﻿A new species of the genus Liotyphlops Peters, 1881 (Serpentes, Anomalepididae) from Colombia and the synonymization of Liotyphlopsbeui (Amaral, 1924) with Liotyphlopsternetzii (Boulenger, 1896)

﻿Abstract A new species of Liotyphlops Peters, 1881, Liotyphlopspalauophissp. nov., is described from the neighborhoods of Bogota, Colombia from a previous syntype of L.anops, and a lectotype is designated for the latter species. The new species is readily distinguished from congeners by having the frontal scale divided (vs single), and a central foramen in the parabasisphenoid (vs foramen absent). High-resolution x-ray computed tomography (HRXCT) was used to study and present data on the skull of the holotype of the new species, the lectotype of L.anops, and the holotype of L.ternetzii. Additionally, extensive study of skull characters and external morphology failed to find diagnostic characters to differentiate L.beui and L.ternetzii, and the former is here considered a junior synonym of L.ternetzii, which is also redescribed.

Helminthophis anops was described by Cope (1899) based on two specimens; he wrote: "The collection which furnishes the basis of the investigation presented in the following pages was made in Colombia, near Bogota, for the World's Exposition of Chicago, where it was exhibited in the department of New Granada. The number of species is fifty-four, of which nine are new to science. I have not been able to ascertain the exact localities at which the specimens were obtained, but most of them, it is believed, were found in the neighborhood of Bogota" (Cope 1899: 3). Subsequently, Dunn (1944) transferred H. anops to Liotyphlops, also in Anomalepididae. Cope (1899: 10-11) also wrote: "This species has a tendency to subdivision of scales. In one of the two specimens the frontal is divided into two regular scales, and in another the lower extremity of the first labial is cut off on one side". The holotype of the new species described here (AMNH R-9550) is one of the two syntypes of H. anops and distinct from the other syntype (AMNH R-17540) in possessing, among other diagnostic characters, the frontal scale divided (vs single), and a central foramen in the parabasisphenoid (vs foramen absent). The other syntype (AMNH R-17540) is consistent with the species currently identified in Colombia as L. anops.
Taxonomic changes over the past century have also included two other species of Liotyphlops: L. beui and L. ternetzii. The original description of L. ternetzii, by Boulenger (1896, as Helminthophis ternetzii) was based on a single specimen from "Paraguay" (holotype BMNH 1946.1.11.77). Later, Smith and Grant (1958) recognized Liotyphlops as a genus distinct from Helminthophis, highlighting as a diagnostic character the separation of prefrontal scales in Liotyphlops, while in Helmintophis the prefrontal scales are widely in contact. They transferred Boulenger's species to Liotyphlops. Boulenger (1896: 584) characterized this species as: "rostral two fifths the width of the head, extending nearly to the level of the eyes, forming a broad, straight suture with the frontal, which is about twice as broad as long; eye scarcely distinguishable through the ocular; two superposed preoculars and a subocular; four upper labials, first largest, second and third in contact with the lower preocular, third and fourth in contact with the subocular. Diameter of body 52 times in total length; tail nearly twice as long as broad, ending in a spine. 22 scales round the body. Olive above and beneath; head and anal region yellowish. Total length 335 mm." Liotyphlops beui was originally described by Amaral (1924), as Helminthophis beui from Butantan, São Paulo, Brazil (holotype IB 1806 and paratypes IB 281, IB 282, IB 652, and IB 1041). Amaral (1924: 29) characterized his new species as: "snout acutely rounded; rostral about two fifths the width of the head, not extending posteriorly to the vertical plane of the eyes, rounded posteriorly and forming a narrow suture with the frontal; frontal only about three times as wide as long; one subocular; two preoculars; eye under the suture between the ocular and lower preocular; four upper labials, 1 st largest, 2 nd and 3 rd in contact with the subocular, which separates them from the lower preocular; prefrontal separated from the 2 nd labial by the lower preocular, nasal and subocular. Tail more than twice as long as broad, ending in a spine. 22 scale rows around the body. Dark brown to blackish brown; head, as well as anal region and surroundings, light yellow; terminal spine yellowish. Total length, 290 mm; tail, 10 mm." Only five years after the original description, H. beui was placed in the synonymy of H. ternetzii by Amaral (1929) himself, but 55 years later Dixon and Kofron (1984) believed the species was valid and removed it from the synonymy of L. ternetzii based on the possession of 20 scale rows around the posterior body (22 in L. ternetzii) and fewer dorsal scales, 384-455 (vs 463-510 in L. ternetzii). Despite some authors maintaining L. beui as synonym of L. ternetzii (e.g., Peters et al. 1986), most subsequent authors have followed Dixon and Kofron (1984) and treated L. beui as a valid species (McDiarmid et al. 1999;Freire et al. 2007;Haad et al. 2008;Centeno et al. 2010;Wallach et al. 2014;Santos and Reis 2018;Boundy 2021;Linares-Vargas et al. 2021).
Here it is important to highlight the research of Dixon and Kofron (1984). They observed that most of the characters utilized for described forms are variable within populations, and occasionally the squamation is different on each side of the head in an individual. Also, according to Dixon and Kofron (1984), the nasal scale is divided and is variously called upper and lower nasals, preseminasals and postseminasals, anterior nasals and postnasals, or just nasals; additionally, the lateral and dorsomedian head scales are variously called subocular(s), preocular(s), ocular, supraocular(s), frontal, and postfrontal(s). They explained that much depends upon one's concept of the position of the scales as to whether there are two suboculars and one preocular, or two preocular and one subocular, or two supraoculars and one preocular, or two preoculars and one supraocular, etc. Accordingly to Dixon and Kofron (1984) the presence or absence of the division and/or fusion of scales on one side of the head and not on the other has been largely ignored by most describers of Liotyphlops species, which has, therefore, resulted in poor species concepts; the only scales that appear to be consistently defined in all writings are the rostral, prefrontal, and frontal scales.
In this paper, the validity of L. beui is revisited and L. ternetzii is redescribed. A new species of Liotyphlops is also described from the neighborhoods of Bogota, Colombia, a lectotype is designated for L. anops, and that lectotype is also redescribed. High-resolution x-ray computed tomography (HRXCT) was used to present data on the skull of the holotype of L. ternetzii and the holotype of the new species.

Materials and methods
I adopted the definition of the Unified Species Concept (Queiroz 2007), in which species are equated with independently evolving metapopulation lineages. In the absence of autapomorphy for species, consistent morphological difference among separate populations is used as a proxy for lineage independence. The study of external morphology was conducted under a stereomicroscope. The terminology used for the head squamation and scale counts follows Dixon and Kofron (1984) and Santos and Reis (2018). Measurements were taken with digital calipers and are presented as percent of total length (TL), except for subunits of the head, which are presented as percent of head length (HL). The results of morphometric analyzes are presented in the description. Specimens were not sexed and only adult specimens were examined (see Appendix 1). The photographs were obtained using a digital Nikon D5100 camera. For drawing preparation, a Wacom Intuos Draw CTL490DW digital tablet was used with the desktop digital stereomicroscope COSMOS LCD.
For the comparisons of Liotyphlops ternetzii and L. beui, the holotype of the former and paratypes of the latter were used. In addition, 50 specimens of each of these two species were measured and counted for the comparisons.
The head of the holotype of L. ternetzii and paratype of L. beui were studied by high-resolution x-ray computed tomography (HRXCT) at the high-resolution x-ray CT facility of the University of Texas at Austin using an Xradia microCT Scanner, and the holotype of the new species of Liotyphlops was studied by HRXCT at the high-resolution x-ray CT facility at Pontifícia Universidade Católica do Rio Grande do Sul using a Skyscan 1173 microfocus x-ray CT. The datasets were rendered in three dimensions using CTvox v. 3.2 (Bruker microCT, Inc., Billerica, MA) for Windows.
The terminology used for bones follows Rieppel et al. (2009), Santos and Reis (2018), and Santos and Reis (2019). The locality of the specimens was plotted using Google Earth Pro v. 7.3.2.5495, and the map was built with ArcMap (ArcGis) v. 10.4.1 for desktop using the WGS1984 geodetic datum. Geographical coordinates for historical specimens with imprecise locality records were approximated using the best evidence available and plotted with Google Earth. Only specimens actually examined were used in the map. Institutional abbreviations of specimens examined follow Sabaj (2020), with the addition of CEPB (Centro de Estudos e Pesquisas Biológicas da Pontifícia Universidade Católica de Goiás, Goiânia, Brazil).
Diagnosis. Liotyphlops palauophis sp. nov. is distinguished from all other Liotyphlops by having the frontal scale divided (vs single) and a central foramen in the parabasisphenoid (vs foramen absent). It is further distinguished from L. albirostris, L. argaleus, L. bondensis, L. caissara, L. haadi, L. trefauti, and L. wilderi in having two scales (vs one scale) contacting the posterior edge of the nasal between the second supralabial and prefrontal. It is further distinguished from L. beui, L. schubarti, L. taylori, and L. ternetzii by having four (vs three) scales contacting the posterior edge of the prefrontal. It is distinguished from L. anops by having 28/26/26 scales around the body and 19 subcaudal scales (vs 26/24/24 scales around the body and 12-14 subcaudal scales), and from L. sousai in having 573 dorsal scales and 561 ventral scales (vs 439 dorsal scales and 427 ventral scales).
Description. Meristic data in Table 1. Total length 361.2 mm, head length 5.3 mm (1.5% TL), snout-vent length 353 mm (97.7% TL), tail length 8.2 mm (2.3% TL), head width 3.8 mm (71.7% HL), and head height 3.1 mm (58.5% HL). Body covered with cycloid scales. Rostral scale large, longer than wide, contacting nasals anterolaterally, prefrontals laterally, and divided frontal posteriorly. Pair of triangular prefrontals bordered anterolaterally by rostral, ventrally by large divided nasal, and dorsoposteriorly by frontal. Posterior edge of prefrontals passing posterior edge of rostral. Frontal scale divided. Nasal scale divided and bordered anteriorly by rostral, dorsally by prefrontal, ventrally by first and second supralabials, and posteriorly by two scales that lie between prefrontal and second supralabial. Eye spot not visible. Four scales contacting posterior edge of prefrontal (three cycloid scales + frontal). Two scales contacting posterior edge of nasal between second supralabial and prefrontal. Six scales in first vertical row of dorsal scales. Mental triangular, not divided, wider than long, contacting first infralabials. Supralabials four, infralabials three. Scales around body 28/26/26. Dorsal scales 573, vental scales 561, and subcaudal scales 19.
Description of skull. High-resolution x-ray computed tomography of skull bones in Figs 3-5. Main body of premaxilla on ventral surface of snout. Maxillapremaxilla contact widely separated. Lateral maxillary foramina absent. Maxilla alveolar row oriented transversely. Nasal fused. Nasal-frontal boundary convex pos-      ipating in internal sidewall of neurocranium. External surface (dorsoposterior) of supraoccipital without transverse ridge. Supraoccipital-prootic contact narrow, less than half supraoccipital-parietal contact. Splenial not present as discrete element. Coronoid and angular separated by prearticular portion of compound bone. Retroarticular process long, longer than articular facet. Teeth present in maxilla, but lacking in dentary, premaxilla, palatine, and pterygoid. Coloration in alcohol. Dorsal and ventral body pale cream with brown pigmentation points along dorsal region of body.
Distribution. Known only from the type locality in the neighborhood of Bogota, Colombia (Fig. 6), according to the information provided by Cope (1899).
Etymology. The species name is in honor of Alfredo Palau Peña (June 10, 1969-August 8, 2020), a Brazilian herpetologist and my friend, who was killed by the COV-ID-19 virus. A combination of his name Palau and the Greek ophis, meaning snake.
Distribution. Central Colombia (neighborhood of Bogota and Villavicencio in the department of Meta) (Fig. 6).
[Liotyphlops] ternetzi- Smith andGrant 1958: 207. Liotyphlops ternetzii-Peters andOrejas-Miranda 1970: 183, in part;included L. beui in the synonymy. Diagnosis. Liotyphlops ternetzii is distinguished from L. anops, L. argaleus, L. sousai, and L. trefauti in having three (vs four) scales contacting the posterior edge of the prefrontal scale. It is distinguished from L. albirostris, L. bondensis, L. caissara, L. haadi, and L. wilderi in having two scales (vs one scale) contacting the posterior edge of the nasal between the second supralabial and the prefrontal. It is distinguished from L. taylori by having three (vs two) infralabial scales, and from L. palauophis sp. nov. in having a single frontal scale (vs frontal scale divided). Is distinguished from L. schubarti in the pale cream, dark brown, or black coloration (vs light brown).

Discussion
The description of new species based on single specimens is generally discouraged due to the obvious limitations, for example, in describing variation and geographical distribution (Santos and Reis 2018). More material will provide data on morphological variation, as well as ecological information that may be useful in conservation efforts. The redescription here of the lectotype of L. anops (AMNH R-17540) was based on photographs sent by the curators of the AMNH due to the great fragility of the specimen, making impossible the packing, transport, and the use of invasive techniques. The examination of these photographs of the lectotype was complemented by data obtained by the examination of other specimens of L. anops, providing the redescription with data of external morphology and osteology of the skull.
The specimens of L. beui (two paratypes and 50 non-types) and L. ternetzii (the holotype and 50 non-types) examined  showed limited variation in meristic characters (Table 2), which does not warrant the recognition of these taxa as separate species. The number and disposition of head scales do not distinguish the two taxa ( Fig. 12): (1) three scales contacting the posterior edge of the prefrontal; (2) two scales contacting the posterior edge of the nasal between the second supralabial and the prefrontal; (3) five or six scales in the first vertical row of dorsal scales; (4) four supralabial scales, and (5) three infralabial scales. The supratemporal bone of anomalepidid snakes is either very reduced or absent (Anomalepis), and the highresolution x-ray tomography showed that the two paratypes of L. beui (MCZ R-16702 and MCZ R-17842) lack a supratemporal, which is instead present, although highly reduced, in all other specimens of L. beui scanned and examined.
Liotyphlops beui was removed from the synonymy of L. ternetzii by Dixon and Kofron (1984) (Table 2). In addition, all other meristic characters (Table 2), coloration pattern, and an extensive study of skull bone characters showed no significant variation that can be used as diagnostic characters for L. beui. After a detailed morphological examination of specimens of L. beui and L. ternetzii, including the relevant type materials, L. beui is considered a junior synonym of L. ternetzii.
It is important to highlight that, in view of the limitation of diagnostic phenotypic characters for species of the genus Liotyphlops and the lack of knowledge about the evolutionary relationships of their species, there is a need for fieldwork to collect samples of fresh tissue to obtain genetic material, which will allow studying the systematics and testing the limits of Liotyphlops species from a molecular perspective.