16S, 53 samples), subunit IV of NADH dehydrogenase (ND4, 36 136 samples), cytochrome b (CYT-B, 42 samples), as well as three nuclear genes, oocyte 137 maturation factor mos (C-MOS, 42 samples), neurotrophin-3 (NT3, 46 samples), and 138 recombination-activating gene 1 (RAG1, 45 samples). We extracted genomic DNA 139 from tissue samples stored in absolute or 95% ethanol using a guanidinium 140 isothiocyanate extraction protocol (see details in Torres-Carvajal & Hinojosa, 2020). 141 We performed Polymerase Chain Reaction (PCR) amplification of gene fragments in a

142 final volume of 25 μl reactions using 1 X PCR Buffer (–Mg), 3 mM MgCl2, 0.2 mM 143 dNTP mix, 0.2 μM of each primer, 0.1 U/μl of Taq DNA Polymerase and 1 μl of 144 extracted DNA. Primers and protocols are listed in Supplementary Table 1. All 145 amplified markers were obtained from the same voucher specimen to avoid problems 146 inherent with chimeric sequences. 147 We analysed polymerase chain reaction products on 1% agarose gels by 148 horizontal electrophoresis (the target fragment size was estimated from molecular 149 weight markers) using SYBR1 Safe (Invitrogen, Carlsbad, CA) staining, with a 150 Molecular Imager1 Gel DocTM XR+ Imaging System (Bio Rad, Hercules, CA). We 151 treated amplified products with ExoSAP-IT (Affymetrix, Cleveland, OH) to remove 152 remaining dNTPs and primers, as well as extraneous single-stranded DNA produced in 153 the PCR. We performed double stranded sequencing of the PCR products in both 154 directions with Macrogen Inc. New sequences are presented in Supplementary Table 2 155 and museum vouchers are housed at CORBIDI, MNRJ, MZUSP, QCAZ and UFMT-R. 156 In addition, we obtained GenBank sequences for 33 outgroup species 157 (Supplementary Table 2), limiting our sampling to species for which we verified Journal of Zoological Systematics and Evolutionary Research Page 6 of 81

158 directly the identifications of the vouchers. Specimen MZUTI 5409 (ANF 2390) listed 159 as Atractus touzeti (Arteaga et al., 2017) was re-identified. We also excluded sequences 160 of distantly related species [Atractus albuquerquei Cunha & Nascimento, 1983, A. 161 reticulatus (Boulenger, 1885), A. trihedrurus Amaral, 1926 and Atractus zebrinus (Jan, 162 1862)], without vouchers or locality data and separated the chimeras (terminals 163 composed of sequences from different individuals or different species; Supplementary 164 Table 3) presented by Grazziotin et al. (2012), Pyron, Burbrink, & Wiens (2013), Pyron, 165 Guayasamin, Peñafiel, Bustamante, & Arteaga (2015), Figueroa et al. (2016), Pyron, 166 Arteaga, Echevarría, & Torres-Carvajal (2016), and Arteaga et al. (2017). 167 168 2.2 PHYLOGENETIC ANALYSESFor Review Only 169 We assembled and aligned the data in Mega 7.0 (Kumar, Stecher, & Tamura, 2016) 170 under default settings for the alignment program Clustal W (Thompson, Higgins, & 171 Gibson, 1994) and built a concatenated matrix using SequenceMatrix (Vaidya, Lohman, 172 & Meier, 2011). We then selected the best-fit nucleotide substitution models and 173 partitioning scheme simultaneously using PartitionFinder 2 (Lanfear, Frandsen, Wright, 174 Senfeld, & Calcott, 2016) under the Bayesian Information Criterion (Sullivan & Joyce, 175 2005), after partitioning the matrix by gene (16S, CYTB, CMOS, ND4, NT3, RAG1) 176 (Supplementary Table 4). We analysed the concatenated, partitioned matrix under a 177 Bayesian inference method using MrBayes v3.2.1 (Ronquist et al., 2012). We chose this 178 strategy because Montingelli et al. (2019) demonstrated the advantages of multiple 179 information, where a concatenated matrix has an important role summarizing few 180 phylogenetically informative sites from the nuclear loci and filling the gap between 181 some non-overlapping taxon sampling; and even if contradictory, the phylogenetic 182 signal of the concatenated nuclear loci will reduce the influence of the mitochondrial 183 data. All parameters except topology and branch lengths were unlinked between 184 partitions. Four independent runs, each with four MCMC chains, were set for 10 million 185 generations each, sampling every 10,000 generations. We used Tracer v1.6 (available 186 from http://beast.bio.ed.ac.uk/Tracer) to assess convergence, stationarity, and effective 187 sample sizes (ESS>200) of model parameters. We used LogCombiner 1.8.4 188 (Drummond, Suchard, Xie, & Rambaut, 2012) to summarize the runs. Of 4,000 189 resulting trees, 10% were discarded as “burn-in” from each run. We used the resultant 190 trees to calculate posterior probabilities (PP) in a maximum clade credibility tree in 191 TreeAnnotator v1.8.3 (Drummond et al., 2012). We edited and visualized the Page 7 of 81 Journal of Zoological Systematics and Evolutionary Research

192 phylogenetic trees using FigTree v1.4.4 (Available in 193 http://tree.bio.ed.ac.uk/software/figtree/). We also performed Maximum likelihood 194 (ML) analyses on the partitioned dataset using RAXML (Stamatakis, 2014) under the 195 GTRCAT approximation (Stamatakis, Hoover, & Rougemont, 2008). Support of nodes 196 was assessed using the rapid-bootstrapping algorithm with 1000 non-parametric 197 bootstraps. We considered the clades with posterior probability values >0.95 and ML 198 bootstraps >70 as strongly supported. 199

200 2.3 MORPHOLOGY, SPECIES BOUNDARIES AND PRESENTATION RATIONALE 201 Terminology for cephalic shields follows Savage (1960) and Peters (1964), 202 whereas ventral and subcaudalFor Reviewcounts follow Dowling Only (1951). Condition of the loreal 203 scale follows Passos, Fernandes, & Borges-Nojosa (2007b). Measurements were taken 204 with a Mitutoyo® digital caliper to the nearest 0.1 mm, except for snout-vent length 205 (SVL) and tail length (TL), which were measured with a ruler to the nearest 1 mm. 206 Measurements and descriptions of paired cephalic scales are strictly based on the left 207 side of the head. We follow the definitions of Passos, Prudente, & Lynch (2016) for 208 body marks (blotches, spots, and dots), which were counted separately on each side of 209 the dorsum, and used the term ‘‘blotch’’ to refer to broad (two or more scales long and 210 wide) dorsal marks located on the vertebral and paravertebral regions. Colours were 211 described following the standard catalogue of Köhler (2012). We analysed 330 212 specimens of banded Atractus. The sex of specimens was determined by verifying 213 presence-absence of hemipenes through a ventral incision at the base of the tail, except 214 when hemipenes had been previously everted. We examined maxillae in situ under a 215 Luxeo 4Z (Labomed) stereoscope through a narrow lateromedial incision between the 216 supralabials and the maxillary arch. After removing tissues covering the maxillary bone, 217 we counted teeth and empty sockets. The method for preparation of preserved 218 hemipenes was modified from Pesantes (1994) by replacing potassium hydroxide 219 (KOH) with distilled water, and filling the hemipenes with petroleum jelly (Passos et 220 al., 2016). Terminology for hemipenial descriptions follows Dowling & Savage (1960) 221 and Zaher (1999), with a few minor modifications based on Passos et al. (2016). 222 Data from additional specimens of Atractus previously examined are listed in 223 Meneses-Pelayo & Passos (2019) and such studies included examination of nearly all 224 types available for the genus Atractus. Nonetheless, we list in the 'Species Account' the 225 type specimens of the Atractus snethlageae complex, in addition to sympatric Journal of Zoological Systematics and Evolutionary Research Page 8 of 81

226 morphologically similar congeners listed in the Appendix. In addition to phylogenetic 227 relationships, we consider unique diagnostic characters distinguishing a putative taxon 228 from the others in the A. snethlageae complex as species delimitation criteria (de 229 Queiroz, 2007). Notwithstanding, we search for concordance between the discrete and 230 continuous characters. Since some features such as colour patterns, morphometrics, and 231 hemipenial morphology are likely uncorrelated with each other; the correspondence 232 between these data sources might represent independent evidence for robust species 233 boundaries (Passos, Prudente, Ramos, Caicedo-Portilla, & Lynch, 2018). 234 The rationale along the species account include detailed morphological 235 comparisons following at least one of these four types of hierarchical levels of similarity 236 or historical association:For (i) species Review with close phylogenetic Only relationships, recovered 237 along the present study; (ii) sympatric or parapatric taxa presenting similar colour 238 pattern, scutelation and morphometric features; (iii) phenotypically similar taxa even 239 known to occur parapatrically or allopatrically regarding the A. snethlageae species 240 complex (e.g., A. schach and closely related species; see Melo-Sampaio et al., 2019); 241 (iv) species previously confused with the former species in literature or scientific 242 collections (e.g., A. flammigerus; see Passos, Ramos, Fouquet, & Prudente, 2017). We 243 follow Passos, Fernandes, Bérnils, & Moura-Leite (2010c) regarding conditions of 244 morphological characters used in diagnoses and descriptions. All species recognized 245 herein as monophyletic exhibited unambiguous phenotypic diagnostic characters or 246 exclusive combinations of traits. 247 248 3 RESULTS

249 3.1 Phylogenetic Analyses 250 Maximum Likelihood (ML) and Bayesian Inference (BI) trees are in agreement 251 (Figure 1; Figures S1-S2). We recovered a paraphyletic A. snethlageae (sensu lato) in 252 four distinct clades. A clade from Ecuador was recovered with maximum support in 253 both trees as sister to a larger clade, containing Guianan clade (Atractus dapsilis (A. 254 schach, A. trefauti)) as sister to clade containing the other three A. snethlageae (sensu 255 lato) clades. A sample from southeastern Peru comprises a unique clade. A clade 256 composed by eastern Amazonia specimens represented by Tocantins and Xingu river is 257 nested with specimens from Madeira, Purus and Aripuanã river in south Amazonia and 258 Caxiuanã bay in eastern Amazonia. This clade is recognized as Atractus snethlageae 259 sensu stricto. The last clade is composed of Peruvian and Ecuadorian specimens from Page 9 of 81 Journal of Zoological Systematics and Evolutionary Research

260 populations adjacent to Amazonian slopes of Andes and the extreme western lowland 261 Amazonian specimens. To better reflect the relationships among sampled populations 262 and maintain species monophyly, we describe below three of the four clades of A. 263 snethlageae (sensu lato) as separate species, each supported by putative autopomorphies 264 or unique combinations of morphological characters.

265

266 3.2 Phenotypic variation among different populations of Atractus snethlageae 267 We discovered two additional unnamed species upon examination of hemipenial 268 morphology, maxillary dentition, morphometrics, colouration and scale counts. In the 269 absence of molecular data for these species, we compared them with geographically 270 close and morphologicallyFor similar Review congeners according Only to our own delimitation criteria 271 (see above). 272 Morphological comparisons with respect to the cis-Andean Atractus complexes, 273 species groups or previously recovered clades south of the Amazon River: A. schach 274 complex (including A. snethlageae and other allied species) differs from the A. elaps 275 species group [except for A. latifrons (Günther, 1868)], A. emmeli (Boettger, 1888), A. 276 occiptoalbus (Jan, 1862), A. roulei Depax, 1910 and A. trilineatus Wagler, 1828 groups 277 or complexes of species (in parentheses) in having 17 dorsal scale rows (vs. 15 dorsal 278 scale rows); from the Atractus collaris Peracca, 1897 group in lacking apical pits and 279 supracloacal tubercles (vs. presence of apical pits in male and females, and supracloacal 280 tubercles in sexually mature males); from the A. elaps (Günther, 1858) species group 281 [including A. latifrons (Günther, 1868)] in having loreal scales twice as long as high and 282 rostral scale wider than high (vs. loreal as long as high and rostral higher than long); 283 from A. maculatus (Günther, 1858) and A. pantostictus Fernandes & Puorto, 1993 284 species groups in having hemipenes usually with lobular crests and belly scattered with 285 dark brown dots or spots eventually forming continuous stripes (vs. hemipenes lacking 286 lobular crests, with alary spines and lateral projections [A. pantostictus group], and belly 287 usually immaculate but, when darkened uniformly pigmented [in A. francoi Passos, 288 Fernandes, Bérnils, & Moura-Leite, 2010, A. serranus Amaral, 1930 and A. trihedrurus 289 Amaral, 1926], never light with dispersed dark marks); from A. bocki Werner, 1909, A. 290 edioi Silva, Silva, Ribeiro, Souza, & Souza, 2005, A. natans Hoogmoed & Prudente, 291 2003, A. thalesdelemai Passos, Fernandes, & Zanella, 2005, and A. torquatus Duméril, 292 Bibron, & Duméril, 1854 in having 17 dorsal scale rows, seven suparalabial, seven Journal of Zoological Systematics and Evolutionary Research Page 10 of 81

293 infralabials, two postoculars, belly ground colour light covered with regular or irregular 294 dark marks, semicapitate and semicalyculate hemipenis with capitular region similar in 295 length or longer than the hemipenial body and usually lobular symmetry (vs. 15 dorsals 296 and belly mostly cream in A. edioi; six supralabials in A. bocki; lobular asymmetry and 297 capitulum shorter than hemipenial body in A. natans; six supralabials, six infralabials, 298 single postocular and belly uniformly cream in A. thalesdelemai; usually eight 299 supralabials, eight infralabials, single postocular, and non-capitate non-calyculate organ 300 in A. torquatus). 301

302 3.3 Species account 303 Order Squamata Oppel,For 1811 Review Only 304 Family Dipsadidae, Bonaparte, 1838 305 Subfamily Dipsadinae, Bonaparte, 1838 306 Tribe Dipsadini, Bonaparte, 1838 307 Genus Atractus Wagler, 1828 308 309 Atractus snethlageae Cunha & Nascimento, 1983. 310 Atractus badius – specimen “E” Boulenger (1894). 311 Atractus badius – Cunha & Nascimento (1978). 312 Atractus flammigerus snethlageae – Cunha & Nascimento (1983), partim 313 (MPEG 16382 represents Atractus trefauti); Cunha & Nascimento (1993). 314 Atractus snethlageae – Vanzolini (1986b); Jorge-da-Silva (1993); Doan & 315 Arizabal (2000); Frota (2000); Frota (2004); Giraudo & Scrocchi (2000); Giraudo 316 (2004); Bernarde & Abe (2006); Gutsche et al. (2007); Gonzales & Embert (2008); 317 Passos & Fernandes (2008), partim; Prudente & Passos (2008), partim; Ávila-Pires et 318 al. (2009); Marçal et al. (2011); Bernarde et al. (2012); Schargel et al. (2013), partim; 319 Vaz-Silva et al. (2015), partim. 320 Atractus flammigerus – Dixon & Soini (1986); Duellman & Salas (1991); Jorge- 321 da-Silva (1993); Duellman (2005); Doan & Arizabal (2002); Carrillo de Espinoza & 322 Icochea (1995). 323 Atractus schach – Nascimento et al. (1988); Passos & Fernandes (2008), partim; 324 Prudente & Santos-Costa (2005); Prudente & Passos (2008), partim; Bernarde, 325 Machado, & Turci (2011); Bernarde et al. (2012); Waldez et al. (2013); Vaz-Silva et al. Page 11 of 81 Journal of Zoological Systematics and Evolutionary Research

326 (2015), partim; Rodrigues, Maschio, & Prudente (2016); Bernarde, Turci, & Machado 327 (2017). 328 Atractus sp. – Silva, Souza, & Bernarde (2012); Melo-Sampaio et al. (2019). 329 Holotype: MPEG 10131, adult male from Brazil, Pará, Colônia Nova, 10 km 330 near Gurupi river, on the BR-316 highway. The type locality is herein restricted to 331 municipality of Cachoeira do Piriá (1.821ºS, 46.402ºW), 29 m above sea level (hereafter 332 asl), collected by O.R. da Cunha and F.P. do Nascimento on 3 October 1976 (Figure 2; 333 Figure S3). 334 Paratypes (n=11): All collected by O.R. da Cunha and F.P. do Nascimento in 335 Brazil. Six females: Pará, Ananindeua, Lago Azul (1.383ºS, 48.406ºW), 33 m asl, 336 MPEG 16383, 16384 For(the last incorrectlyReview numbered Only as MPEG 16387); Capanema, Bela 337 Vista (now Tauari Village), (1.131ºS, 47.065ºW) 37 m asl; MPEG 2543 (stated as male 338 in original description); Belém, Mosqueiro Island (1.151ºS, 48.378ºW) 34 m asl, MPEG 339 2595; Maranhão, Junco do Maranhão, Nova Vida (1.825ºS, 46.107ºW) 31 m asl, MPEG 340 14986, 15422. Five males: Pará, Ananindeua, Lago Azul (1.383ºS, 48.406ºW) 33 m asl, 341 MPEG 16385; Belém, São João da Pratinha (1.368ºS, 48.476ºW), 20 m asl, MPEG 342 10137; Capanema, Bela Vista (now Tauari Village), (1.131ºS, 47.065ºW), 37 m asl, 343 MPEG 6845, 15973; Santa Bárbara do Pará (formerly Benevides) (1.228ºS, 48.292ºW), 344 16 m asl, (MPEG 3955) (Figure 3). 345 Diagnosis: Atractus snethlageae can be distinguished from all congeners by the 346 unique combination of the following characters: (1) smooth dorsal scale rows 17/17/17; 347 (2) postoculars two; (3) loreal moderately long; (4) temporal formula 1+2; (5) 348 supralabials usually seven, third and fourth contacting eye; (6) infralabials usually eight, 349 first four contacting chinshields; (7) maxillary teeth six; (8) gular scale rows three; (9) 350 preventrals one to four (usually three); (10) ventrals 147–163 in females, 137–155 in 351 males; (11) subcaudals 21–24 in females, 27–34 in males; (12) in preservative, dorsum 352 pale brown with irregular pale yellowish transversal bands; (13) in preservative, venter 353 usually cream with brown midventral line; (14) body moderately long in females 354 (maximum 400 mm SVL) and males (maximum 310 mm SVL); (15) tail short in 355 females (8.2–11.3% of SVL) and moderately long in males (13.3–16% of SVL); and 356 (16) hemipenes slightly to moderately bilobed (≤ half the length of capitulum), 357 semicapitate and semicalyculate. 358 Comparisons: We compared Atractus snethlageae with the remaining allopatric 359 species of the clade: A. dapsilis, A. schach, and A. trefauti. Atractus snethlageae differs Journal of Zoological Systematics and Evolutionary Research Page 12 of 81

360 from A. dapsilis by having ≤ 163 ventrals in females, ventral mean=145.7 in males, 361 fully everted hemipenes slightly to moderately bilobed, and retracted organs extending 362 to the level of eighth subcaudal (vs. ≥ 167 ventrals in females, ventral mean=159.5 in 363 males, fully everted hemipenes strongly bilobed, and retracted organs extending to the 364 level of 10–12th subcaudal); from A. schach by having usually six maxillary teeth and 365 hemipenes semicapitate (vs. seven maxillary teeth and hemipenes without capitular 366 groove); from A. trefauti by having usually six maxillary teeth, and tail ≥ 13.3% of SVL 367 in males and dorsal ground colour varying from Vandyke brown to Amber in the “light” 368 morph (vs. often seven maxillary teeth, tail ≤ 13.2% of SVL in males and black 369 dorsum). Among Amazonian sympatric congeners, A. snethlageae shares only with A. 370 badius, A. boimirim Passos,For Prudente, Review & Lynch, 2016Only, A. flammigerus, A. major 371 Boulenger, 1894, A. natans Hoogmoed & Prudente, 2003, Atractus tartarus Passos, 372 Prudente, & Lynch, 2016 and A. torquatus usually a blotched or banded colour pattern. 373 Atractus snethlageae differs from A. boimirim and A. tartarus by having 17/17/17 374 dorsal scale rows (vs. 15/15/15 dorsal scale rows); from A. badius by having dorsum 375 light covered with dark marks or dark with light marks (vs. dorsum red with 376 conspicuous black dyads separated by white band; see Passos et al. in press); from A. 377 flammigerus by having usually seven supralabials and lacking keels on the dorsal scales 378 (vs. eight supralabials and keels covering dorsal scale rows); from A. major by having 379 usually four infralabials contacting chinshields and seven maxillary teeth (vs. three 380 infralabials contacting chinshields and six maxillary teeth); from A. natans by having 381 ventral surface of body mostly creamish white covered by few dark brown blotches 382 usually linearly arranged on the midventer (vs. belly almost entirely black except for the 383 lateral sides of ventral scales creamish white); from A. torquatus by having two 384 postoculars and seven supralabials (vs. usually one postocular and eight supralabials; 385 see Passos & Prudente, 2012). We refer to Table 1 for additional comparisons between 386 Atractus snethlageae and other Amazonian congeners. 387 Description of the holotype: We refer to Cunha & Nascimento (1983) and 388 describe additional features. SVL 182 mm, TL 30 mm; midbody diameter 5.3 mm. 389 Rounded head in lateral view; symphysial three times wider than high; gular scale rows 390 three; preventrals four; loreal moderate touching second and third supralabials; nasal 391 divided; internasals reduced, slightly wider than long; rostral not visible from above, 392 about twice as wide as high; prefrontals longer than wide, fused posteriorly; supraocular 393 small, squared, shorter than eye diameter; frontal triangular as long as wide; parietals Page 13 of 81 Journal of Zoological Systematics and Evolutionary Research

394 twice as long as wide; postoculars 2/2; temporals 1+2, posterior ones larger; pupil 395 rounded; supralabials seven, with first shorter and narrower than second; second and 396 third supralabials in contact with loreal; infralabials eight, first four (right) and first five 397 (left) touching chinshields; ventrals 137; subcaudals 29; maxillary teeth six. 398 Colour in preservative: Dorsum of head olive-brown [colour 278] including 399 posterior portion of parietals; nuchal band Straw Yellow [colour 53] (two and a half 400 scales long), reaching occipital region, except for the first dorsal scale row; ventral 401 portion of supralabials cream [colour 12]; infralabials and gular region cream [colour 402 12]; symphysial, first two pair of infralabials and first third of chinshields stained with 403 sepia dots; body dorsum Vandyke brown [colour 38] with third row beige [colour 254], 404 light yellow ocher [colourFor 13] bandsReview one-half to oneOnly scale long connected or not on the 405 vertebral axis; venter cream with an irregular midventral sepia stripe [colour 279], 406 increasing in width gradually on posterior portion of body; tail ventrally Raw siena 407 [colour 32] (Figure 2). 408 Colour in life: The extreme dark morph is represented by MZUSP 21474, with 409 sepia dorsum of head covering frontal and postoculars, and a pale buff [colour 1] band 410 covering temporals, parietals, one row of occipitals and two last supralabials; sepia 411 body dorsum with buff transverse bands one-half to one scale long interrupted or not on 412 the vertebral axis; venter light buff with Burnt umber spots forming midventral line; 413 venter of specimen UFRO-H 2549 immaculate chamois [colour 83] on the first quarter, 414 with tiny sepia dots increasing in number posterior to midbody, forming a diffuse 415 midventral line at the level of six scales before cloaca (Figure 3). 416 Colour variation in preservative (n=128): Dorsum of head with olive-brown 417 [colour 278] cephalic-cap, extending from rostral to medial region of parietal; posterior 418 portion of parietal and temporal region ground cinnamon [colour 270], constituting an 419 occipital band of variable size, but variable in width and length; occipital band ground 420 cinnamon with medial constriction or trapezoidal-shaped, being posteriorly limited by 421 background colour of body; occipital band eventually wider, extending dorsally over 422 posterior region of parietal until first dorsal with cinnamon pigmented; occipital band 423 occasionally indistinct, making head entirely sepia dorsal and laterally; head laterally 424 sepia [colour 279] to anterior region of sixth supralabial and predominantly cream after 425 postorbital region; temporal and occipital regions generally creamy, with irregular 426 brown spots invading anterior portion of temporals; temporal and occipital regions 427 become uniformly cream with enlargement of occipital band; supralabial usually brown Journal of Zoological Systematics and Evolutionary Research Page 14 of 81

428 to the level of sixth scale and predominantly cream on the remained supralabials; 429 supralabial occasionally entirely sepia; gular region cream with black spots on the 430 symphysial, first pair of infralabials, mesial region of infralabials and anterior portion of 431 chinshields; preventrals usually uniformly cream; gular region eventually densely 432 stained by irregular brown dots; belly with background colour varying from uniform 433 ground cinnamon to predominantly grayish horn [colour 268], usually with round hair 434 brown spots located on the middle of each ventral scales forming conspicuous (but 435 frequently discontinuous) midlines throughout body; spots occasionally laterally 436 expanded, constituting enlarged ventral bands or concentrating on anterior part of 437 ventral scales; belly occasionally thickly stained with brown scoring, more concentrated 438 on the middle and posteriorFor third Review of body; tail usually Only hair brown uniform or irregularly 439 invaded by ground cinnamon spots; body dorsally sepia with 24–34 transverse bands; 440 bands brownish olive or yellowish cream (one to two scales in length) alternated on the 441 flanks and, occasionally connected to opposite one on the vertebral region forming 442 narrow rings separated by sepia interspaces (four to seven scales long); bands generally 443 extending into paraventral region; possibly with an irregular pattern, with diffuse and 444 longitudinally oriented spots (Figure 4). For the specimens with “light” morphotype the 445 colours varying from dorsum Amber (MZUSP 21464) to Cinnamon-Drab with fuscous 446 bands (UFRO-H 2548), connected on the vertebral axis, two or three scales long. 447 Parietal band is often cream white (also cinnamon-drab to cinnamon brown). 448 Hemipenial morphology (n=6): Organs in situ (retracted) extend to the level of 449 eighth subcaudal and bifurcate at the level of seventh subcaudal (n=2). Fully everted 450 and maximally expanded hemipenes renders a slightly (more frequently, ≤ than half 451 capitulum length) to moderately (more rarely, ≥ than half capitulum length) bilobed, 452 semicapitate and semicalyculate organ; lobular region wider than hemipenial body; 453 lobes centrolinearly oriented and flattened, clavated or attenuated (rarely) on the tip; 454 lobes with lateromesial expansion toward intrasulcar region (Figure 5B–F), sometimes 455 only evident nearly to lobe bifurcation (Figure 5E); lobes symmetrical and presenting 456 proximal portion from intrasulcar region with straight base; straight base more 457 conspicuous in conical or attenuate organs (Figure 5); lobes uniformly covered with 458 spinulate calyces on both sides of hemipenes; spinules usually replaced by irregular 459 papillae toward apices of lobes; capitular groove usually distinct on both faces of organ; 460 capitular groove generally indistinct in slightly bilobed hemipenes (Figure 5) and more 461 conspicuous in moderately bilobed organs (Figure 5); capitulum with transversal Page 15 of 81 Journal of Zoological Systematics and Evolutionary Research

462 spinulated flounces formed by union of horizontal walls of calyces; calyces lacking 463 vertical walls along the sulcate and asulcate faces of capitulum; conspicuous transversal 464 calyculated flounces with irregular rows on the intrasulcar region; hemipenial body 465 elliptical covered with enlarged hooked spines; larger spines generally located laterally 466 below sulcus spermaticus bifurcation; distal region of hemipenial body on maximally 467 expanded organ with rows of spines concentrating in the middle of asulcate face; sulcus 468 spermaticus bifurcates approximately on the 50% of organ length; sulcus spermaticus 469 margins relatively thick at level of division and along the capitular region; sulcus 470 spermaticus not bordered by spinules; basal naked pocket restricted to most basal region 471 of hemipenial body; proximal region of hemipenes body covered with few hooked 472 spines and dispersed spinulesFor (FigureReview 5). Only 473 Quantitative variation (n=128): Largest female 420 mm SVL, 40 mm TL; largest 474 male 335 mm SVL, 55 mm TL; ventrals 147–163 (mean=155.1; n=61; SD=5.3) in 475 females, 137–158 (mean=145.7; n=54; SD=7.9) in males; subcaudals 17–26 476 (mean=22.9; n=61; SD=0.9) in females, 26–39 (mean=31.4; n=54; SD=2.5) in males; 477 supralabials seven (n=27 sides) or eight (n=7 sides); infralabials seven (n=2 sides), eight 478 (n=31 sides) or nine (n=1 side); preventrals one (n=4), two (n=10), three (n=88) or four 479 (n=11); maxillary teeth five (n=2 sides), six (n=52 sides), seven (n=144 sides) or eight 480 (n=10 sides). 481 Distribution: Atractus snethlageae occurs in southern Amazonia (from 482 Maranhão state in Brazil to Peru), with a few records in northern Amazonia (on the 483 north side of the Amazon River) only in the western portion of Amazonia (Amazonas 484 state in Brazil, Colombia, Peru, and Ecuador) (Figure 6). 485 Remarks: Cunha & Nascimento (1978) first identified A. snethlageae as 486 Atractus badius. However, the corresponding specimen was not included within the 487 type series of A. flammigerus snethlageae (Cunha & Nascimento, 1983). When 488 examining the type series, we found that there is a typographical error in relation to 489 female paratype MPEG 16387, which is MPEG 16384. Moreover, we were not able to 490 find the “tubercles" mentioned by Cunha & Nascimento (1983), since the dorsal scales 491 are entirely smooth. Hoogmoed & Prudente (2003) have hipotesized that anomalous 492 conditions of internasal scales in prefrontal suture should be evaluated regarding 493 relationship among Atractus. Atractus snethlageae was described based on a series of 494 13 specimens from the states of Amapá, Maranhão and Pará, Brazil. However, the 495 specimen MPEG 16382 from Serra do Navio was recently recognized as A. trefauti, Journal of Zoological Systematics and Evolutionary Research Page 16 of 81

496 making the original type-series of A. snethlageae composite (Melo-Sampaio et al., 497 2019). The central Amazonian records of A. snethlageae in the Manaus region by Fraga, 498 Lima, Prudente, & Magnusson (2013) belong to a dark morph of A. dapsilis. The 499 Argentinian record of A. snethlageae (Giraudo & Scrocchi, 2000) is tentatively 500 maintained as A. cf. snethlageae due to the presence of a conspicuous light parietal band 501 (see Giraudo, 2004). On the other hand, the Ecuadorian populations assigned to A. 502 snethlageae by Schargel et al. (2013) are recognized herein as two separate species (see 503 below). 504 505 Atractus nawa sp. nov. 506 urn:lsid:zoobank.org:act:E96A9F8B-8D29-4A9F-A13B-650DAE22DCACFor Review Only 507 Atractus major – Ávila-Pires et al. (2009). 508 Atractus schach – Passos & Fernandes (2008), partim; Prudente & Passos 509 (2008), partim. 510 Holotype: MPEG 20376 (field number LJV 6299), adult female from Brazil, 511 Acre, Porto Walter (8.258ºS, 72.776ºW), 212 m asl, collected by L.J. Vitt, T.C.S. Avila- 512 Pires, J.P. Caldwell and V. Oliveira on 28 February 1996 (Figure 7). 513 Paratype: UFACF 3771, adult female from Brazil, Acre, Cruzeiro do Sul, km 80 514 of the BR-364 highway on the route to Tarauacá (7.750ºS, 72.366ºW), 200 m asl, 515 collected by R.A. Machado on February 2010. 516 Etymology: The specific epithet ‘nawa’ corresponds to the self-designation or to 517 an indicator of otherness (other people) of many Pano-speaking societies living along 518 Juruá River basin (Montagner, 2007). The word also refers to the distinction of the new 519 species from its congeners by indigenous people on the region. 520 Diagnosis: Atractus nawa can be distinguished from all congeners by unique 521 combination of the following characters: (1) smooth dorsal scale rows 17/17/17; (2) 522 postoculars two; (3) loreal moderately long; (4) temporal formula 1+2; (5) supralabials 523 seven, third and fourth contacting eye; (6) infralabials seven, first four contacting 524 chinshields; (7) maxillary teeth seven; (8) gular scale rows four; (9) preventrals four; 525 (10) ventrals 166–169 in females, unknown in males; (11) subcaudals 16–20 in females, 526 unknown in males; (12) in preservative, dorsum Brussels brown [colour 33] with Raw 527 umber spots; (13) in preservative, venter predominantly pale cinnamon with warm sepia 528 dots near cloaca and midventral portion of tail; (14) body size moderately long in Page 17 of 81 Journal of Zoological Systematics and Evolutionary Research

529 females (maximum 405 mm SVL); and (15) tail short in females (6.2–7.7% of SVL); 530 (16) hemipenis unknown. 531 Comparisons: We compared the new species with geographically closer and 532 sympatric congeners occurring along the state of Acre, Brazil. Atractus nawa differs 533 from A. albuquerquei, A. boimirim, A. emmeli and A. elaps by having 17/17/17 dorsal 534 scale rows (vs. 15/15/15). Regarding the species with 17 dorsal scales rows, A. nawa 535 differs from A. major in having first four infralabials in contact with chinshields, dorsal 536 Brussels brown with Raw umber spots, pale cinnamon with warm sepia dots near 537 cloaca, and ≤ 20 subcaudals in females (vs. first three infralabials in contact with 538 chinshields, dorsum gray or reddish, followed by square sepia dots from midbody to 539 posterior region of bellyFor and > Review27 subcaudals in females); Only from A. snethlageae in 540 having tail < 8% of SVL in females, ≤ 20 subcaudals in females, ≥ 166 ventrals in 541 females, and pale orange snout (vs. tail > 8% of SVL in females, > 26 subcaudals in 542 females, ≤ 163 ventrals, and olive brown snout); from A. latifrons in having loreal 543 scales twice as long as high and rostral scale wider than high (vs. loreal as long as high 544 and rostral higher than long). We refer to Table 1 for additional comparisons between 545 Atractus nawa and other Amazonian congeners. 546 Description of the holotype: SVL 390 mm, tail length 30 mm (7.7% of SVL); 547 head slightly distinct from body; head length 11.1 mm (3% of SVL); head width 7.0 548 mm (63.1% head length); rostral–orbit distance 3.9 mm; nostril–orbit distance 2.8 mm; 549 interorbital distance 4.1 mm; head rounded in lateral view; snout truncate in dorsal 550 view, truncate in lateral view; canthus rostralis conspicuous; rostral subtriangular in 551 frontal view, 2.0 mm wide, 1.2 mm high, well visible in dorsal view; internasal 0.9 mm 552 long, 1.1 mm wide; internasal suture sinistral with respect to prefrontal suture; 553 prefrontal 2.7 mm long, 2.6 mm wide; supraocular subtrapezoidal, 1.1 mm long, 1.1 554 mm wide at broadest point; frontal bell-shaped, 2.5 mm long, 3.7 mm wide; parietal 5.0 555 mm long, 3.0 mm wide; nasal entirely divided, nostril divided in both parts; prenasal 1.0 556 mm high, 0.5 mm long; postnasal 1.0 mm high, 0.9 mm long; loreal 2.2 mm long, 1.0 557 mm high; second and third supralabials contacting loreal; third and fourth supralabials 558 entering the orbit; eye diameter 1.2 mm; pupil rounded; two postoculars similar in 559 height, being lower shorter than upper; upper postocular 0.7 mm long, 0.8 mm high; 560 lower postocular 0.6 long, 0.7 mm high; temporal formula 1+2; first temporal 2.0 mm 561 long, 1.4 mm high; upper posterior temporals 3.8 mm long, 1.2 mm wide; supralabials 562 seven, third and fourth contacting eye; first supralabial shorter (0.8 mm high) than Journal of Zoological Systematics and Evolutionary Research Page 18 of 81

563 second (1.1 mm high) and smaller in length (0.7 mm) than second (1.0 mm); third 564 supralabial pentagonal, larger in height (1.4 mm) and longer (1.4 mm) than second; 565 sixth supralabial taller than third (1.7 mm); seventh longer than third (2.5 mm) 566 supralabial; symphysial triangular, 1.8 mm wide, 0.5 mm long; first pair of infralabial 567 preventing symphysial–chinshields; infralabials seven, first four contacting chinshields; 568 chinshields 3.5 mm long, 1.7 mm wide; gular scale rows four; preventrals four; ventrals 569 169; subcaudals 18 respectively from left to right side; dorsal scale rows 17/17/17, 570 lacking apical pits and supracloacal tubercles; midbody diameter 10.1 mm (2.6% of 571 SVL); caudal spine 1.1 mm long, smaller than last subcaudal scale (1.3 mm); anal 572 single. Maxillary bone arched upward anteriorly in lateral view, ventral portion curved 573 on anterior and nearlyFor flattened Review on median to posterior Only portion; maxillary with seven 574 teeth; teeth angular in cross section, robust at base, narrower at apices, slightly curved 575 posteriorly; teeth similar in size and spacing; last teeth slightly smaller and in same 576 spacing to anterior ones; maxillary “diastema” absent; lateral process of maxilla well 577 developed. 578 Dorsum of head Brussels brown with interocular stripe Raw umber and center of 579 frontal extending into subocular region; snout with a small Raw umber spot between 580 internasals and posterior part of rostral; supralabials and infralabials trogon yellow 581 [colour 81] to light orange yellow [colour 77]; dorsal scales flesh ocher splecked [colour 582 57] on dorsum of body Brussels brown with Raw umber spots rectangular (four scales 583 wide and two scales long), bordered with cream spots; nuchal band Raw umber with 584 five scales long; paraventral line extending along the body at the level of second dorsal 585 scale row; venter predominantly pale cinnamon with warm sepia dots near cloaca and 586 midventral portion of tail (Figure 8). 587 Colour in life (n=2): Dorsum of head black with snout and parietals orange, 588 followed by black parietal band and black spots bordered by yellowish pigment; belly 589 well-distinct reddish-pink, limited by narrow black paraventral lines. Iris brown with 590 rounded black pupil (Figure 9). The paratype agrees well with the holotype in dorsal 591 colour pattern. Some differences are found in the extension of Raw umber bands being 592 inconspicuous light-bordered. 593 Quantitative variation (n=2): Largest female 390 mm SVL, 30 mm TL; ventrals 594 166–169 (mean=167.5; n=2; SD=2.1) in females; subcaudals 18–20 (mean=19; n=2; 595 SD=1.4) in females; adult midbody diameter 9.2–10.1 mm in females; maxillary teeth 596 six (n=2 sides) or seven (n=2 sides). Page 19 of 81 Journal of Zoological Systematics and Evolutionary Research

597 Distribution: This species occurs only in two localities in upper Juruá river 598 basin, state of Acre, Brazil (Figure 10). 599 Remarks: The type locality where A. nawa was collected in Porto Walter has 600 been deforested for cattle pasture (Vitt, Ávila-Pires, Caldwell, & Oliveira, 1997; P.R. 601 Melo-Sampaio, pers. obs.). The site where paratype was collected is also suffering with 602 deforestation along BR-364 highway. 603 604 Atractus pachacamac sp. nov. 605 urn:lsid:zoobank.org:act:BE85090A-B08C-4AA5-8EAD-3211ED8396C7 606 Atractus snethlageae – Schargel et al. (2013), partim; Yanéz-Muñoz & Venegas 607 (2008); Camper & ZartFor (2014) .Review Only 608 Atractus touzeti – Arteaga et al. (2017). 609 Atractus aff. snethlageae – von May & Mueses-Cisneros (2010). 610 Atractus cf. snethlageae – Rodríguez & Knell (2004). 611 Atractus major – Duellman (1978). 612 Holotype: QCAZ 12630, adult male from Ecuador, Napo, Sumaco Wildlife 613 Sanctuary, Sendero Benavides (0.676ºS, 77.600ºW), 1496 m asl, collected by F. Ayala 614 on 4 April 2014. 615 Paratopotypes (n=4): QCAZ 11986, 12804, adult females, collected by J. 616 Camper on 23 July 2014; QCAZ 10639, adult male collected by J. Camper on 11 July 617 2010; QCAZ 16083, adult male same data as holotype. 618 Paratypes (n=24): Ecuador: Napo, Gonzalo Pizzaro, El Reventador (0.016°N, 619 77.377°W), 1720 m asl, DHMECN 11556, adult female, collected by M. Yánez-Muñoz 620 and team between 11–14 December 2014. Orellana, Dayuma, Pozo Petrolero Capirón 621 (0.665°S, 76.468°W), 240 m asl, EPN 7358, adult female, collected by A. Almendáriz 622 on 17 March 2000; Joyas de los Sachas, San Sebastian del Coca (0.142°S, 76.992°W), 623 350 m asl, EPN 11701, adult female, collected by A. Almendáriz on 6 May 2002. 624 Pastaza, Tzarentza (1.356ºS, 78.058ºW), 1355 m asl, MZUTI 5409, collected by J. 625 Vieira on 10 May 2016. Morona Santiago, General Proaño, Central Hidroeléctrica 626 Abanico (2.245°S, 78.195°W), 1600 m asl, EPN 11454-11455, adult females, collected 627 by Y. Sagredo on 29 March 2007; Puerto Misahuallí, Reserva Biológica Jatún Sacha 628 (1.050ºS, 77.590ºW), 406 m asl, QCAZ 3477, adult female collected by G. Vigle on 20 629 September 1986; Villano, K10 camp (1.476ºS, 77.534ºW), 475 m asl, QCAZ 8367, 630 collected by E. Carrillo on 11 October 2008; Cotundo, at km 32 from the road Journal of Zoological Systematics and Evolutionary Research Page 20 of 81

631 Cocodrilos-Tena (0.681ºS, 77.800ºW), 1476 m asl, QCAZ 11075, adult female, 632 collected by F. Velásquez on 28 November 2010. Zamora-Chinchipe, Yanzatza, Los 633 Encuentros (3.765ºS, 78.502ºW; 1450 m asl), QCAZ 15174, adult female collected by 634 D. Chávez on 18 November 2016, and MECN 8437 adult female collected by P. Galvis 635 on November 2017; Reserva Natural Maycu (4.248°S, 78.658°W), 879 m asl, QCAZ 636 15414, adult female collected by D. Almeida on 28 February 2017; El Zarza (3.776°S, 637 78.497°W), 1450 m asl, adult female QCAZ 15537, collected by P. Baldeón on 2 March 638 2017. Orellana: Nuevo Paraíso, Village Juan Pablo II (0.302ºS, 77.179ºW), 291 m asl, 639 EPN 5471, adult male collected by A. Almendáriz on 12 April 1996; Aguarico 640 (0.792°S, 75.901°W), 217 m asl, EPN 10548, adult male collected by A. Almendáriz 641 and F. Grefa on 23 MarchFor 2006. Review Sucumbios: Lago Only Agrio, El Eno (0.064°S, 76.879°W), 642 278 m asl, FHGO 5813, adult male collected by M. Alcoser. Pastaza: Villano, Kurintza 643 (1.507ºS, 77.511ºW), 362 m asl, QCAZ 8278, juvenile collected at by D. Paucar on 29 644 July 2008; K10 camp (1.475ºS, 77.534ºW), 468 m asl, QCAZ 11833, adult male 645 collected by J. Brito on 1 October 2013. Orellana: Alejandro Labaka, Yasuní National 646 Park (0.492ºS, 76.557ºW) 252 m asl, QCAZ 10614, adult male collected by M. Read on 647 29 January 1996. Napo: El Chaco, Sardinas (0.340ºS, 77.810ºW) 1300 m asl, QCAZ 648 1493, adult male collected by G. Onore on 18 October 1986. Morona-Santiago: Taisha, 649 Makuma, Paatim Shuar Center (2.133ºS, 77.650ºW), 715 m asl, FHGO 10080, adult 650 male collected by native people on 16 April 2014. Zamora-Chinchipe: El Pangui, 651 Concesión Minera Princesa (3.896°S, 78.516°W), 1587 m asl, MECN 13078, adult 652 male collected by D. B. Zapata on 24 January 2016. Peru: Loreto, Datem del Marañon, 653 Cahuapana (5.664°S, 76.839°W), 1150 m asl, CORBIDI 13797 adult male collected by 654 P. J. Venegas on October 2014; Maynas, Güeppi (0.018ºS, 75.358ºW), 220 m asl, 655 CORBIDI 167, adult male, collected by P.J. Venegas and M. Yanez-Muñoz on 22 656 October 2007. 657 Etymology: According to the Inca mythology, Pachacámac or Pacha Kamaq 658 ('creator of the land'; being pacha: 'land' and kamaq: 'creator' or ‘created’ in Quechua 659 language) was a god. The Pachacámac, son of the sun, came to our world and climbed 660 to the highest summit (perhaps a volcano) to throw four stones to the four cardinal 661 points, thereby taking possession of everything that covered his sight and reached his 662 stones. The name also refers to fossorial lifestyle in high elevations. 663 Diagnosis: Atractus pachacamac can be distinguished from all congeners by 664 unique combination of the following characters: (1) smooth dorsal scale rows 17/17/17; Page 21 of 81 Journal of Zoological Systematics and Evolutionary Research

665 (2) postoculars two; (3) loreal moderately long; (4) temporal formula 1+2; (5) 666 supralabials seven, third and fourth contacting eye; (6) infralabials usually eight, first 667 four contacting chinshields; (7) maxillary teeth seven; (8) gular scale rows three; (9) 668 usually two preventrals; (10) ventrals 162–175 in females, 158–167 in males; (11) 669 subcaudals 31–33 in females, 39–45 in males; (12) in preservative, dorsum sepia with 670 saval brown bands; (13) in preservative, venter chamois anteriorly and sepia with 671 chamois spots posteriorly; (14) long body in females (maximum 620 mm SVL) and 672 moderately long in males (maximum 460 mm SVL); (15) tail size moderately long in 673 females (12.1–13.6% of SVL) and long in males (17.5–19.5% of SVL); and (16) 674 hemipenes moderately bilobed (≤ half the length of capitulum), semicapitate, and 675 semicalyculate. For Review Only 676 Comparisons: Atractus pachacamac differs from A. snethlageae in having ≥ 158 677 ventrals, ≥ 39 subcaudals, and > 320 mm SVL in males (vs. ≤ 155 ventrals, ≤ 34 678 subcaudals and < 300 mm SVL in males); from A. nawa in having > 30 subcaudals, tail 679 > 12.0% SVL, and > 600 mm SVL in females, (vs. ≤ 20 subcaudals, tail ≤ 8.0% of 680 SVL, and ≤ 405 mm SVL in females); from A. dapsilis in having > 600 mm SVL, tail > 681 12.0% of SVL in females, ≥ 17.5% of SVL in males, and 31–33 subcaudals in females, 682 39–45 in males (vs. < 600 mm SVL, and < 12.0% of SVL in females, ≤ 17.6% of SVL 683 in males, and 21–26 subcaudals in females, 30–37 in males); from A. ukupacha in 684 having immaculate preventrals, > 600 mm SVL in females and moderately bilobed 685 hemipenes (vs. spotted preventrals, < 500 mm SVL in females, and strongly bilobed 686 hemipenes); from A. schach and A. trefauti in having > 600 mm SVL, tail > 12.0% of 687 SVL in females, > 17% of SVL in males, 162–175 ventrals in females, 158–167 in 688 males, and 31–33 subcaudals in females, 39–45 in males (vs. < 230 mm SVL, tail < 689 12.0% of SVL in A. schach and A. trefauti; 148–150 ventrals in females, 142–151 in 690 males, 19–21 subcaudals in females, 25–31 in males of A. schach; and 153–158 ventrals 691 in females, 139–149 in males, 21–24 subcaudals in females, 24–29 in males of A. 692 trefauti); from A. touzeti in having < 700 mm SVL in females, seven supralabials, third 693 and fourth supralabial contacting eye, seven maxillary teeth, postocular equal in size, 694 and upper and lower temporal posterior equal in size (vs. > 700 mm SVL in females, 695 eight supralabials, fourth and fifth supralabial contacting eye, eight maxillary, very 696 small lower postocular, and very large upper temporal posterior). We refer to Table 1 697 for additional comparisons between Atractus pachacamac and other Amazonian 698 congeners. Journal of Zoological Systematics and Evolutionary Research Page 22 of 81

699 Description of the holotype: SVL 385 mm, tail length 72 mm (18.7% of SVL); 700 head slightly distinct from body; head length 15.1 mm (3.9% of SVL); head width 8.15 701 mm (2.1% of head length); rostral-orbit distance 5.53 mm; nostril-orbit distance 3.92 702 mm; interorbital distance 5.75 mm; head rounded in lateral view; snout rounded in 703 dorsal view, truncate in lateral view; canthus rostralis barely conspicuous; rostral 704 subtriangular in frontal view, 2.42 mm wide, 1.63 mm high, slightly visible in dorsal 705 view; internasal 1.28 mm long, 1.39 mm wide; internasal suture sinistral with respect to 706 prefrontal suture; prefrontal 3.77 mm long, 2.93 mm wide; supraocular subtrapezoidal, 707 1.98 mm long, 1.56 mm wide at broadest point; frontal bell-shaped, 3.89 mm long, 4.34 708 mm wide; parietal 6.51 mm long, 3.28 mm wide; nasal partially divided, nostril located 709 mostly in prenasal; prenasalFor 1.56 Review mm high, 1.06 mmOnly long; postnasal 1.41 mm high, 710 0.94 mm long; loreal 3.15 mm long, 1.0 mm high; eye diameter 1.67 mm; pupil 711 rounded; two postoculars similar in shape and size; upper postocular 1 mm long, 1 mm 712 high; lower postocular 0.66 mm long, 0.95 mm high; temporal formula 1+2; first 713 temporal 2.61 mm long, 1.47 mm high; upper posterior temporal 4.64 mm long, 1.1 mm 714 wide; supralabials seven, second and third contacting loreal, third and fourth contacting 715 eye; first supralabial smaller (1.15 mm high x 0.93 mm long) than second (1.55 mm x 716 1.0 mm); third supralabial pentagonal, higher (1.82 mm) and longer (2.45 mm) than 717 adjacent supralabials; sixth supralabial as tall as third; seventh longer than third (3.82 718 mm) supralabial; symphysial subtriangular, 2.10 mm wide, 0.60 mm long; first pair of 719 infralabials preventing symphysial–chinshields contact; infralabials eight, first four 720 contacting chinshields; chinshields 4.90 mm long, 2.12 mm wide; gular scale rows 721 three; preventrals 3; ventrals 160; subcaudals 42/42 on left and right sides, respectively; 722 dorsal scale rows 17/17/17, lacking apical pits and supracloacal tubercles; midbody 723 diameter 9.95 mm (2.6% of SVL); caudal spine 1.32 mm long, larger than last 724 subcaudal scale (0.89 mm); anal single. 725 Retracted hemipenes extend to the level of 12th subcaudal, bifurcating at 10th 726 subcaudal. Maxillary bone arched upward anteriorly in lateral view, ventral portion 727 curved on anterior and nearly flattened on median to posterior portion; maxillary with 728 seven teeth; teeth angular in cross section, robust at base, narrower at apices, slightly 729 curved posteriorly; teeth similar in size and spacing; last teeth slightly smaller and in 730 same spacing to anterior ones; maxillary diastema absent or indistinct from interspaces; 731 lateral process of maxilla well developed. Page 23 of 81 Journal of Zoological Systematics and Evolutionary Research

732 Dorsum of head cinnamon brown [colour 43] through its extension, covering 733 parietals and two rows of occipital scales; lateral sides of head mikado brown; upper 734 portion of supralabials chamois [colour 84] giving a “moustache” pattern; sixth 735 supralabial with a fuscous spot; postocular and anterior temporal; gular region chamois 736 with infralabials and symphysial region cinnamon-brown spotted, being two last 737 infralabials almost completely chamois; first ventral with a cinnamon-brown spot 738 followed by seven immaculate chamois ventrals; eighth ventral with cinnamon-brown 739 rectangular spot; ninth to fourteenth ventrals immaculate chamois; venter chamois with 740 few dispersed sepia dots until midbody; posterior region of ventral scales sepia 741 pigmented with chamois spots; ventral surface of tail sepia with scattered chamois 742 spots; dorsal ground colourFor sepia Review [colour 279] with Only conspicuous sayal brown bands 743 (29), one-half to two scale rows wide; interspaces between sayal brown [colour 41] 744 bands three to five scales long; first dorsal scale row with chamois and sepia “chess 745 pattern”; dorsal surface of tail sepia with seven conspicuous bands; tip of tail sayal 746 brown. 747 Colour in life of holotype: Rostral, symphisial and first pair of infralabials 748 almost completely dark brownish olive [colour 127]; snout with tiny chrome orange 749 [colour 74] dots; iris rounded and brown; cephalic-cap brownish olive [colour 276]; 750 dorsum sepia [colour 279] with true cinnamon [colour 260] to raw umber [colour 26] 751 bands (Figure 11). 752 Colour in life variation (n=4): Dorsum jet black [colour 300] with salmon 753 [colour 251] and white parietal band; venter white suffused with fuscous [colour 283] in 754 contact between ventral never forming a line in juvenile QCAZ 8278 (Figures 11 and 755 12). 756 Colour in preservative variation (n=27): The paratypes agree with holotype in 757 overall similarity, differing sometimes in visibility of bands by having inconspicuous 758 bands that are olive brown [colour 278] in QCAZ 8367 to cinnamon drab [colour 259] 759 in QCAZ 3476 and QCAZ 10614. All specimens present dorsal bands connected on the 760 vertebral axis. Reverse pattern of colouration is also found in this species. Specimens 761 QCAZ 11986, EPN 10548 and EPN 11701 have dorsum of head beige [colour 254] 762 followed by a dark grayish brown [colour 284] collar; dorsum beige to cinnamon drab 763 [colour 259] with dark grayish brown square spots, each covering two scales. Juveniles 764 possess white parietal band and narrow dorsal bands (one-half to one scale long). Journal of Zoological Systematics and Evolutionary Research Page 24 of 81

765 Hemipenial morphology (n=6): Fully everted and maximally expanded 766 hemipenes renders a moderately bilobed (≤ than capitulum half length), semicapitate, 767 and semicalyculate organ; lobular region as wide as hemipenial body; lobes 768 centrolinearly oriented, attenuated, conical with acute apices; lobes symmetrical, 769 uniformly covered with spinulate calyces on both sides of hemipenes; spinules replaced 770 with papillae toward apices of lobes; capitular groove distinct on both sides of organ; 771 capitulum with transversal spinulated flounces formed by union of horizontal walls of 772 calyces; calyces lacking vertical walls along the sulcate and asulcate faces of capitulum, 773 except for the intrasulcar region with irregular spinulated flounces; transversal 774 calyculate flounces with irregular rows on the intrasulcar region; flounces conspicuous 775 along lateral region ofFor capitulum; Review midportion of sulcate Only face of capitulum with 776 disconnection among transversal flounces; lateral regions of capitulum with regular 777 rows of spinulate calyces; capitulum slightly longer than hempenial body; hemipenial 778 body elliptical covered with enlarged hooked spines; larger spines generally located 779 laterally below sulcus spermaticus bifurcation; distal region of hemipenial body on 780 maximally expanded organ with rows of spines concentrating on the middle of asulcate 781 face; sulcus spermaticus bifurcating approximately at 40% of organ length; sulcus 782 spermaticus margins relatively thick at level of division and along capitular region; 783 sulcus spermaticus not bordered by spinules; basal naked pocket restricted to most basal 784 region of hemipenial body; proximal region of hemipenes body sometimes with shallow 785 groove on the asulcate face, covered with few hooked spines and dispersed spinules; 786 proximal region of hemipenes almost nude (Figure 13). 787 Quantitative variation (n=28): Largest female 620 mm SVL, 75 mm TL; largest 788 male 460 mm SVL, 87 mm TL; ventrals 162–175 (mean=169; n=3; SD=6.5) in females, 789 158–167 (mean=161; n=5; SD=3.6) in males; subcaudals 31–33 (mean=32; n=3; 790 SD=1.1) in females, 39–45 (mean=43; n=5; SD=2.5) in males; supralabials seven (n=16 791 sides); infralabials seven (n=8 sides) or eight (n=8 sides); preventrals two (n=2), three 792 (n=4) or four (n=2); adult midbody diameter 6.8–8.5 mm; maxillary teeth seven (n=14 793 sides). 794 Distribution: Atractus pachacamac inhabits Amazon foothills between 350– 795 1500 m asl from Colombia to Peru, including the surroundings of Sumaco volcano in 796 Ecuador and Peru (Figure 14). 797 Remarks: Camper & Zart (2014) reported earthworms in the diet of this species. 798 After careful examination of museum specimens, we concluded that the specimen listed Page 25 of 81 Journal of Zoological Systematics and Evolutionary Research

799 as A. touzeti (ANF 2590 = MZUTI 5409) by Arteaga et al. (2017) belongs to A. 800 pachacamac. Furthermore, the specimen referred as Atractus snethlageae without 801 voucher, photographed by Arteaga et al. (2017) also is similar to A. pachacamac. Thus, 802 as far we know, A. touzeti is still known only from its type series from the Cordillera de 803 los Guacamayos in northeastern Ecuador (> 1800 m asl; see Schargel et al., 2013 and 804 Passos et al., 2019). 805 806 Atractus akerios sp. nov. 807 urn:lsid:zoobank.org:act:A233B343-8D5E-4DC6-B0F7-98C6E46E17E1 808 Atractus badius – Cunha & Nascimento (1978), partim. 809 Atractus schachFor – Cunha Review & Nascimento (1983) Only; Passos & Fernandes (2008), 810 partim; Prudente & Passos (2008), partim; Prudente, Sarmento, Ávila-Pires, Maschio, 811 & Sturaro (2018); Cunha & Nascimento (1993). 812 Atractus snethlageae – Prudente et al. (2018), partim. 813 Holotype: MPEG 12255, adult male from Brazil, Maranhão, Junco do 814 Maranhão, at Nova Vida (1.822ºS, 46.109ºW), 39 m asl, collected by O.R. da Cunha 815 and F.P. do Nascimento on 1 June 1976. 816 Paratopotypes (n=4): MPEG 15791, adult female collected on 15 August 1979. 817 MPEG 10347, MPEG 15003, and MPEG 15790, adult males collected by O.R. Cunha 818 and F.P. do Nascimento on 31 October 1975, 1 October 1978, and 15 August 1979. 819 Paratypes (n=6): Brazil: Pará: Bragança, Bom Jesus (1.077ºS, 46.861ºW) 44 m 820 asl, MPEG 11374, adult female collected on 7 November 1975; Marabá, São Felix 821 (5.211ºS, 49.046ºW), 102 m asl, MPEG 11569, MPEG 15165 adult males collected by 822 O.R. Cunha and F.P. do Nascimento on 23 September 1976 and on 15 October 1977; 823 Viseu, Bela Vista (1.444ºS, 46.372ºW) 21 m asl, MPEG 2295, MPEG 3713, MPEG 824 10106, adult males collected by O.R. Cunha and F.P. do Nascimento on 21 September 825 1977, 22 May 1973 and 18 June 1976. 826 Etymology: The specific epithet "akerios" in the Greek means lifeless. The word 827 is related to the Greek goddess Keres, (Κῆρες) who personifies violent death and 828 destruction. We draw a parallel, where this new species is named after the sudden 829 disappearance in one of the main areas explored and well-studied in relation to 830 Amazonian snakes (see remarks). 831 Diagnosis: Atractus akerios can be distinguished from all congeners by unique 832 combination of the following characters: (1) smooth dorsal scale rows 17/17/17; (2) Journal of Zoological Systematics and Evolutionary Research Page 26 of 81

833 postoculars two; (3) loreal moderately long; (4) temporal formula 1+2; (5) supralabials 834 usually seven, third and fourth contacting eye; (6) infralabials seven, first four 835 contacting chinshields; (7) maxillary teeth usually six; (8) gular scale rows three; (9) 836 usually four preventrals; (10) ventrals 149–156 in females, 140–154 in males; (11) 837 subcaudals 19–20 in females, 24–33 in males; (12) in preservative, dorsum salmon 838 coloured to antique brown with russet spots; (13) in preservative, venter sayal brown 839 with Verona brown spots; (14) body moderately long in females (maximum 360 mm 840 SVL) and short in males (290 mm SVL); (15) tail short in females (9.4% of SVL) and 841 moderately long in males (11.7–16.3% of SVL); and (16) hemipenes moderately 842 bilobed (≥ half the length of capitulum), semicapitate and semicalyculate. 843 Comparisons: ForAtractus akeriosReview differs from Only A. pachacamac by having ≤ 156 844 ventrals in both sexes, ≤ 20 and 33 subcaudals in females and males, respectively, and < 845 390 mm of maximum SVL (vs. ≥ 158 ventrals in both sexes, ≥ 30 and 39 subcaudals in 846 females and males, respectively, and > 400 mm maximum SVL in both sexes); from A. 847 snethlageae by having usually four preventrals, 31–56 spots, light dorsal colour with 848 small dark spots, ≤ 20 subcaudals in females, and hemipenes presenting attenuated 849 lobes without lateromesial expansion and with dispersed hooked spines entering the 850 proximal region (vs. frequently three preventrals, 24–34 spots, dorsum Vandyke brown 851 with light yellow ocher bands, ≥ 21 subcaudals in females, hemipenes presenting 852 flattened, clavated or conical lobes always with lateromesial expansion and organs 853 without hooked spine entering basalmost region); from A. dapsilis by having usually 854 four preventrals, 31–56 spots, light dorsal colour with small dark spots, ≤ 20 subcaudals 855 in females, and hemipenes presenting attenuated lobes without laterosmesial expansion 856 and with dispersed hooked spines entering the proximal region (vs. usually three 857 preventrals, 24–34 spots, dorsum sepia with light bands, ≥ 21 subcaudals in females, 858 hemipenes presenting flattened, clavated or conical lobes always with lateromesial 859 expansion and organs without hooked spine entering basalmost region); from A. trefauti 860 by having usually four preventrals, 31–56 spots, light dorsal colour with small dark 861 spots, ≤ 20 subcaudals in females, and hemipenes moderately bilobed without 862 conspicuous capitular groove on both sides (vs. three preventrals, 24–34 spots, dorsum 863 sepia with light bands, ≥ 21 subcaudals in females, hemipenes slightly bilobed with 864 evident capitular groove on the asulcate side). We refer to Table 1 for additional 865 comparisons between Atractus akerios and other Amazonian congeners. Page 27 of 81 Journal of Zoological Systematics and Evolutionary Research

866 Description of the holotype: SVL 220 mm, tail length 36 mm (16.4% of SVL); 867 head slightly distinct from body; head length 9.7 mm (4.4% of SVL); head width 5.5 868 mm (56.7% head length); rostral–orbit distance 3.1 mm; nostril–orbit distance 2.25 mm; 869 interorbital distance 3.2 mm; head rounded in lateral view; snout acuminated in dorsal 870 view, truncate in lateral view; canthus rostralis little conspicuous; rostral subtriangular 871 in frontal view, 1.6 mm wide, 0.7 mm high, not visible in dorsal view; internasal 0.8 872 mm long, 0.9 mm wide; internasal suture sinistral with respect to prefrontal suture; 873 prefrontal 2.0 mm long, 1.7 mm wide; supraocular subtrapezoidal, 0.9 mm long, 0.8 874 mm wide at broadest point; frontal bell-shaped, 2.6 mm long, 2.6 mm wide; parietal 4.3 875 mm long, 2.3 mm wide; nasal entirely divided, nostril divided; prenasal 0.8 mm high, 876 0.4 mm long; postnasalFor 0.8 mm Review high, 0.5 mm long; Only loreal 1.7 mm long, 0.7 mm high; 877 second and third supralabials contacting loreal; third and fourth supralabials entering the 878 orbit; eye diameter 1.1 mm; pupil rounded; two postoculars similar in size; upper 879 postocular 0.4 mm long, 0.7 mm high; temporal formula 1+2; first temporal 1.7 mm 880 long, 1.3 mm high; upper posterior temporals 2.9 mm long, 1.1 mm wide; supralabials 881 eight, fourth and fifth contacting eye in the right side; supralabials seven, third and 882 fourth contacting eye in the left side; first supralabial as high (0.9 mm) as second and 883 smaller in length (0.6 mm) than second (0.7 mm); third supralabial pentagonal, larger in 884 height (0.9 mm) and longer (1.1 mm) than second; sixth supralabial taller than third (1.2 885 mm); seventh as long than sixth (1.7 mm) supralabial; symphysial subtriangular, 1.2 886 mm wide, 0.4 mm long; first pair of infralabial preventing symphysial–chinshields; 887 infralabials eight, first four contacting chinshields; chinshields 3.0 mm long, 1.2 mm 888 wide; gular scale rows three; preventrals 4; ventrals 143; subcaudals 33 respectively 889 from left to right side; dorsal scale rows 17/17/17, lacking apical pits and supracloacal 890 tubercles; midbody diameter 5.8 mm (2.6% of SVL); caudal spine 0.9 mm long, larger 891 than last subcaudal scale (0.5 mm); anal single. Hemipenes extend at the level of 7th 892 subcaudal. Maxillary bone arched upward anteriorly in lateral view, ventral portion 893 curved on anterior and nearly flattened on median to posterior portion; maxillary with 894 seven teeth; teeth angular in cross section, robust at base, narrower at apices, slightly 895 curved posteriorly; teeth similar in size and spacing; last teeth slightly smaller and in 896 same spacing to anterior ones; maxillary “diastema” absent or indistinct from 897 interspaces; lateral process of maxilla well developed. 898 Dorsum of head antique brown into tip of parietal, where a light occipital band 899 appears being posteriorly bordered by russet collar; lateral of head antique brown Journal of Zoological Systematics and Evolutionary Research Page 28 of 81

900 covering into sixth supralabial, postoculars and anterior temporal; dorsum salmon to 901 antique brown with russet spots (one or two scales long), light bordered; venter sayal 902 brown with Verona brown spots forming conspicuous midventral line (Figure 15). 903 Colour in life: Unknown. 904 Colour variation (n=11): The dorsal pattern is quite variable, but predominantly 905 Verona brown to cinnamon, with numerous Burnt umber spots with irregular border, 906 disposed irregularly from the vertebral line until ventral borders. Many specimens show 907 a Raw umber narrow line extending above vertebral axis connecting paravertebral spots; 908 vertebral line most visible on the region posterior to incomplete Raw umber occipital 909 band; anterior part of head Verona brown, followed by incomplete clay collar covering 910 parietals and part of temporals;For Review first four pairs of infralabialsOnly and symphysial with 911 irregular Raw umber spots, sometimes extending to chinshields; venter clay colour to 912 tawny olive with small Raw umber spots in each scale, extending in a straight line into 913 cloacal plate (Figure 16). 914 Quantitative variation (n=10): Largest female 360 mm SVL, 34 mm TL; largest 915 male 290 mm SVL, 43 mm TL; ventrals 149–156 (mean=152.5; n=2; SD=4.9) in 916 females, 140–154 (mean=144; n=9; SD=5.1) in males; subcaudals 19–20 (mean=19.5; 917 n=2; SD=0.7) in females, 24–33 (mean=29.2; n=9; SD=2.7) in males; supralabials 918 seven (n=23 sides) or eight (n=1 side); preventrals one (n=2), two (n=1), three (n=4) or 919 four (n=5); adult midbody diameter 5.6–7.4 mm; maxillary teeth six (n=18 sides) or 920 seven (n=6 sides). 921 Hemipenial morphology (n=2): Fully everted and almost maximally expanded 922 hemipenes renders a moderately bilobed (≥ than half capitulum length), semicapitate, 923 and semicalyculate organ; lobular region as wide as hemipenial body; lobes 924 centrolinearly oriented, attenuated with rounded apices; lobes symmetrical; lobes 925 densely covered with spinulate calyces, just above bifurcation of sulcus spermaticus; 926 spinules replaced by irregular papillae toward apices of lobes; capitular groove 927 indistinct on the both sides of organ; capitulum with transversal spinulated flounces 928 formed by union of horizontal walls of calyces; calyces lacking vertical walls along the 929 sulcate and asulcate faces of capitulum, except for the intrasulcar region with irregular 930 spinulated flounces; transversal calculate flounces with irregular rows on the intrasulcar 931 region; flounces conspicuous along lateral region of capitulum; midportion of sulcate 932 face of capitulum almost nude by interruption of transversal flounces; lateral regions of 933 capitulum with regular and conspicuous rows of spinulate calyces; hemipenial body Page 29 of 81 Journal of Zoological Systematics and Evolutionary Research

934 elliptical covered with enlarged hooked spines; larger spines generally located laterally 935 below sulcus spermaticus bifurcation; distal region of hemipenial body on maximally 936 expanded organ with rows of spines concentrating on the middle of asulcate face; sulcus 937 spermaticus bifurcates approximately on the 50% of organ length; sulcus spermaticus 938 margins relatively thick at level of division and along the capitular region; sulcus 939 spermaticus narrow, not bordered by spinules; basal naked pocket extending along of 940 hemipenial body, with a large spine in its base; proximal region of hemipenes body 941 covered with few hooked spines and dispersed spinules (Figure 20A). 942 Distribution: Atractus akerios is found in primary forest, except one individual 943 that was found in the secondary forest (Cunha & Nascimento, 1983), in five sites at 944 extreme eastern portionFor of the BrazilianReview Amazonia, Only being one in Junco do Maranhão, 945 state of Maranhão and four in state of Pará between 20–110 m asl (Figure 17). 946 Remarks: Cunha & Nascimento (1978) reported Atractus akerios as A. badius 947 from six localities in the state of Pará and one in the state of Maranhão in Brazil. They 948 mentioned colour variation ranging from sepia to reddish brown with dark transversal 949 bands, and white or yellowish narrow bands disposed irregularly. Posteriorly, the darker 950 specimens were described as Atractus flammigerus snethlageae (Cunha & Nascimento, 951 1983) whereas the other specimens were associated to A. schach (Cunha & Nascimento, 952 1984) until recently (see Melo-Sampaio et al., 2019). Since we did not obtain new 953 records of specimens collected in the region, the species may have experienced a 954 population decline following changes in land use on this region. 955 956 Atractus ukupacha sp. nov. 957 urn:lsid:zoobank.org:act:C95B553B-1F95-46B5-A73B-4D8AEA48D81D 958 Atractus snethlageae – Schargel et al. (2013), partim; Maynard et al. (2017). 959 Holotype: QCAZ 12504, adult male from Ecuador, Napo, El Chaco (0.371ºS, 960 77.821ºW), 1606 m asl, collected by P. Medrano, on 16 April 2014. 961 Paratypes (n=18): Ecuador: Napo, El Reventador (0.041ºS, 77.526ºW), 1543 m 962 asl, QCAZ 444, MNRJ 24596 (formerly QCAZ 205) adult females collected by G. 963 Onore on 10 January 1985 and 1 January 1986; collected at San Francisco de Borja 964 (0.424ºS, 77.837ºW; 1500 m asl), MNRJ 24597, (formerly QCAZ 1320), adult female 965 collected by G. Scacco on 18 April 1992; Sardinas (0.340ºS, 77.810ºW), 1300 m asl, 966 QCAZ 1494, adult female collected by G. Onore on 18 October 1992; Santa Rosa, 967 Quijos (0.394ºS, 77.822ºW), 1623 m asl, QCAZ 12715, adult female collected by P. Journal of Zoological Systematics and Evolutionary Research Page 30 of 81

968 Medrano on 24 April 2014. San Rafael, San Rafael stream (0.103ºS, 77.580ºW), 1190 969 m asl, QCAZ 0004, 3256, adult males collected by G. Onore on 10 January 1984 and 7 970 April 1996; San Francisco de Borja (0.438ºS, 77.820ºW), 1670 m asl, QCAZ 1606, 971 adult male collected by V. Utreras on 4 January 1992. El Chaco (0.340ºS, 77.810ºW), 972 1600 m asl, QCAZ 4047, collected by R. Cárdenas on 2 November 2006. San Francisco 973 de Borja (0.403ºS, 77.834ºW), 1665 m asl, QCAZ 12596, adult male by P. Medrano on 974 17 April 2014; San Francisco de Borja, Borja (0.433°S, 77.849°W), 1744 m asl, 975 DHMECN 80, adult male collected by E. Asanza on September 1980. Piedra Fina 976 (0.128ºS, 77.609ºW), 1299 m asl, QCAZ 4812, 4942–44, juveniles collected by M. 977 Wilkinson on 18 March 2012; Orellana, San José de Payamino (0.464ºS, 77.297ºW), 978 370 m asl, QCAZ 11651–52,For juveniles Review collected byOnly R. Lynch on 8 June 2013. 979 Etymology: According to Inca mythology, Uku Pacha represents the 980 underworld, containing the dead and everything that was under terrestrial or aquatic 981 surface. The general legend considers the fountains, caves, volcanoes and any opening 982 in the earth's crust as a means of communication between Uku Pacha and Kay Pacha. 983 The word Pacha in Quechua Indian language means “time and space”, but into a more 984 general sense means “Earth”. Thus, the specific epithet allows inference to this semi- 985 fossorial lifestyle on the Reventador Volcano. 986 Diagnosis: Atractus ukupacha can be distinguished from all congeners by the 987 following combination of characters: (1) smooth dorsal scale rows 17/17/17; (2) 988 postoculars two; (3) loreal moderately long; (4) temporal formula 1+2; (5) supralabials 989 seven, third and fourth contacting eye; (6) infralabials eight, first four contacting 990 chinshields; (7) maxillary teeth seven; (8) gular scale rows three; (9) preventrals two; 991 (10) ventrals 161–170 in females, 153–165 in males; (11) subcaudals 23–33 in females, 992 38–42 in males; (12) in preservative, dorsum dusky brown with olive-brown bands; (13) 993 in preservative, venter olive-brown with small cream white dots; (14) body moderately 994 long in females (maximum 464 mm SVL) and males (maximum 390 mm SVL); (15) 995 tail moderately long in female (9.9–12.7% of SVL) and long in males (15.5–20.6% of 996 SVL); and (16) hemipenes strongly bilobed (≥ length of capitulum), semicapitate, and 997 semicalyculate. 998 Comparisons: Atractus ukupacha differs from A. dapsilis, A. schach, A. trefauti 999 by having 161–170 ventrals in females, 153–165 in males, 23–33 subcaudals in 1000 females, 38–42 in males, and belly olive-brown with small cream white dots (vs. 1001 ventrals 148–150 in females, 142–151 in males, 19–21 subcaudals in females, 25–32 in Page 31 of 81 Journal of Zoological Systematics and Evolutionary Research

1002 males of A. schach; 30–37 subcaudals in males of A. dapsilis; ventrals 153–158 in 1003 females, 139–149 in males, 21–24 subcaudals in females, 24–29 in males of A. trefauti; 1004 and belly mottled in A. schach and A. trefauti, and cream in A. dapsilis usually with 1005 brown dots forming an inconspicuous midline); from A. schach by having hemipenes 1006 with conspicuous capitular groove, 161–170 ventrals in females, 153–165 in males; 23– 1007 33 subcaudals in females, 38–42 in males, and tail > 15% of SVL in males (vs. 1008 hemipenes without capitular groove, 148–150 ventrals in females, 142–151 in males, 1009 19–21 subcaudals in females, 25–32 in males, and tail < 15% of SVL in males); from A. 1010 trefauti by having tail > 13.3% of SVL and strongly bilobed hemipenes in males (vs. 1011 shorter tail < 13.2% and slightly bilobed hemipenes in males); from A. nawa by having 1012 eight infralabials, and For≥ 23 subcaudals Review and > 9% ofOnly SVL in females (vs. seven 1013 infralabials, < 21 subcaudals, and tail < 9% of SVL in females); from A. pachacamac by 1014 having spotted preventrals, < 500 mm SVL in females, and hemipenes strongly bilobed 1015 with lobes centrifugally oriented and flattened on the apices (immaculate preventrals, > 1016 600 mm SVL in females, and hemipenes moderately bilobed with lobes centrolinearly 1017 oriented and attenuated on the apices); from A. akerios by having ≥ 23 and ≥ 38 1018 subcaudals in females in males, respectively, ≥ 161 and 153–165 ventrals in females 1019 and males respectively, and > 390 mm SVL (≤ 20 and ≤ 33 subcaudals in females and 1020 males, respectively, ≤ 156 and 140–154 ventrals in females and males respectively, and 1021 maximum SVL < 360 mm); from A. snethlageae by lacking light parietal band, present 1022 > 154 ventrals in males, and > 38 subcaudals in males (vs. presence of incomplete light 1023 parietal band, < 154 ventrals and < 35 subcaudals in males); touzeti by having < 700 1024 mm SVL in females, seven supralabials, third and fourth supralabial contacting eye, 1025 seven maxillary teeth, postocular equal in size, and upper and lower temporal posterior 1026 equal in size (vs. > 700 mm SVL in females, eight supralabials, fourth and fifth 1027 supralabial contacting eye, eight maxillary, very small lower postocular, and very large 1028 upper temporal posterior). We refer to Table 1 for additional comparisons between 1029 Atractus ukupacha and other Amazonian congeners. 1030 Description of the holotype: SVL 310 mm, tail length 60 mm (19.3% of SVL); 1031 head slightly distinct from body; head length 11.6 mm (3.7% of SVL); head width 7.1 1032 mm (61.2% head length); rostral–orbit distance 4.3 mm; nostril–orbit distance 3.1 mm; 1033 interorbital distance 4.3 mm; head rounded in lateral view; snout truncate in dorsal 1034 view, truncate in lateral view; canthus rostralis little conspicuous; rostral subtriangular 1035 in frontal view, 2.2 mm wide, 1.1 mm high, well visible in dorsal view; internasal 1.0 Journal of Zoological Systematics and Evolutionary Research Page 32 of 81

1036 mm long, 1.1 mm wide; internasal suture sinistral with respect to prefrontal suture; 1037 prefrontal 3.0 mm long, 2.5 mm wide; supraocular subtrapezoidal, 1.5 mm long, 1.3 1038 mm wide at broadest point; frontal bell-shaped, 2.7 mm long, 3.3 mm wide; parietal 4.7 1039 mm long, 3.1 mm wide; nasal entirely divided, nostril divided; prenasal 1.1 mm high, 1040 0.6 mm long; postnasal 1.1 mm high, 0.8 mm long; loreal 2.1 mm long, 0.7 mm high; 1041 second and third supralabials contacting loreal; third and fourth supralabials entering the 1042 orbit; eye diameter 1.5 mm; pupil rounded; two postoculars distinct in height, being 1043 lower shorter than upper; upper postocular 0.7 mm long, 1.0 mm high; lower postocular 1044 0.4 long, 0.7 mm high; temporal formula 1+2; first temporal 1.9 mm long, 1.5 mm high; 1045 upper posterior temporals 3.7 mm long, 1.3 mm wide; supralabials seven, third and 1046 fourth contacting eye;For first supralabial Review shorter (1.0 Only mm high) than second (1.4 mm high) 1047 and smaller in length (0.6 mm) than second (1.0 mm); third supralabial pentagonal, 1048 larger in height (1.5 mm) and longer (1.5 mm) than second; sixth supralabial as tall as 1049 third; seventh longer than third (2.5 mm) supralabial; symphysial triangular, 1.7 mm 1050 wide, 0.6 mm long; first pair of infralabial preventing symphysial–chinshields; 1051 infralabials eight, first four contacting chinshields; chinshields 4.0 mm long, 1.5 mm 1052 wide; gular scale rows three; preventral one; ventrals 165; subcaudals 38 respectively 1053 from left to right side; dorsal scale rows 17/17/17, lacking apical pits and supracloacal 1054 tubercles; midbody diameter 8.5 mm (2.7% of SVL); caudal spine 1.8 mm long, twice 1055 larger than last subcaudal scale (0.9 mm). Retracted hemipenes extends to the level of 1056 12th subcaudal, bifurcating at 10th. Maxillary bone arched upward anteriorly in lateral 1057 view, ventral portion curved on anterior and nearly flattened on median to posterior 1058 portion; maxillary with seven teeth; teeth angular in cross section, robust at base, 1059 narrower at apices, slightly curved posteriorly; teeth similar in size and spacing; last 1060 teeth slightly smaller and in same spacing to anterior ones; maxillary “diastema” absent; 1061 lateral process of maxilla well developed. 1062 Dorsum of head dusky brown through its extension; supralabials and lateral 1063 portion of head dusky brown; first four infralabials and gular region olive-brown burnt 1064 umber spotted; venter olive-brown with few dispersed cream white dots; lateral portions 1065 of ventral scales with cream white spots; ventral surface of tail entirely fuscous; dorsal 1066 ground colour dusky brown with conspicuous olive-brown bands (33); interspaces 1067 between olive-brown bands covered by three to five scales long; first dorsal scale row 1068 with light cream, forming spots in contact with ventral scales; dorsal surface of tail 1069 dusky brown with conspicuous six bands; ventral surface of tail fuscous (Figure 18). Page 33 of 81 Journal of Zoological Systematics and Evolutionary Research

1070 Colour in life: Dorsum of head dusky brown through its extension; internasals, 1071 prefrontals, loreal, supralabials and lateral portion of head dusky brown with orange 1072 flecks; first four infralabials and gular region olive-brown burnt umber spotted; dorsal 1073 ground colour dusky brown with conspicuous olive-brown bands; interspaces between 1074 olive-brown bands covered by three to five scales long and tiny orange flecks; first 1075 dorsal scale row with light cream, forming spots in contact with ventral scales (Figure 1076 19). 1077 Colour variation (n=17): Dorsum of head olive brown with inconspicuous dark 1078 grayish brown [colour 284]; dorsum of body sepia to dusky brown [colour 285] with 1079 dorsal bands varying from pale cinnamon [colour 55] to cinnamon [colour 21], with one 1080 to two scales long; chinshields,For Review gulars and anterior Only portion of venter cream suffused 1081 with fuscous [colour 283] dots; paraventral line connecting light band sometimes 1082 present near cloaca. Juveniles present a white parietal band with medial suture diffuse 1083 with dark pigments, and dorsal bands two scales long (Figure 19). 1084 Hemipenial morphology (n=6): Fully everted and almost maximally expanded 1085 hemipenes renders a strongly bilobed (≥ than the entire capitulum), semicapitate, and 1086 semicalyculate organ; lobular region twice as wide as hemipenial body; lobes 1087 centrifugally oriented, clavate with nearly flattened apices on the medial region and 1088 laterally conical; lobes uniformly covered with spinulate calyces on both sides of 1089 hemipenes; spinules replaced by irregular papillae toward apices of lobes; capitular 1090 groove distinct on the both sides of organ; capitulum with transversal spinulated 1091 flounces formed by union of horizontal walls of calyces; calyces lacking vertical walls 1092 along the sulcate and asulcate faces of capitulum, except for the intrasulcar region with 1093 irregular spinulated flounces; transversal calyculate flounces with irregular rows on the 1094 intrasulcar region; flounces very concentrated and irregular distributed along lateral 1095 region of capitulum; high concentration of calyces on the sulcate side of organ 1096 sometimes forming longitudinal crests; capitulum slightly shorter than hemipenial body; 1097 hemipenial body elliptical covered with enlarged hooked spines; larger spines generally 1098 located laterally below sulcus spermaticus bifurcation; distal region of hemipenial body 1099 on maximally expanded organ with rows of spines concentrating on the middle of 1100 asulcate face; sulcus spermaticus bifurcates approximately on the 50% of organ length; 1101 sulcus spermaticus margins relatively thick at the level of division and along the 1102 capitular region; sulcus spermaticus not bordered by spinules; basal naked pocket runs Journal of Zoological Systematics and Evolutionary Research Page 34 of 81

1103 beyond sulcus spermaticus bifurcation; proximal region of hemipenes body covered 1104 with few hooked spines and dispersed spinules (Figure 20B–F). 1105 Quantitative variation (n=17): Largest female 464 mm SVL, 46 mm TL; largest 1106 male 390 mm SVL, 75 mm TL; ventrals 161–170 (mean=166; n=6; SD=3.7) in females, 1107 154–165 (mean=159; n=8; SD=1.5) in males; subcaudals 23–31 (mean=27; n=6; 1108 SD=3.6), in females, 38–42 (mean=39; n=8; SD=1.5) in males; supralabials seven 1109 (n=27 sides), or eight (n=1 sides); infralabials seven (n=6 sides) or eight (n=22 sides); 1110 preventrals two (n=4), three (n=8) or four (n=2); adult midbody diameter 8.5 mm. 1111 Distribution: Known from San José de Payamino (370 m asl) to Quijos river 1112 Valley on elevations (1800 m asl), Ecuador (Figure 21). 1113 Remarks: AlthoughFor Atractus Review ukupacha was Only found in a site near the type locality 1114 of A. touzeti, they occupy different altitudes (1600 m asl for A. ukupacha versus 2200 m 1115 asl for A. touzeti). Thus, ecological requirements of these two species are probably quite 1116 different among them and the recently discovered Atractus atlas at elevations >1700 m 1117 asl (Passos et al., 2019). 1118

1119 4 DISCUSSION 1120 We assessed the systematics of the Atractus snethlageae complex based on 1121 multiple and putatively independent characters (DNA sequences, measurements, 1122 scutelation, colour pattern, and male genitalia). Atractus snethlageae was described as a 1123 subspecies of A. flammigerus without any evidence of close relationships except for 1124 sharing an external banded-pattern. We have recovered here both taxa on distinct clades 1125 (Figure 1). Our phylogeny is consistent with a recent study focusing on the Atractus 1126 schach species complex (Melo-Sampaio et al., 2019). Populations of Atractus 1127 snethlageae sensu stricto are parapatric with all other species in the complex, except for 1128 allopatric A. ukupacha. All species were diagnosed upon differences in body size, 1129 meristic data and hemipenial morphology (Figures 5, 13 and 20). Prior to this work, the 1130 hemipenis of A. snethlageae was never illustrated or even properly described, although 1131 this character is very useful for species delimitation (and identification) in the genus 1132 Atractus (Passos et al., 2018). We describe the male genitalia based on the paratype of 1133 A. snethlageae (MPEG 10137) and show its variation across its distribution. In this 1134 sense, we hope to encourage more descriptions/preparations of this important source of 1135 characters for both new and old species. Page 35 of 81 Journal of Zoological Systematics and Evolutionary Research

1136 Although species of the former A. snethlageae complex are similar in 1137 morphology to other banded species as A. atlas, A. dapsilis, A. major, A. schach, A. 1138 touzeti and A. trefauti, hemipenial morphology markedly differs among them (condition 1139 unknown in A. atlas and A. touzeti) and molecular data support their recognition as 1140 distinct taxa. For example, both Atractus pachacamac and Atractus ukupacha were 1141 sometimes misidentified as Atractus major. Both species can be easily diagnosed by the 1142 number of infralabials in contact with chinshields. Although Atractus atlas and A. 1143 touzeti also have banded dorsal patterns, they are giant species that differ markedly 1144 from A. snethlageae complex in scutelation, morphometrics and colouration. 1145 Interestingly, the banded dorsal pattern has evolved multiple times in Atractus (e.g., A. 1146 multicinctus and A. flammigerusFor Review clades also share Onlythis condition). The discovery of 1147 additional new species of the A. snethlageae complex is expected (see above), 1148 especially in Peru and Colombia, where sampling is scanty and diversity is high (see 1149 Meneses-Pelayo & Passos, 2019). Further taxonomic studies (using an expanded 1150 molecular dataset and hemipenial morphology) are needed to evaluate the status of 1151 other related clades. 1152 Atractus akerios, A. nawa and A. ukupacha have apparently restricted 1153 distributions and probably through denser sampling it is possible to detect other distinct 1154 lineages also with narrow ranges of distribution. Many recent studies have pointed that 1155 the diversity of the Amazonian herpetofauna is underestimated due to cryptic diversity 1156 (Melo-Sampaio et al., 2018, 2019, 2020; Rojas et al., 2018). The increasing efforts to 1157 accurately delimit species in the Neotropics have profited from the museum specimens 1158 with available tissue samples for molecular studies, allowing falsification of previous 1159 hypotheses with premature or misguided taxonomic decisions. We have found a striking 1160 diversity of Atractus throughout its distribution in the last decade, mostly triggered by 1161 the study of a new system of phenotypic characters (Melo-Sampaio et al., 2019; Passos 1162 et al., 2017, 2018). On the other hand, many taxa along Amazonian lowlands remain 1163 hidden in the shelves of museum collections without proper systematic studies. For 1164 example, Atractus altagratiae was described only five decades after its own collection 1165 event (Passos & Fernandes, 2008), while A. hoogmoedi still remained misidentified as 1166 A. zidoki for ~40 years after the collection of its type-specimens (Prudente & Passos, 1167 2010). In the last two decades, herpetological collections from Brazil, Ecuador and Peru 1168 have increased abruptly. Most of these collections come from environmental studies for 1169 high-impact projects, allowing biologists to reach to unexplored corners of the Amazon Journal of Zoological Systematics and Evolutionary Research Page 36 of 81

1170 rainforest. In fact, current fieldwork and sampling efforts throughout the Amazonian 1171 lowlands is paradoxical, as the best-studied snake faunas are those sampled in the 1172 course of ‘development projects’ (e.g., oil exploration, hydroelectric plants, mining, 1173 monocultures) that will most likely have a negative impact on biodiversity (Marçal et 1174 al., 2011; Morato et al., 2014; Prudente, Maschio, Santos-Costa, & Feitosa, 2010). 1175 We have found areas with high alpha diversity of Atractus in both eastern and 1176 western extremes of Amazonia. Although we were unable to obtain more tissue samples 1177 from western Amazonia (Colombia and Peru), we believe that better sampling including 1178 taxa that occur there can improve significantly our understanding of the relationships of 1179 Atractus. Current documentation of biodiversity is crucial to policies and management 1180 of natural resources. BlackburnFor Review et al. (2019) show Only a recent tendency to describe cryptic 1181 species using non-morphological (e.g., DNA sequence) data that may impact positively 1182 by shortening branch lengths in the tree of life. We highlight that eastern Amazonia is 1183 probably the most threatened area of the biome, since 2/3 of its territory has been 1184 deforested (Silva, Rylands, & Fonseca, 2005; Sonter et al., 2017), although small 1185 patches of forest harbor a substantial number of reminiscent snakes (Ávila-Pires et al., 1186 2018). Based on our results and considering previous studies on Atractus (Passos et al., 1187 2017, 2018; Prudente & Passos, 2010), at least four species are restricted to this region 1188 (i.e., A. alphonsehogei, A. hoogmoedi, A. akerios, and A. tartarus), suggesting it should 1189 be a relevant conservation area for snakes. 1190 Deforestation, in the 21st century has predominantly occurred in the south- 1191 eastern part of the Amazon, making the seasons drier and more frequent (Staal et al., 1192 2020); with currently deforestation rates, we will face rainfall decreasing up to 20% 1193 with consequences in resilience and regrowth (Zemp, Schleussner, Barbosa, & Rammig, 1194 2017). Studies in the region are crucial to improve the knowledge about impacts on the 1195 biota experiencing fast change on vegetation covering and land use. However, we are 1196 just filling some gaps in the knowledge on snake diversity from western Amazonia 1197 (Nogueira et al., 2019). 1198 1199 ACKNOWLEDGMENTS 1200 We thank to Dr. Stephan Koblmüller and two anonymous reviewers by suggestions that 1201 improved the manuscript. We are grateful to the following people for allowing us to 1202 examine specimens and/or loan tissue samples under their care: C. Aguilar and J. 1203 Cordova (MUSM), A. Almendáriz (EPN), J. Aparicio (CBF), M.B. Souza (UFAC-RB), Page 37 of 81 Journal of Zoological Systematics and Evolutionary Research

1204 P. Bernarde, R. Machado and L. Turci (UFACF), Y. Bittar (UFPA), J. Lynch and M. 1205 Calderón (ICN), P. Campbell (NHM), R. Cassalas and J. Espítia (MLS), F. Curcio 1206 (UFMT-R), A. Fouquet (CNRS), F. Franco and F. Grazziotin (IBSP), L. Gonzalez and 1207 R. Montaño (MNK), M. Messias (UNIR), D. Salazar-Valenzuela (MZUTI), J. Valencia 1208 (FHGO), W. Vaz (PUC-GO), F. Werneck (INPA), M. Yanez-Muñoz (DHMECN), and 1209 H. Zaher (MZUSP). We deeply indebted to S. Albuquerque, R. Gaiga, R. Lynch, J. 1210 López-Rojas, M. Martins, R. Maynard, K. Mebert, D. Meneghelli, L. Moraes, S. 1211 Morato, R. Oliveira, U. Oliveira, D. Neira, D. Paiva, M. Pingleton, D. Quirola, D. 1212 Rosenberg, W. Vaz and L. Vitt for providing photographs of live specimens. We thank 1213 D. Fernandes, R. Fernandes, B. Jennings and P. Pinna for helpful suggestion on early 1214 version of the manuscript.For PRMS Review thanks L. Lima, Only K. Silva, and R. Oliveira for 1215 friendship and patience during this study. PRMS especially thank to I. Prates 1216 (University of Michigan) for sharing his time and knowledge with prolific discussions 1217 and M. T. Rodrigues (USP) for all supporting with workspace, samples and loan of 1218 specimens. PRMS thanks K. de Queiroz, E. Langan and J. Poindexter (USNM), D.A. 1219 Kizirian, L. Vonnnahme (AMNH) for information and photos of specimens under their 1220 care. PRMS also acknowledge QCAZ team (Y. Sagredo, M. Stacey, S. Valverde, F. 1221 Ayala-Varela, E. Guerra, V. Chasiluisa, M. Navarrete, G. Maldonado, and S. Guamán), 1222 CORBIDI team (J. Chavez-Arribasplata, A. Barbosa), MUSM team (J. Villegas, J. Cusi 1223 and G. Gutierrez) for logistic facilities at Quito and Lima, respectively. Financial 1224 support for PRMS was provided by Coordenação de Aperfeiçoamento de Pessoal de 1225 Nível Superior (CAPES). Financial support for PP was provided by Conselho Nacional 1226 de Desenvolvimento Científico e Tecnológico (#309560/2018-7) and Fundação Carlos 1227 Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (#E-26/202.737/2018 1228 and #E-26/211.154/2019). Financial support for ALCP was provided by Conselho 1229 Nacional de Desenvolvimento Científico e Tecnológico (#302611/2018-5 and 1230 #440413/2015-0). Molecular work and fieldwork by OTC were funded by SENESCYT 1231 under the ‘Arca de Noé’ Initiative (PIs: S. Ron and OTC). 1232 1233 ORCID 1234 Paulo Roberto Melo-Sampaio 0000-0003-1858-1643 1235 Paulo Passos 0000-0002-1775-0970 1236 Ana Prudente 0000-0002-4164-6815 1237 Pablo Venegas 0000-0002-6501-4492 Journal of Zoological Systematics and Evolutionary Research Page 38 of 81

1238 Omar Torres-Carvajal 0000-0003-0041-9250 1239 1240 REFERENCES 1241 Arteaga, A., Mebert, K., Valencia, J. H., Cisneros-Heredia, D. F., Peñafiel, N., Reyes- 1242 Puig, C., Vieira-Fernandes, J. L., & Guayasamin, J. M. (2017). Molecular 1243 phylogeny of Atractus (Serpentes, Dipsadidae), with emphasis on Ecuadorian 1244 species and the description of three new taxa. ZooKeys, 661, 91–123. 1245 doi:10.3897/zookeys.661.11224 1246 Ávila-Pires, T. C. S., Alves-Silva, K. R., Barbosa, L., Correa, F. S., Cosenza, J. F. A., 1247 Costa-Rodrigues, A. P. V., Cronemberger A. A., Hoogmoed, M. S., Lima-Filho, 1248 G. R., Maciel, ForA. O., Missassi, Review A. F. R., Nascimento, Only L. R. S., Nunes, A. L. S., 1249 Oliveira, L. S., Palheta, G. S., Pereira Jr., A. J. S., Pinheiro, L., Santos-Costa, M. 1250 C., Pinho, S. R. C., Silva, F. M., Silva, M. B., & Sturaro, M. J. (2018). Changes 1251 in amphibian and reptile diversity over time in Parque Estadual do Utinga, Pará 1252 State, Brazil, a protected area surrounded by urbanization. Herpetology Notes, 1253 11, 499–512. 1254 Ávila-Pires, T. C. S., Vitt, L. J., Sartorius, S. S., & Zani, P. A. (2009). Squamata 1255 (Reptilia) from four sites in southern Amazonia, with a biogeographic analysis 1256 of Amazonian lizards. Boletim do Museu Paraense Emílio Goeldi Ciências 1257 Naturais, 4, 99–118. 1258 Bernarde, P. S., & Abe, A. S. (2006). A snake community at Espigão do Oeste, 1259 Rondônia, southwestern Amazon, Brazil. South American Journal of 1260 Herpetology, 1, 102–113. 1261 Bernarde, P. S., Albuquerque, S., Barros, T. O., & Turci, L. C. B. (2012). Serpentes do 1262 estado de Rondônia, Brasil. Biota Neotropica, 12, 154–182. 1263 Bernarde, P. S., Machado, R. A., & Turci, L. C. B. (2011). Herpetofauna of Igarapé 1264 Esperança area in the Reserva Extrativista Riozinho da Liberdade, Acre – Brazil. 1265 Biota Neotropica, 11, 117–144. 1266 Bernarde, P. S., Turci, L. C. B., & Machado, R. A. (2017). Serpentes do Alto Juruá: 1267 Acre-Amazônia brasileira. Rio Branco: Edufac. 1268 Blackburn, D. C., Giribet, G., Soltis, D. E., & Stanley, E. L. (2019). Predicting the 1269 impact of describing new species on phylogenetic patterns. Integrative 1270 Organismal Biology, 1, 1–12. doi:10.1093/iob/obz028 Page 39 of 81 Journal of Zoological Systematics and Evolutionary Research

1271 Boie, F. (1827). Bemerkungen über Merrem's Versuch eines Systems der Amphibien, 1. 1272 Lieferung: Ophidier. Isis von Oken, 20, 508–566. 1273 Boulenger, G. A. (1894). Catalogue of the snakes in the British Museum (Natural 1274 History). Volume II. London: Trustees of the British Museum. 1275 Böhm, M., Collen, B., Baillie, J. E. M., Bowles, P., Chanson, J., Cox, N., Hammerson, 1276 G., Hoffmanng, M., Livingstone, S. R., Ram, M., Rhodin, A. G. J., Stuart, S. N., 1277 van Dijk, P. P., Young, B. E., Afuang, L.E., Aghasyan, A., García, A., Aguilar, 1278 C., Ajtic, R., Akarsu, F., Alencar, L. R. V., Allison, A., Ananjeva, N., Anderson 1279 S., Andrén, C., Ariano-Sánchez, D., Arredondo, J.C., Auliya, M., Austin, C. C., 1280 Avci, A., Bakeraf, P. J., Barreto-Lima, A. F., Barrio-Amorós, C. L., Basu, D., 1281 Bates, M. F., Batistella,For ReviewA., Bauer, A., Bennett, Only D., Böhm, W., Broadley, D., 1282 Brown R., Burgess, J., Captain, A., Carreira, S., Castañeda, M. R., Castro, F., 1283 Catenazzi, A., Cedeño-Vázquez, J. R., Chappleay, D. G., Cheylan, M., Cisneros- 1284 Heredia, D. F., Cogalniceanu, D., Cogger, H., Corti, C., Costa, G. C., Couper, P. 1285 J., Courtney, T., Crnobrnja-Isailovic, J., Crochet, P-A., Crother, B., Cruz, F., 1286 Daltry, J. C., Daniels, R. J. R., Das, I., de Silva, A., Diesmos, A. C., Dirksen, L., 1287 Doan, T. M., Dodd Jr., C. K., Doody, J. S., Dorcas, M. E., Filho, J. D. B., Egan, 1288 V. T., El Mouden, E. H., Embert, D., Espinoza, R. E., Fallabrino, A., Feng, X., 1289 Feng, Z-J., Fitzgerald, L., Flores-Villela, O., França, F. G. R., Frost, D., 1290 Gadsden, H., Gamble, T., Ganesh, S. R., Garcia, M. A., García-Pérez, J. E., 1291 Gatus, J., Gaulke, M., Geniez, P., Georges, A., Gerlach, J., Goldberg, S., 1292 Gonzalez, J-C. T., Gower, D. J., Grant, T., Greenbaum, E., Grieco, C., Guo, P., 1293 Hamilton, A. M., Hare, K., Hedges, S. B., Heideman, N., Hilton-Taylor, C., 1294 Hitchmough, R., Hollingsworth, B., Hutchinson, M., Ineich, I., Iverson, J., 1295 Jaksic, F. M., Jenkins, R., Joger, U., Jose, R., Kaskado, Y., Kaya, U., Keogh, J. 1296 S., Köhler, G., Kuchling, G., Kumlutaş, Y., Kwet, A., La Marca, E., Lamar, W., 1297 Lane, A., Lardner, B., Latta, C., Latta, G., Lau, M., Lavin, P.,Lawson, D., Le 1298 Breton, M., Lehr, E., Limpus, D., Lipczynski, N., Lobo, A. S., López-Luna, M. 1299 A., Luiselli, L., Lukoschekek, V., Lundberg, M., Lymberakis, P., Macey, R., 1300 Magnusson, W. E., Mahler, D. L., Malhotra, A., Mariaux, J., Maritz, B., 1301 Marques, O. A. V., Márquez, R., Martins, M., Masterson, G., Mateo, J. A., 1302 Mathew, R., Mathews, N., Mayer, G., McCranie, J. R., Measey, G. J., Mendoza- 1303 Quijano, F., Menegon, M., Métrailler, S., Milton, D. A., Montgomery, C., 1304 Morato, S. A. A., Mott, T., Muñoz-Alonso, A., Murphy, J., Nguyenao, T. Q., Journal of Zoological Systematics and Evolutionary Research Page 40 of 81

1305 Nilson, G., Nogueira, C.,Núñez, H., Orlov, N., Ota, H., Ottenwalder, J., 1306 Papenfuss, T., Pasachnik, S., Passos, P., Pauwels, O. S. G., Pérez-Buitrago, N., 1307 Pérez-Mellado, V., Pianka, E. R., Pleguezuelos, J., Pollock, C., Ponce-Campos, 1308 P., Powell, R., Pupin, F., Quintero Díaz, G. E., Radder, R., Ramer, J., 1309 Rasmussen, A. R., Raxworthy, C., Reynolds, R., Richmana, N., Rico, E. L., 1310 Riservato, E., Rivas, G., Rocha, P. L.B.,Rödel, M-O., Schettino, L. R., 1311 Roosenburg, W. M, Ross, J. P.,Sadek, R., Sanders, K., Santos-Barrera, G., 1312 Schleich, H. H., Schmidt, B. R., Schmitz, A., Sharifi, M., Shea, G., Shi, H-T., 1313 Shine, R., Sindaco, R., Slimani, T., Somaweera, R., Spawls, S., Stafford, P., 1314 Stuebing, R., Sweet, S., Sy, E., Temple, H. J., Tognellic, M. F., Tolley, K., 1315 Tolson, P. J., Tuniyev,For B.,Review Tuniyev, S., Üzüm, Only N., Buurt, G., Van Sluys, M., 1316 Velasco, A., Vences, M., Veselý, M., Vinke, S., Vinke, T., Vogel, G., Vogrin, 1317 M., Vogt, R.C., Wear, O. R., Werner, Y. L.,Whiting, M., Wiewandt, T., 1318 Wilkinson, J., Wilson, B., Wrenb, S.,Zamin, T., Zhouhu, K., & Zug, G. (2013). 1319 The conservation status of the world’s reptiles. Biological Conservation, 157, 1320 372–385. doi:10.1016/j.biocon.2012.07.015 1321 Camper, J. D., & Zart, D. J. (2014) Atractus snethlageae (ground snake): diet. 1322 Herpetological Review, 45, 705. 1323 Carrillo de Espinoza, N., & Icochea, J. (1995). Lista taxonómica preliminar de los 1324 reptiles vivientes del Perú. Publicaciones del Museo Historia Natural 1325 Universidad Nacioanal Mayor San Marcos, 47, 1–27. 1326 Ceballos, G., Ehrlich, P. R., Barnosky, A. D., García, A., Pringle, R. M., & Palmer, T. 1327 M. (2015). Accelerated modern human–induced species losses: Entering the 1328 sixth mass extinction. Science Advances, 1: e1400253. 1329 doi:10.1126/sciadv.1400253 1330 Cunha, O. R., & Nascimento, F. P. (1978). Ofídios da Amazônia X – As cobras da 1331 região leste do Pará. Publicações Avulsas do Museu Paraense Emílio Goeldi, 31, 1332 1–218. 1333 Cunha, O. R., & Nascimento, F. P. (1983). Ofídios da Amazônia XX – As espécies de 1334 Atractus Wagler, 1828, na Amazônia oriental e Maranhão (Ophidia, 1335 Colubridae). Boletim do Museu Paraense Emílio Goeldi Nova Série Zoologia, 1336 123, 1–38. Page 41 of 81 Journal of Zoological Systematics and Evolutionary Research

1337 Cunha, O. R., & Nascimento, F. P. (1984). Ofídios da Amazônia XXI – Atractus zidoki 1338 no leste do Pará e notas sobre A. alphonsehogei e A. schach. Boletim do Museu 1339 Paraense Emílio Goeldi Nova Série Zoologia, 1, 219–228. 1340 Cunha, O. R., & Nascimento, F. P. (1993). Ofídios da Amazônia X - As cobras da 1341 região Leste do Pará. Boletim do Museu Paraense Emilio Goeldi Nova Série 1342 Zoologia, 9, 1–191. 1343 de Queiroz, K. (2007). Species concepts and species delimitation. Systematic Biology, 1344 56, 879–886. doi:10.1080/10635150701701083 1345 Dixon, J. R., & Soini, P. (1986). The reptiles of the upper Amazon basin, Iquitos region, 1346 Peru. Milwaukee: Milwaukee Public Museum. 1347 Doan, T. M., & Arizabal,For W. (2000). Review New herpetological Only records for the Tambopata 1348 Province, Department of Madre de Dios, Peru. Herpetological Review, 31, 188– 1349 189. 1350 Doan, T. M., & Arizabal, W. (2002). Microgeographic variation in species composition 1351 of the herpetofaunal communities of Tambopata region, Peru. Biotropica, 34, 1352 101–117. doi:10.1111/j.1744-7429.2002.tb00246.x 1353 Dowling, H. G. (1951). A proposed standard system of counting ventrals in snakes. 1354 British Journal of Herpetology, 1, 97–99. 1355 Dowling, H. G., & Savage, J. M. (1960). A guide to snake hemipenis: a survey of basic 1356 structure and systematic characteristics. Zoologica, 45, 17–28. 1357 Duellman, W. E. (1978). The biology of an equatorial herpetofauna in Amazonian 1358 Ecuador. Miscellaneous Publication University of Kansas Museum of Natural 1359 History, 65, 1–352. 1360 Duellman, W. E. (2005). Cusco Amazónico: the lives of amphibians and reptiles in an 1361 Amazonian rainforest. Ithaca: Cornell University Press. 1362 Duellman, W. E., & Salas, A. W. (1991). Annotated checklist of the amphibians and 1363 reptiles of Cuzco Amazónico, Peru. Occasional Papers of the Museum of 1364 Natural History University of Kansas, 143, 1–13. 1365 Duméril, A. M. C., Bibron, G., & Duméril, A. H. A. (1854). Erpétologie générale ou 1366 histoire naturelle complète des Reptiles. Tome 7. Paris: Librairie 1367 Encyclopédique de Roret. 1368 Drummond, A. J., Suchard, M. A., Xie, D., & Rambaut, A. (2012). Bayesian 1369 phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and 1370 Evolution, 29, 1969–1973. doi:10.1093/molbev/mss075 Journal of Zoological Systematics and Evolutionary Research Page 42 of 81

1371 Ferrer-Paris, J. R., Zager, I., Keith, D. A., Oliveira-Miranda, M. A., Rodríguez, J. P., 1372 Josse, C., González-Gil, M., Miller, R. M., Zambrana-Torrelio, C., & Barrow, E. 1373 (2019). An ecosystem risk assessment of temperate and tropical forests of the 1374 Americas with an outlook on future conservation strategies. Conservation 1375 Letters, 12: e12623. doi:10.1111/conl.12623 1376 Figueroa, A., McKelvy, A. D., Grismer, L. L., Bell, C. D., Lailvaux, S. P. (2016). A 1377 species-level phylogeny of extant snakes with description of a new colubrid 1378 subfamily and genus. Plos One, 11: e0161070. 1379 doi:10.1371/journal.pone.0161070 1380 Fraga, R., Lima, A. P., Prudente, A. L. C., & Magnusson, W. E. (2013). Guia de cobras 1381 da região de ManausFor – ReviewAmazônia central . OnlyManaus: INPA. 1382 Frota, J. G. (2000). Geographic distribution: Serpentes: Atractus snethlageae. 1383 Herpetological Review, 31, 254. 1384 Frota, J. G. (2004). As serpentes da região de Itaituba, médio rio Tapajós, Pará, Brasil 1385 (Squamata). Comunicações do Museu de Ciências e Tecnologia da Pontifícia 1386 Universidade Católica do Rio Grande do Sul – Série Zoologia, 17, 1, 9–19. 1387 Giraudo, A. R. (2004). Serpientes de la selva paranaense y del Chaco húmedo. Buenos 1388 Aires: Lola. 1389 Giraudo, A. R., & Scrocchi, G. J. (2000). The genus Atractus (Serpentes: Colubridae) in 1390 northeastern Argentina. Herpetological Journal, 10, 81–90. 1391 Gonzales, L., & Embert, D. (2008). Geographic distribution: Serpentes: Atractus 1392 snethlageae. Herpetological Review, 39, 370. 1393 Grazziotin, F. G., Zaher, H., Murphy, R. W., Scrocchi, G., Benavides, M. A., Zhang, 1394 Y., & Bonatto, S. L. (2012). Molecular phylogeny of the New World Dipsadidae 1395 (Serpentes: Colubroidea): a reappraisal. Cladistics, 28, 1–23. 1396 doi:10.1111/j.1096-0031.2012.00393.x 1397 Gutsche, A., Kwet, A., Kucharzewski, C., Lingnau, R., & Günther, R. (2007). Wilhelm 1398 Ehrhardt and an evaluation of his amphibians and reptiles held in the 1399 Herpetological Collection of the Museum für Naturkunde, Berlin. Mitteilungen 1400 aus dem Museum für Naturkunde in Berlin Zoologische Reihe, 83, 80–93. 1401 doi:10.1002/mmnz.200600019 1402 Hoogmoed, M. S. (1980). Revision of the genus Atractus in Surinam, with the 1403 ressurection of two species (Colubridae, Reptilia). Notes on the Herpetofauna of 1404 Surinam VII. Zoologische Verhandelingen, 175, 1–47. Page 43 of 81 Journal of Zoological Systematics and Evolutionary Research

1405 Hoogmoed, M. S. (1982). Snakes of Guianan region. Memórias do Instituto Butantã, 1406 46, 219–254. 1407 Hoogmoed, M. S, & Prudente, A. L. C. (2003). A new species of Atractus (Reptilia: 1408 Ophidia: Colubridae: Dipsadinae) from the Amazon forest region in Brazil. 1409 Zoologische Mededelingen, 77, 15-36, 425–439. 1410 Jorge-da-Silva, N. (1993). The snakes from Samuel hydroelectric power plant and 1411 vicinity, Rondônia, Brasil. Herpetological Natural History, 1, 37–86. 1412 Köhler, G. (2012). Color catalogue for field biologists. Affenbach: Herpeton. 1413 Kumar, S., Stecher, G., & Tamura, K. (2016). MEGA7: Molecular Evolutionary 1414 Genetics Analysis version 7.0 for bigger datasets. Molecular Biology and 1415 Evolution, 33, For1870–1874. Review doi:10.1093/molbev/msw054 Only 1416 Lanfear, R., Frandsen, P. B., Wright, A. M., Senfeld, T., & Calcott, B. (2016). 1417 PartitionFinder 2: new methods for selecting partitioned models of evolution for 1418 molecular and morphological phylogenetic analyses. Molecular Biology and 1419 Evolution, 34, 772–773. doi:10.1093/molbev/msw260 1420 Marçal, A. S., Gomes, I. B. R., & Coragem, J. T. (Orgs.). (2011). UHE Santo Antônio – 1421 Guia das espécies de fauna resgatadas. Porto Velho: Scriba Comunicação 1422 Corporativa. 1423 Maynard, R., Lynch, R. L., Maier, P., & Hamilton, P. S. (2017). Reptiles of San José de 1424 Payamino, Orellana, Ecuador. Chicago, IL: Field Museum. 1425 Melo-Sampaio, P. R., Oliveira, R. M., & Prates, I. (2018). A new nurse frog from Brazil 1426 (Aromobatidae: Allobates), with data on the distribution and phenotypic 1427 variation of Western Amazonian species. South American Journal of 1428 Herpetology, 13, 131–149. doi:10.2994/SAJH-D-17-00098.1 1429 Melo-Sampaio, P. R., Passos, P., Fouquet, A., Prudente, A. L. C., & Torres-Carvajal, O. 1430 (2019). Systematic review of Atractus schach (Serpentes: Dipsadidae) species 1431 complex from the Guiana Shield with description of three new species. 1432 Systematics and Biodiversity, 17, 207–229. 1433 doi:10.1080/14772000.2019.1611674 1434 Melo-Sampaio, P. R., Prates, I., Peloso, P. L. V., Recoder, R., Dal Vechio, F., Marques- 1435 Souza, S., & Rodrigues, M. T. (2020). A new nurse frog from southwestern 1436 Amazonian highlands, with notes on the phylogenetic affinities of Allobates 1437 alessandroi (Aromobatidae). Journal of Natural History, 54, 43–62. 1438 doi:10.1080/00222933.2020.1727972 Journal of Zoological Systematics and Evolutionary Research Page 44 of 81

1439 Meneses-Pelayo, E., & Passos, P. (2019). New polychromatic species of Atractus 1440 (Serpentes: Dipsadidae) from the Eastern portion of the Colombian Andes. 1441 Copeia, 2019, 250–261. doi:10.1643/CH-18-163 1442 Montagner, D. (2007). Construção da etnia Náwa. Revista de Estudos e Pesquisas, 4, 1443 33–108. 1444 Montingelli, G. G., Grazziotin, F. G., Battilana, J., Murphy, R. W., Zhang, Y. P., & 1445 Zaher, H. (2019). Higher-level phylogenetic affinities of the Neotropical genus 1446 Mastigodryas Amaral, 1934 (Serpentes: Colubridae), species-group definition 1447 and description of a new genus for Mastigodryas bifossatus. Journal of 1448 Zoological Systematics and Evolutionary Research, 57, 205–239. 1449 doi:10.1111/jzs.12262For Review Only 1450 Morato, S. A. A., Calixto, P. O., Mendes, L. R., Gomes, R., Galatti, U., Trein, F. L., 1451 Oliveira, F. S., & Ferreira, G. N. (2014). Guia fotográfico de identificação da 1452 herpetofauna da Floresta Nacional de Saracá-Taquera, Estado do Pará. 1453 Curitiba: STCP Engenharia de Projetos Ltda. 1454 Nascimento, F. P., Ávila-Pires, T. C., & Cunha, O. R. (1988). Répteis Squamata de 1455 Rondônia e Mato Grosso coletados através do Programa Pólo Noroeste. Boletim 1456 do Museu Paraense Emílio Goeldi, 4, 21–66. 1457 Nogueira, C.C., Argôlo, A.J.S., Arzamendia, V., Azevedo J.A., Barbo, F.E., Bérnils, 1458 R.S., Bolochio, B. E., Borges-Martins, M., Brasil-Godinho, M., Braz, H., 1459 Buononato, M. A., Cisneros-Heredia, D. F., Colli, G. R., Costa, H. C., Franco, F. 1460 L., Giraudo, A., Gonzalez, R. C., Guedes, T., Hoogmoed, M. S., Marques, O. A. 1461 V., Montingelli, G. G., Passos, P., Prudente, A. L.C., Rivas G. A., Sanchez P. 1462 M., Serrano F. C., Silva Jr. N. J., Strüssmann, C., Vieira-Alencar, J. P. S., 1463 Zaher, H., Sawaya, R. J., & Martins, M. C. M. (2019). Atlas of Brazilian snakes: 1464 verified point-locality maps to mitigate the Wallacean shortfall in a megadiverse 1465 snake fauna. South American Journal of Herpetology, 14, 1–274. 1466 doi:10.2994/SAJH-D-19-00120.1 1467 Passos, P., Cisneros-Heredia, D. F., Rivera, D. E., Aguilar, C., & Schargel, W. (2012). 1468 Rediscovery of Atractus microrhynchus and reappraisal of the taxonomic status 1469 of A. emersoni and A. natans (Serpentes: Dipsadidae). Herpetologica, 68, 375– 1470 392. doi:10.1655/HERPETOLOGICA-D-11-00078.1 Page 45 of 81 Journal of Zoological Systematics and Evolutionary Research

1471 Passos, P., Chiesse, A., Torres-Carvajal, O., & Savage, J. (2010). Testing species 1472 boundaries within Atractus occipitoalbus complex (Serpentes: Dipsadidae). 1473 Herpetologica, 65, 384-403. doi:10.1655/08-024.1 1474 Passos, P., & Fernandes, R. (2008). A new species of the colubrid snake genus Atractus 1475 (Reptilia: Serpentes) from the central Amazon of Brazil. Zootaxa, 1849, 59–66. 1476 doi:10.11646/zootaxa.1849.1.4 1477 Passos, P., Fernandes, R., Bérnils, R. S., & Moura-Leite, J. C. (2010c). Taxonomic 1478 revision of Atlantic Forest Atractus (Serpentes: Dipsadidae). Zootaxa, 2364, 1– 1479 63. doi:10.11646/zootaxa.2364.1.1 1480 Passos, P., Fernandes, D. S., & Borges-Nojosa, D. M. (2007b). A new species of 1481 Atractus (Serpentes:For Dipsadinae) Review from a relictual Only forest in northeastern Brazil. 1482 Copeia, 2007, 788–797. doi:10.1643/0045- 1483 8511(2007)7[788:ANSOAS]2.0.CO;2 1484 Passos, P., Kok, P. J. R., Albuquerque, N. R., & Rivas, G. (2013). Groundsnakes of the 1485 Lost World: a review of Atractus (Serpentes: Dipsadidae) from the Pantepui 1486 region, northern South America. Herpetological Monographs, 27, 52–86. 1487 doi:10.1655/HERPMONOGRAPHS-D-12-00001R2.1 1488 Passos, P., Prudente, A. L. C., & Lynch, J. D. (2016). Redescription of Atractus 1489 punctiventris and description of two new Atractus (Serpentes: Dipsadidae) from 1490 Brazilian Amazonia. Herpetological Monographs, 30, 1–20. 1491 doi:10.1655/HERPMONOGRAPHS-D-14-00009 1492 Passos, P., Prudente, A. L. C., Ramos, L. O., Caicedo-Portilla, J. R., & Lynch, J. D. 1493 (2018). Species delimitations in the Atractus collaris complex (Serpentes: 1494 Dipsadidae). Zootaxa, 4392, 491–520. doi:10.11646/zootaxa.4392.3.4 1495 Passos, P., Ramos, L. O., Fouquet, A., & Prudente, A. L. C. (2017). Taxonomy, 1496 morphology and distribution of Atractus flammigerus Boie, 1827 (Serpentes: 1497 Dipsadidae). Herpetologica, 73, 349–363. doi:10.1655/Herpetologica-D-16- 1498 00086 1499 Passos, P., Scanferla, A., Melo-Sampaio, P. R., Brito, J., & Almendáriz, A. (2019). A 1500 giant on the ground: another large-bodied Atractus (Serpentes: Dipsadinae) from 1501 Ecuadorian Andes, with comments on the dietary specializations of the goo- 1502 eaters snakes. Anais da Academia Brasileira de Ciências, 91, e20170976. 1503 doi:10.1590/0001-3765201820170976 Journal of Zoological Systematics and Evolutionary Research Page 46 of 81

1504 Pesantes, O. (1994). A method for preparing hemipenis of preserved snakes. Journal of 1505 Herpetology, 28, 93–95. 1506 Peters, J. A. (1964). Dictionary of Herpetology. New York: Hafner. 1507 Pérez-Santos, C. E., & Moreno, A. G. (1988). Ofidios de Colombia. Museo Regionale 1508 di Scienzi Naturali Torino, 6, 6–517. 1509 Pimm, S. L., Jenkins, C. N., Abell, R., Brooks, T. M., Gittleman, J. L., Joppa, L. N., 1510 Raven, P. H., Roberts, C. M., & Sexton, J. O. (2014). The biodiversity of species 1511 and their rates of extinction, distribution, and protection. Science, 344, 1246752- 1512 1–10. doi:10.1126/science.1246752 1513 Prudente, A. L. C., Maschio, G. F., Santos-Costa, M. C., & Feitosa, D. T. (2010). 1514 Serpentes da baciaFor petrolífera Review de Urucu, Município Only de Coari, Amazonas, Brasil. 1515 Acta Amazônica, 40, 381–386. doi:10.1590/S0044-59672010000200016 1516 Prudente, A. L. C., & Passos, P. (2008). A new species of Atractus Wagler, 1828 1517 (Serpentes: Dipsadinae) from Guyana plateau in Northern Brazil. Journal of 1518 Herpetology, 42, 723–732. doi:10.1670/07-115R3.1 1519 Prudente, A. L. C., & Passos, P. (2010). New cryptic species of Atractus (Serpentes: 1520 Dipsadidae) from Brazilian Amazonia. Copeia, 2010, 397–404. 1521 doi:10.1643/CH-08-193 1522 Prudente, A. L. C., & Santos-Costa, M. C. (2005). Checklist of snakes from Ferreira 1523 Penna Scientific Station, Eastern Amazonia, Pará State, Brazil. Boletim do 1524 Museu Paraense Emílio Goeldi – Série Ciências Naturais, 1, 3, 203–208. 1525 Prudente, A. L. C., Sarmento, J. F. M., Ávila-Pires, T. C. S., Maschio, G. F., & Sturaro, 1526 M. J. (2018). How much do we know about the diversity of Squamata (Reptilia) 1527 in the most degraded region of Amazonia?. South American Journal of 1528 Herpetology, 13, 117–130. doi:10.2994/SAJH-D-17-00009.1 1529 Pyron, R. A., Arteaga, A., Echevarría, L. Y., & Torres-Carvajal, O. (2016). A revision 1530 and key for the tribe Diaphorolepidini (Serpentes: Dipsadidae) and checklist for 1531 the genus Synophis. Zootaxa, 4171, 293–320. doi:10.11646/zootaxa.4171.2.4 1532 Pyron, R. A., Burbrink, F. T., & Wiens, J. J. (2013). A phylogeny and revised 1533 classification of Squamata, including 4161 species of lizards and snakes. 1534 BioMedCentral Evolutionary Biology, 13, 1–53. doi:10.1186/1471-2148-13-93 1535 Pyron, R. A., Guayasamin, J. M., Peñafiel, N., Bustamante, L., & Arteaga, A. (2015). 1536 Systematics of Nothopsini (Serpentes, Dipsadidae), with a new species of Page 47 of 81 Journal of Zoological Systematics and Evolutionary Research

1537 Synophis from the Pacific Andean slopes of southwestern Ecuador. ZooKeys, 1538 541, 109–147. doi:10.3897/zookeys.541.6058 1539 Rodrigues, G. M., Maschio, G. F., & Prudente, A. L. C. (2016). Snake assemblages of 1540 Marajó Island, Pará state, Brazil. Zoologia, 33, e20150020. doi:10.1590/S1984- 1541 4689zool-20150020 1542 Rodríguez, L. O., & Knell, G. (2004). Amphibians and Reptiles. In: N. Pitman, R.C. 1543 Smith, C. Vriesendorp, D. Moskovits, R. Piana, G. Knell, & T. Wachter, (Eds.), 1544 Perú: Ampiyacu, Apayacu, Yaguas, Medio Putumayo. Rapid Biological 1545 Inventories Report 12 (pp. 152–155). Chicago, IL: The Field Museum. 1546 Rojas, R. R., Fouquet, A., Ron, S. R., Hernández-Ruz, E. J., Melo-Sampaio, P. R., 1547 Chaparro, J. C.,For Vogt, R.Review C., Carvalho, V. OnlyT., Pinheiro, L. C, Ávila, R. W., 1548 Farias, I. P., Gordo, M., & Hrbek, T. (2018). A Pan-Amazonian species 1549 delimitation: high species diversity within the genus Amazophrynella (Anura: 1550 Bufonidae). PeerJ, 6, e4941. doi:10.7717/peerj.4941 1551 Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D. L., Darling, A., Höhna, S, 1552 Larget, B., Liu, L., Suchard, M. A., & Huelsenbeck, J. P. (2012). MrBayes 3.2: 1553 Efficient Bayesian phylogenetic inference and model choice across a large 1554 model space. Systematic Biology, 61, 539–542. doi:10.1093/sysbio/sys029 1555 Sabaj, M. (2020). Codes for natural history collections in ichthyology and herpetology. 1556 Copeia, 108, 3, 593–669. doi:10.1643/ASIHCODONS2020 1557 Savage, J. M. (1960). A revision of the Ecuadorian snakes of the colubrid genus 1558 Atractus. Miscellaneous Publications of the Museum of Zoology University of 1559 Michigan, 112, 1–86. 1560 Schargel, W. E., Lamar, W. W., Passos, P., Valencia, J. H., Cisneros-Heredia, D. F., & 1561 Campbell, J. A. (2013). A new giant Atractus (Serpentes: Dipsadidae) from 1562 Ecuador, with notes on some other large Amazonian congeners. Zootaxa, 3721, 1563 455–474. doi:10.11646/zootaxa.3721.5.2 1564 Scheffers, B. R., Oliveira, B. F., Lamb, I., & Edwards, D. P. (2019). Global wildlife 1565 trade across the tree of life. Science, 366, 71–76. doi:10.1126/science.aav5327 1566 Serié, P. (1915). Suplemento a la fauna erpetologica argentina. Annales del Museo 1567 Nacional de Historia Natural Buenos Aires, 27, 93–109. 1568 Silva, J. M. C., Rylands, A. B., & Fonseca, G. A. B. (2005). The fate of the Amazonian 1569 areas of endemism. Conservation Biology, 19, 689–694. doi:10.1111/j.1523- 1570 1739.2005.00705.x Journal of Zoological Systematics and Evolutionary Research Page 48 of 81

1571 Silva, M. V., Souza, M. B., & Bernarde, P. S. (2012) “2010”. Riqueza e dieta de 1572 serpentes do Estado do Acre, Brasil. Revista Brasileira de Zoociências, 12, 165– 1573 176. 1574 Silva-Haad, J. J. (2004). Las serpientes del género Atractus Wagler, 1828 (Colubridae; 1575 Xenodontinae) en la Amazonia colombiana. Revista de la Academia 1576 Colombiana de Ciencias Exactas Físicas y Naturales, 108, 409–446. 1577 Sonter, L. J., Herrera, D., Barrett, D. J., Galford, G. L., Moran, C. J., & Soares-Filho, B. 1578 S. (2017). Mining drives extensive deforestation in the Brazilian Amazon. 1579 Nature Communications, 8, 1013. doi:10.1038/s41467-017-00557-w 1580 Staal, A., Flores, B. M., Aguiar, A. P. D., Bosmans, J. H. C., Fetzer, I., & Tuinenburg. 1581 O. A. (2020). FeedbackFor Reviewbetween drought andOnly deforestation in the Amazon. 1582 Environmental Research Letters, 15, 044024. doi:10.1088/1748-9326/ab738e 1583 Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis and post- 1584 analysis of large phylogenies. Bioinformatics, 30, 1312–1313. 1585 doi:10.1093/bioinformatics/btu033 1586 Stamatakis, A., Hoover, P., & Rougemont, J. (2008). A rapid bootstrap algorithm for 1587 the RAxML Web servers. Systematic Biology, 57, 758–771. 1588 doi:10.1080/10635150802429642 1589 Sullivan, J., & Joyce, P. (2005). Model selection in phylogenetics. Annual Review of 1590 Ecology, Evolution and Systematics, 36, 445–466. 1591 doi:10.1146/annurev.ecolsys.36.102003.152633 1592 Thompson, J. D., Higgins, D. G., & Gibson, T. J. (1994). CLUSTAL W: improving the 1593 sensitivity of progressive multiple sequence alignment through sequence 1594 weighting, position-specific gap penalties and weight matrix choice. Nucleic 1595 Acids Research, 22, 4673–4680. doi:10.1007/978-1-4020-6754-9 1596 Torres-Carvajal, O., & Hinojosa, K. C. (2020). Hidden diversity in two widespread 1597 snake species (Serpentes: Xenodontini: Erythrolamprus) from South America. 1598 Molecular Phylogenetics and Evolution 146, 106772. 1599 doi:10.1016/j.ympev.2020.106772 1600 Uetz, P., Freed, P., & Hošek, J. (eds.) (2020). The Reptile Database, http://www.reptile- 1601 database.org, accessed 23 July 2020. 1602 Vaidya, G., Lohman, D. J., & Meier, R. (2011). SequenceMatrix: concatenation 1603 software for the fast assembly of multi-gene datasets with character set and Page 49 of 81 Journal of Zoological Systematics and Evolutionary Research

1604 codon information. Cladistics, 27, 171–180. doi:10.1111/j.1096- 1605 0031.2010.00329.x 1606 Vanzolini, P. E. (1986a). Addenda and corrigenda to Part I Snakes. In: J. A. Peters & B. 1607 Orejas-Miranda. Catalogue of the Neotropical Squamata. Part I, Snakes. 1608 Washington: Smithsonian Institution. 1609 Vanzolini, P. E. (1986b). Levantamento herpetológico da área do estado de Rondônia 1610 sob influência da rodovia BR 364. Relatório de Pesquisa n.1. Brasília: 1611 MCT/CNPq. 1612 Vaz-Silva, W., Oliveira, R. M., Gonzaga, A. F. N., Pinto, K. C., Poli, F. C., Bilce, T. 1613 M., Penhacek, M., Wronski, L., Martins, J. X., Junqueira, T. G., Cesca, L. C. C., 1614 Guimarães, V.For Y., & Pinheiro, Review R. D. (2015). Only Contributions to the knowledge of 1615 amphibians and reptiles from Volta Grande do Xingu, northern Brazil. Brazilian 1616 Journal of Biology, 75, 3, S205–S218. doi:10.1590/1519-6984.00814BM 1617 Vitt, L. J., Ávila-Pires, T. C. S., Caldwell, J. P., & Oliveira, V. R. L. (1997). The impact 1618 of individual tree harvesting on thermal environments of lizards in Amazonian 1619 rain forest. Conservation Biology, 12, 654–664. doi:10.1111/j.1523- 1620 1739.1998.96407.x 1621 von May, R., & Mueses-Cisneros, J. J. (2010). Amphibians and Reptiles. In: N. Pitman, 1622 C. Vriesendorp, D. K. Moskovits, R. von May, D. Alvira, T. Wachter, D. F. 1623 Stotz, & Á. del Campo (Eds.), Perú: Yaguas-Cotuhé. Rapid Biological 1624 Inventories Report 23 (pp. 230–237). Chicago, IL: The Field Museum. 1625 Waldez, F., Menin, M., & Vogt, R.C. (2013). Diversidade de anfíbios e répteis 1626 Squamata na região do baixo rio Purus, Amazônia Central, Brasil. Biota 1627 Neotropica, 13, 300–316. 1628 Wilson, E. O. (1988). The current state of biological diversity. In E. O. Wilson & F. M. 1629 Peter. Biodiversity (pp. 3–18). Washington: National Academy Press. 1630 Yanez-Muñoz, M., & Venegas, P. J. (2008). Amphibians and Reptiles. In: W. S. 1631 Alverson, C. Vriesendorp, Á. del Campo, D. K. Moskovits, D. F. Stotz, M. G. 1632 Donayre, & L. A. Borobr (Eds.), Ecuador-Peru: Cuyabeno-Güeppi, Rapid 1633 Biological Inventories Report 20 (pp. 215–221). Chicago, IL: The Field 1634 Museum. 1635 Zaher, H. (1999). Hemipenial morphology of the South American xenodontine snakes, 1636 with a proposal for a monophyletic Xenodontinae and a reappraisal of colubroid 1637 hemipenes. Bulletin of the American Museum of Natural History, 240, 1–168. Journal of Zoological Systematics and Evolutionary Research Page 50 of 81

1638 Zaher, H., Grazziotin, F.G., Cadle, J. E., Murphy, R. W., Moura-Leite, J.C., & Bonnato, 1639 S. L. (2009). Molecular phylogeny of advanced snakes (Serpentes, Caenophidia) 1640 with an emphasis on South American Xenodontines: a revised classification and 1641 descriptions of new taxa. Papéis Avulsos de Zoologia, 49, 115–153. 1642 doi:10.1590/S0031-10492009001100001 1643 Zemp, D. C., Schleussner, C.-F., Barbosa, H. M. J., & Rammig, A. (2017). 1644 Deforestation eﬀects on Amazon forest resilience. Geophysics Research Letters, 1645 44, 6182–6190. doi:10.1002/2017GL072955 1646 1647 1648 FIGURE LEGENDSFor Review Only 1649 1650 Figure 1. Phylogeny of the Atractus snethlageae complex. Maximum-likelihood tree 1651 based on phylogenetic analysis of six genes (16S, CMOS, CYTB, ND4, NT3 and 1652 RAG1). Posterior probabilities and Bootstrap values are separated by “/”. 1653 1654 Figure 2. Dorsal (A) and ventral (B) views of Atractus snethlageae. Holotype (MPEG 1655 10131) from Cachoeira do Piriá, Pará, Brazil. 1656 1657 Figure 3. Colour pattern variation of Atractus snethlageae. Paratypes: MPEG 3955 (A), 1658 MPEG 15973 (B), MPEG 6845 (C), MPEG 16383 (D), MPEG 2595 (E), and MPEG 1659 10137 (F). 1660 1661 Figure 4. Colour in life of Atractus snethlageae. Uncatalogued specimen from Porto 1662 Velho, Rondônia, Brazil (A, B). MPEG 20362 from Nova Mamoré, Rondônia, Brazil 1663 (C). MTR 18844 from Beruri, Amazonas, Brazil (D). Itaituba, Pará, Brazil (E). Cruzeiro 1664 do Sul, Acre, Brazil (H). Photos by D. Meneghelli (A), U. Oliveira (B), L. Vitt (C), M. 1665 Teixeira Jr. (D), L. Moraes (E), S. Albuquerque (F). 1666 1667 Figure 5. Hemipenial morphology of Atractus snethlageae. Asulcate (left) and sulcate 1668 (right) views of the organs of specimens from Colniza, Mato Grosso, Brazil (UFMT-R 1669 7813 - A); Itaituba, Pará, Brazil (MPEG 24564 - B); Machadinho d’Oeste, Rondônia 1670 (MZUSP 21982 - C); Anamã, Amazonas (INPA-H 9524 - D); Porto Velho, Rondônia 1671 (MPEG 17877 - E); and Belém, Pará (MPEG 10137, paratype - F). Scale bar = 5 mm. Page 51 of 81 Journal of Zoological Systematics and Evolutionary Research

1672 1673 Figure 6. Known distribution of Atractus snethlageae. Type-locality is represented by 1674 white dot. 1675 1676 Figure 7. Dorsal (A), lateral (B) and ventral (C) views of Atractus nawa. Holotype 1677 (MPEG 20376) from the municipality of Porto Walter, state of Acre, Brazil. Scale bar 5 1678 mm. 1679 1680 Figure 8. General view of the holotype of Atractus nawa (MPEG 20376) in life from 1681 Porto Walter, state of Acre, Brazil. Photo by L. Vitt. 1682 For Review Only 1683 Figure 9. Colour in life of Atractus nawa. Uncollected specimen from Santa Luiza 1684 Village, Cruzeiro do Sul, Brazil (A–D). Note the light orange snout. Photos by D. Paiva. 1685 1686 Figure 10. Known distribution of Atractus nawa. Type locality is represented by white 1687 dot. 1688 1689 Figure 11. Dorsal (left) and ventral (right) views. Paratypes of Atractus pachacamac 1690 CORBIDI 13797 (A–B) from Cahuapanas, Datem del Marañón, Peru; CORBIDI 167 1691 (C–D) from Güeppi, Maynas, Peru and QCAZ 8278 (E–F) from Villano, Pastaza, 1692 Ecuador. 1693 1694 Figure 12. Colour pattern variation of Atractus pachacamac. Holotype: QCAZ 12630 1695 (A–B) from Wildsumaco Wildlife Sanctuary, Napo, Ecuador; Paratypes: QCAZ 15414 1696 (C) from La Zarza, Zamora-Chinchipe, Ecuador; QCAZ 16083 (D–E) from type 1697 locality. Photos by Bioweb/QCAZ. 1698 1699 Figure 13. Hemipenial morphology of Atractus pachacamac. Asulcate (left) and sulcate 1700 (right) views of the organs of specimens from Ecuador and Peru: holotype QCAZ 1701 12630 (A), paratype QCAZ 10639 (B), paratype CORBIDI 167 (C), paratype CORBIDI 1702 13797 (D), paratype EPN 5471 (E), and paratype QCAZ 11833 (F). 1703 1704 Figure 14. Known distribution of Atractus pachacamac. Type locality is represented by 1705 white dot. Journal of Zoological Systematics and Evolutionary Research Page 52 of 81

1706 1707 Figure 15. Dorsal (A), lateral (B), and ventral (C) views. Holotype of Atractus akerios 1708 (MPEG 12255) from Nova Vida, Junco do Maranhão, state of Maranhão, Brazil. Scale 1709 bar = 5 mm. 1710 1711 Figure 16. Dorsal (left) and ventral (right) views of Atractus akerios. Holotype MPEG 1712 12255 (A) from Junco do Maranhão, state of Maranhão, Brazil, and paratype MPEG 1713 10106 (B) from same locality. 1714 1715 Figure 17. Known distribution of Atractus akerios in Brazil. The type locality is 1716 represented by white dot.For Review Only 1717 1718 Figure 18. Dorsal (A), lateral (B) and ventral (C) views. Holotype of Atractus ukupacha 1719 (QCAZ 12504) from El Chaco, province of Napo, Ecuador. Scale bar = 5 mm. 1720 1721 Figure 19. Colour in life of Atractus ukupacha. Uncollected specimen from San José de 1722 Payamino, Orellana, Ecuador (A-B). Photos by R. Maynard. 1723 1724 Figure 20. Hemipenial morphology. Asulcate (left) and sulcate (right) views of the 1725 organs of the holotype of Atractus akerios MPEG 12255 (A) from Junco do Maranhão, 1726 Maranhão, Brazil; and of the paratypes of Atractus ukupacha from El Reventador, 1727 Napo, Ecuador QCAZ 205 (B), QCAZ 4047 (C), QCAZ 3476 (D), QCAZ 10614 (E), 1728 QCAZ 4 (F). 1729 1730 Figure 21. Known distribution of Atractus ukupacha. Type locality is represented by 1731 white dot. Beige areas are above 1000 m asl. 1732

1733 SUPPLEMENTARY FIGURES

1734 Figure S1. Phylogeny of Atractus. Maximum-likelihood tree based on phylogenetic

1735 analysis of six genes (16S, CMOS, CYTB, ND4, NT3 and RAG1). Node labels

1736 correspond to bootstrap support values; labels on short branches were removed for

1737 clarity. The Atractus snethlageae complex is shown in red. Page 53 of 81 Journal of Zoological Systematics and Evolutionary Research

1738

1739 Figure S2. Phylogeny of Atractus. Maximum clade credibility tree based on

1740 phylogenetic analysis of six genes (16S, CMOS, CYTB, ND4, NT3 and RAG1). Node

1741 labels correspond to posterior probability values; labels on short branches were removed

1742 for clarity. The Atractus snethlageae complex is shown in red.

1743

1744 Figure S3. Head of the holotype of Atractus snethlageae. A- Dorsal view; B- Lateral 1745 view; C- Ventral view.For Review Only Journal of Zoological Systematics and Evolutionary Research Page 54 of 81

Table 1. Selected features for comparisons among Atractus species closely related to or previously associated (in literature or scientific collections) with

Atractus snethlageae species complex. Abbreviations are as follow: VE =ventrals; SC=subcaudals; SL=supralabials; IL(CH)=Infralabials (in contact with chinshields); MT=maxillary teeth; LPB=light parietal band. The SL, IL(CH), MT, and LPB were placed together for both sexes because these variables are not secondarily dimorphic. NA = not available.

SPECIES VE ForSC ReviewSL IL(CH) MT OnlyLPB Dorsal colouration

F M F M

A. akerios 149–156 140–145 19–20 27–33 7 8 (4) 6 Absent antique brown with russet spots

A. atlas 158–169 NA 28–33 NA 8 8 (4) 8 Absent yellow ocher with black bands

tawny oliver with raw umber bands / A. dapsilis 167–182 152–166 21–26 30–37 7 8 (4) 6–7 Absent grayish brown with cinnamon bands Page 55 of 81 Journal of Zoological Systematics and Evolutionary Research

cinnamon drab with cinnamon brown bands/ A. major 154–181 149–170 27–37 31–45 7 7(3) 6–7 Absent fawn colour with warm sepia bands

A. nawa 166–169 NA 16–20ForNA Review7 7 (4) 7 OnlyAbsent Brussels brown with Raw umber spots

A. pachacamac 162–175 158–167 31–33 39–45 7 8 (4) 7 Absent sepia with saval brown bands

A. schach 148–150 142–151 19–21 25–32 7 8 (4) 7 Absent olive brown with black regular bands

A. snethlageae 147–163 137–155 21–24 27–34 7 8 (4) 6 small/incomplete Vandyke brown with light yellow ocher bands

short pale crossbands A. touzeti 167–170 NA 31 NA 8 7 or 8 (4) 8 Absent edged by black borders Journal of Zoological Systematics and Evolutionary Research Page 56 of 81

A. trefauti 153–158 139–149 21–24 24–29 7 8 (4) 5–7 Incomplete black with beige bands

A. ukupacha 161–170 154–165 23–29 38–42 7 8 (4) 7 Absent dusky brown with olive-brown bands For Review Only Page 57 of 81 Journal of Zoological Systematics and Evolutionary Research

APPENDIX

Material Examined

Specimens marked with asterisk (*) were included in the phylogeny and specimens

highlighted in (bold) were used to prepare hemipenes. The specimens included in the

type series of Atractus akerios, A. nawa, A. pachacamac, A. snethlageae and A.

ukupacha are not listed again here. Institutional abbreviations follow Sabaj (2020).

Atractus aboiporu (n=3). Brazil: Amapá: Serra do Navio: (MPEG 25796, holotype; MPEG 25797,For paratype); Review Pedra Branca Onlydo Amapari: (MPEG 19783, paratype).

Atractus aff. snethlageae (n=1). Peru: Madre de Díos: Tambopata: CORBIDI

15135*).

Atractus atlas (n=4). Ecuador: Zamora-Chinchipe: Paquisha: (MEPN 14203,

holotype); Parroquía Guayzimi: (DHMECN 2972, paratype). Morona-Santiago: Reserva

Biológica Cerro Plateado: (QCAZ 14946*, paratype); Zúñac: (DHMECN 12361,

paratype).

Atractus boimirim (n=22). Brazil: Acre: Cruzeiro do Sul: (UFACF 112).

Rondônia: Machadinho d’Oeste (MZUSP 21871, 21983); Samuel Hydroelectric Power

Plant at Jamari river: Vila Cachoeira do Samuel: (MPEG 17908, holotype; MPEG

17909–11, 17916–17, 17922, 17967–70, paratypes), surroundings of Jaci-Paraná river,

tributary of Madeira river: Jirau Hydroelectric Power Plant: (MPEG 23965, paratype),

Porto Velho (UFRO-H 45). Pará: Altamira (MZUSP 21716); Juriti: (MNRJ 24864,

paratype); Parque Nacional da Amazônia: (MPEG 25259–60, paratypes), Itaituba:

(MPEG 21233*).

Atractus dapsilis (n= 55, including 39 paratypes). Brazil: Amazonas: Manaus:

(IBSP 49430, IMTM 1061, 1440, INPA-H 18466, 32348, 32271, MZUSP 3713, UFAM Journal of Zoological Systematics and Evolutionary Research Page 58 of 81

20.06.09); Presidente Figueredo: (IMTM 1354, 1378, 1501, 1563, 1678, MPEG 17426–

27, 17495, 17527, 17539, 17559, 17568, MZUSP 8659, 9501). Pará: Oriximiná: (IBSP

87633, MNRJ 14910, 14911, 14912, 14913, 14914 [holotype], 14915, 16794*, 16795,

16796*, 16797–801, 16802*, 16804–03, MPEG 20782, 21569–70, 21712–13, 23505,

23759–60); Terra Santa: (INPA-H 31489, MNRJ 17953–54).

Atractus flammigerus (n=3). Brazil: Pará: Almeirim (MPEG 21011, 21013,

21353). Atractus gigasFor (n=17). ReviewEcuador: Carchi: kmOnly 15 El Chical–Gualtal road, Tulcán Chical: (QCAZ 5771*). Cotopaxi: Bosque Protector Río Guajalito (formerly Palmeras

Farm), between San Francisco de Las Pampas and Quito: (FHGO 194, holotype; QCAZ

2099, topotype); Bosque Integral Otonga: (QCAZ 3266), San Francisco de Las Pampas:

(QCAZ 175, 179, 443, 647, 662). Pichincha: Cantón San Miguel de los Bancos,

Tadayapa road, Tandayapa Farm: (FHGO 4791), Chiriboga: (QCAZ 01), Reserva Las

Gralarias: (MZUTI 3286*), Las Palmas: Lloa: (DHMECN 372), Palmeras: (QCAZ

2099), Peñas Coloradas: (QCAZ 4058), Reserva Bella Vista: (QCAZ 6526).

Provenance in error: Piso Tropical Oriental: without specific data: (MEPN 8706). Peru:

Cajamarca: San Ignacio, Santuario Nacional Tabaconas Namballe: Alto Lhuama:

(CORBIDI 877), El Chaupe: (ZFMK 89147).

Atractus pachacamac (n=7). Colombia: Amazonas: Puerto Nariño (IAVH

3872); Caquetá: Florencia: (MLS 2731). Ecuador: Morona-Santiago: Macas: (EPN

8711). Sucumbios: Lago Agrio: (MECN 7873). Napo: (QCAZ 5074). Peru: Putumayo:

Maynas: (MUSM 32287); Andoas: (MUSM 27356). Cajamarca: Jaén: (MUSM 3390).

Atractus schach (n=12). French Guiana: no specific locality: (MNHN

1995.9481, AMNH 139922); Saul: Limonade (AF1716*); Saint-Eugène: (MNHN

1997.2371); Nourágues: (MNHN 2002.615). Guiana: no specific locality (NHM Page 59 of 81 Journal of Zoological Systematics and Evolutionary Research

1939.1.1.95, NHM 1939.1.1.96). Suriname: no specific locality (NHM 1870.3.10.61);

Browns Bergen: (RMNH 12683); Petit Saut: (RMNH 38072); Sinnamary river: (SMNS

2664).

Atractus snethlageae (n=116). Brazil: no specific locality: (MNRJ 9842*).

Acre: Marechal Thaumaturgo: (UFAC-RB 227); Cruzeiro do Sul: (UFACF 475).

Amazonas: Anamã: (INPA-H 9524); Benjamin Constant: (IBSP 33369); Beruri: (INPA-

H 13974–75); Manaus [probably in error] (BMNH 1876.4.23.1). Mato Grosso: Colniza: (UFMT-R 7813, 7820).For Maranhão: Review Açailândia: (MNRJ Only 16507); Estreito: (MZUSP 18694). Pará: Altamira: (MZUSP 9334, MPEG 23462); Ananindeua: (MPEG 18518);

Barcarena (MPEG 23451); Chaves: (MPEG 24960); Itaituba: (MPEG 21138, 24564,

25149–50); Melgaço: (MPEG 18653, 19863–64, 19966, 19968, 20071, 20605, 21581–

83, 23167–68, 23134, 23136); Santarém: (MPEG 19092, 19098); Porto de Moz:

(MZUSP 18654–55); Capitão Poço: (MPEG 13267); Vitória do Xingu: (MPEG 19807,

19811, 26191, 26194, 26198–99, 26206, 26208, 26223, 26112, 25834, 25836, 26195,

26203, MNRJ 25233–34, 25239, 24383–85, MZUSP 21769). Rondônia: Ariquemes:

(IBSP 41530); Candeias do Jamari: (CHUFC 1399, MPEG 17805–06, 17875–76,

17877–78, 17913–14, 17918–19); Espigão do Oeste: (MPEG 21058); Itapuã do Oeste:

(MZUSP 8680); Nova Mamoré: (MPEG 20362); Porto Velho: (INPA-H 27840, 32102,

32372, 34150, 34173, MPEG 23966, 25494–96, MZUSP 19012, 19029–30, 19050–55,

19745, 19985, 20673, UFRO-H 253, 1726–27, 1730, 1732, 1747, 1754, 1757, 2545–

2554, 3034–3035, 3725, 3727); Machadinho d’Oeste: (MZUSP 21464–65, 21474,

21862, 21981, 21982); São Miguel do Guaporé: (MPEG 16554). Peru: Cusco: La

Convención: (CORBIDI 10034, 15991, MHNC 48, 182, 377, 422, 424, 1029, MUSM

30847); Madre de Díos: Manu: (MUSM 34987); Tambopata: (CORBIDI 18264). Journal of Zoological Systematics and Evolutionary Research Page 60 of 81

Loreto: Loreto: (IBSP 49431–32); San Martin: Tarapoto: (MUSM 3338). Bolivia:

Pando: (MNKR 3275). Colombia: Caquetá: (MLS 140); (ICN 10109–10).

Atractus tartarus (n=3). Brazil: Pará: (MPEG 25849, MPEG 26456, MPEG

26457).

Atractus torquatus (n=4). Brazil: Roraima (INPA-H 28565, 28567, 28569,

MZUSP 14281). French Guiana: (AF2281).

Atractus touzeti (n=3). Ecuador: Napo: Cordillera de los Guacamayos: Cosanga–Archidona road:For (FHGO Review 517, holotype); Only La Virgen: (FHGO 2035–36, paratypes).

Atractus trefauti (n=6). French Guiana: Roura: (MNRJ 26709*, holotype);

Mont Tabulaire: (MNHN 2015.56, paratype). Brazil: Amapá: Serra do Navio: (MPEG

25788, MPEG 16382, paratypes); Pará: Almeirim: (MPEG 21354*–55, paratypes). Page 61 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

Figure 1. Phylogeny of the Atractus snethlageae complex. Maximum-likelihood tree based on phylogenetic analysis of six genes (16S, CMOS, CYTB, ND4, NT3 and RAG1). Posterior probabilities and Bootstrap values are separated by “/”. Journal of Zoological Systematics and Evolutionary Research Page 62 of 81

For Review Only

209x85mm (300 x 300 DPI) Page 63 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x297mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 64 of 81

For Review Only

209x205mm (300 x 300 DPI) Page 65 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x297mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 66 of 81

For Review Only

1237x874mm (72 x 72 DPI) Page 67 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x111mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 68 of 81

For Review Only

76x57mm (300 x 300 DPI) Page 69 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

207x141mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 70 of 81

For Review Only

1237x874mm (72 x 72 DPI) Page 71 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x230mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 72 of 81

For Review Only

209x190mm (300 x 300 DPI) Page 73 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x297mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 74 of 81

For Review Only

1237x874mm (72 x 72 DPI) Page 75 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

209x97mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 76 of 81

For Review Only

209x297mm (300 x 300 DPI) Page 77 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

1237x874mm (72 x 72 DPI) Journal of Zoological Systematics and Evolutionary Research Page 78 of 81

For Review Only

209x105mm (300 x 300 DPI) Page 79 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only 209x71mm (300 x 300 DPI) Journal of Zoological Systematics and Evolutionary Research Page 80 of 81

For Review Only

209x297mm (300 x 300 DPI) Page 81 of 81 Journal of Zoological Systematics and Evolutionary Research

For Review Only

1237x874mm (72 x 72 DPI)