The taxonomic status of Tantilla marcovani Lema 2004 (Squamata: Colubridae) Author Mata-Silva, Vicente Author Wilson, Larry David text Zootaxa 2016 4092 3 421 425 journal article 51576 10.11646/zootaxa.4092.3.7 aa04940b-88af-4787-b612-9fc2f8e9a08a 1175-5326 258595 B4AA687F-32AF-46A0-BCDF-7E899FC60533 The taxonomic status of Tantilla marcovani Lema 2004 ( Squamata : Colubridae ) VICENTE MATA-SILVA1 & LARRY DAVID WILSON2 1Department of Biological Sciences, The University of Texas at El Paso, El Paso, Texas 79968-0500, U.S.A. E-mail: vmata@utep.edu 2Centro Zamorano de Biodiversidad, Escuela Agrícola Panamericana Zamorano, Departmento de Francisco Morazán, Honduras ; 16010 SW 207th Avenue, Miami, Florida 33187-1056, U.S.A. E-mail: bufodoc@aol.com The colubrid snake genus Tantilla currently comprises 62 valid species distributed from the southern United States of America to southern Bolivia in the west, and Uruguay and northern Argentina in the east (Wilson & Mata-Silva 2014, 2015). Of these 62 species, 13 are found in South America (Wilson & Mata-Silva 2014, 2015). The most recently described of these Southamerican species is Tantilla marcovani Lema, 2004 . While we were working on a checklist and key to the members of the Tantilla clade (sensu Holm 2008, including the genera Geagras Cope 1876 , Scolecophis Fitzinger 184 3 , Tantilla Baird & Girard 1853 , and Tantillita Smith 1941 ; Wilson & Mata-Silva 2015), it became apparent that the diagnosis of T. marcovani entirely falls to distinguish this nominal taxon from the widespread and extensively variable T. melanocephala Linnaeus 1758 . Tantilla melanocephala , as presently envisioned, is the most broadly distributed member of its genus (Wilson & Mena 1980; Savage 2002; Greenbaum et al . 2004), occurring from Panama southward into the major portion of the South American continent, apart from Chile and southern Argentina . The concept of Tantilla melanocephala has changed somewhat since the work of Wilson & Mena (1980). This conceptual change was effected mainly by two studies. Initially, Savage (2002) resurrected two names ( T. armillata Cope “1875” [1876] and T. ruficeps Cope 1894 ) from the synonymy of T. melanocephala , as predicted by Wilson & Mena (1980). Savage (2002: 693) examined the status of the populations allocated to T. melanocephala by Wilson & Mena (1980) and concluded that “in Costa Rica , where the ranges of T. armillata and T. ruficeps meet, the two behave as distinct species [and that] the populations referred to T. melanocephala from Guatemala to western Panama are considered to constitute two distinct species, T. armillata and T. ruficeps , that are marginally sympatric in central Costa Rica .” Subsequently, Greenbaum et al . (2004) undertook a detailed study of the variation of 42 specimens allocated to T. melanocephala from Panama , Colombia , and Ecuador west of the Andes, as compared to that exhibited by four specimens allocated to T. equatoriana Wilson & Mena 1980 , including the holotype of this nominal species. Greenbaum et al . (2004: 457) concluded that, based on their analysis of several characters of color pattern and the numbers of subcaudals, “and a Principal Components Analysis of morphometric variation of T. equatoriana and T. melanocephala ,” the characters used by Wilson & Mena (1980) to diagnose T. equatoriana all overlap with those evidenced by the material of T. melanocephala they examined; thus, they treated “ T. equatoriana as a junior synonym of T. melanocephala .” As a result of the actions taken by Savage (2002) and Greenbaum et al . (2004), the range of T. melanocephala became restricted to South America and central and eastern Panama . Lema (2004) described Tantilla marcovani , based on a single specimen ( Figs. 1–2 ), an adult male from “Pico do Jabre, highest point of Paraiba State, Brazil at 50 km SW from Teixeira city ( 07° 07’48’’S , 37° 08’60’’W ), Matureia locality” (Lema 2004: 269). He further noted that “the region is located in highlands within Caatinga, with 1090 m above sea level.” Since the original description, the name Tantilla marcovani has appeared infrequently in the literature (Freitas & Silva 2007; Hamdan & Lira-da-Silva 2012; Guedes et al . 2014). Hamdan & Lira-da-Silva (2012) reported a specimen identified as T. marcovani from Santo Inácio in the state of Bahia, which they indicated constituted the second record for the species from a locality ca. 456 km N of the type-locality. They did not state, however, how they identified their specimen as belonging to T. marcovani , as opposed to T. melanocephala . Guedes et al . (2014) indicated that “the determination of this species is problematic and we thus opted to restrict the name T. marcovani only to specimens collected in Pico do Jabre” (the type-locality). Thus, their distribution map (Guedes et al . 2014: figure 26.1) for this species and T. melanocephala indicates a single locality for T. marcovani (Pico do Jabre) located within the range of T. melanocephala , which is recorded from throughout the Caatinga. Based on the opinion expressed about the recognizability of T. marcovani by Guedes et al . (2014) and our own reading of the diagnosis of this nominal taxon in Lema (2004), we have undertaken an examination of the utility of this diagnosis to discriminate between the restricted Tantilla marcovani and the widespread Tantilla melanocephala . FIGURE 1. Dorsal view of the holotype of Tantilla marcovani (MNRJ 6525). Photo by Marcos Bilate. FIGURE 2. Dorsal (A) and lateral (B) views of the head of the holotype of Tantilla marcovani (MNRJ 6525). Photo by Marcos Bilate. The description and diagnosis of Tantilla marcovani are based on the data of their holotype . The comparative statements in the diagnosis are based, apparently, on the features exhibited by a series of 20 additional specimens of T. melanocephala from Argentina , Brazil , and Uruguay , although Lema (2004) did not state this explicitly to be the case. Regardless, Lema did not examine material from elsewhere in the range of T. melanocephala in constructing the diagnosis for T. marcovani . In attempting to decipher the T. marcovani diagnosis, we constructed a table that indicates the condition of nine features of head shape and coloration. In doing so, we have hewed to the language used by Lema (2004), which as will be noted below, is, in some cases, indecipherable. Most of our commentaries are based on an extensive sample of Tantilla melanocephala examined by one of us (LDW), previously listed in Wilson & Mena (1980). In order to establish the taxonomic status of Tantilla marcovani we reviewed the features formally employed in its diagnosis, as following: (1) Head width, Lema (2004: 269) stated the head of T. marcovani is “wider” than that of T. melanocephala , which is indicated to be “elongated” ( Table 1 ). Since neither of these conditions is defined, they cannot be used to distinguish T. marcovani from T. melanocephala ; (2) Snout shape, Lema (2004: 269) indicated the snout to be “wide and straight” in T. marcovani and “slender and long” in T. melanocephala . Neither of the terms “wide” nor “slender” are defined and, so, cannot be used for discrimination. The terms “straight” and “long” are also not defined and, furthermore, do not describe conditions of the same character and also cannot be used for species discrimination; (3) Head diameter, Lema (2004: 269–270) described the head “diameter” (presumably the head width) as “wider than neck” in T. marcovani and “equal to neck” in T. melanocephala . The junior author of this paper (LDW) examined a large number of specimens of T. melanocephala during the writing of Wilson & Mena (1980) and found none in which the head was anything other than “wider than the neck.” Even so, this comparison is not supported by any mensural data and is based on a comparison of the holotype of T. marcovani with only 20 specimens of T. melanocephala from a restricted portion of its total range of distribution and, in our opinion, also is not usable to discriminate between the two taxa; (4) Snout color, Lema (2004: 270) indicated the snout color of T. marcovani to be “light immaculate” meaning presumably “immaculate cream” and that of T. melanocephala as “black or black blotched.” Wilson & Mena (1980), however, characterized both color pattern types C and D, which are the two found in Brazil , as having “pale pigment on internasals and prefrontals” ( Table 1 ). Therefore, this distinction does not hold up to scrutiny; (5) Black head cap, Lema (2004: 270) described the dark head cap as “small, restricted to frontal and parietal region” in T. marcovani and “over all the head from the snout to the nape” in T. melanocephala . Apparently, this distinction reflects Lema’s contention that T. marcovani differed from T. melanocephala in having a pale snout. As we noted above, however, the pattern types found throughout Brazil also have pale snouts, so this feature will not operate for discrimination either; (6) Color of head cap, Lema (2004: 270) stated that the dark head cap of T. marcovani was “grayish brown” (presumably in life) and that of T. melanocephala is “black.” The color photograph of the holotype of T. marcovani (figure 2 of Lema 2004 and Fig. 2 of present study), however, demonstrates the head cap to be dark brown in color with some paler marbling. Furthermore, Wilson & Mena (1980; Table 1 ) stated that the dark head cap in color pattern type C is “brown to dark brown” and “brown” in color pattern type D and not black; (7) Oblique black band over eyes, such a feature described by Lema (2004: 270) was indicated by Wilson & Mena (1980; Table 1 ) to be the “lateral extension of dark head cap.” Lema (2004) stated that in T. marcovani it is “attaining first supralabials, with remainder cream” and in T. melanocephala there is “none present on eyes and last supralabial.” Neither of these descriptions makes sense to us and otherwise do not describe the actual appearance of this feature in these snakes. The color photograph of the holotype of T. marcovani ( Fig. 2 ) illustrates that the lateral extension of the dark head cap does not reach the lip, differing slightly from the condition indicated in Table 1 of Wilson & Mena (1980) for color pattern types C and D, in which the lateral extension is said to reach the lip (as is evident, for example in the drawing of the head of a member of color pattern type D from Venezuela in figure 3 from Wilson & Mena 1980). Given that Lema’s description of this feature in the two taxa is incomprehensible, it is not possible to use it for the purposes of distinction; (8) Posterior end of head, it is not clear how Lema (2004) defined the “posterior end of [the] head,” but perhaps he meant the posterior end of the parietal scales. Otherwise, the posterior end of the head is defined usually as the point to which the posterior end of the lower jaw reaches. Regardless, Lema stated that the “posterior end of the head” is “cream attaining nuchal light collar” in T. marcovani and “not black with pair of narrow cream blotches” in T. melanocephala . These descriptions also are indecipherable, so their utility is indeterminable. Nonetheless, the drawing of the head of the holotype of T. marcovani in Lema (2004) and our photograph ( Fig. 2 ) indicates that the pale nuchal band begins at about the mid-point of the parietals and extends to about the mid-point of the first nuchal scale behind the posteriormost point of the parietal scales. The dark head cap is also almost in contact with an anteriomedial extension of the dark nape band, thus almost completely dividing the pale nuchal band medially; (9) Black cervical collar, this is the feature described by Wilson & Mena (1980) as the dark nape band. Lema (2004) described it in T. marcovani as “evident, only vertebral, three vertebral scales long” and in T. melanocephala as “not narrow, feeble or almost indistinct dark collar.” These statements do not function to describe the actual conditions in these snakes and, therefore, cannot be used to discriminate between them. Nonetheless, in the holotype of T. marcovani , the dark nape band is almost connected to the dark head cap along the common parietal suture and otherwise extends for four middorsal scales posterior to the posterior end of the common parietal suture ( Fig. 2 ). Wilson & Mena (1980: Table 1 ) indicated the mean length of the dark nape band in the members of color pattern type C to be 3.9 middorsal scales and that of those belonging to color pattern type D to be 3.6 middorsal scales. Thus, the condition in the holotype of T. marcovani is close to these mean values. TABLE 1. A synthesis of putative diagnostic features of the holotype of Tantilla marcovani compared with T. melanocephala (based on Lema 2004).

Diagnostic features	Tantilla marcovani	Tantilla melanocephala
Head width	Wider	Elongated
Snout shape	Wide and straight	Slender and long
Head diameter	Wider than neck	Equal to neck
Snout color	Light immaculate	Black or black blotched
Black head cap	Small, restricted to frontal and parietal region	Over all the head from the snout to the nape
Color of head cap	Grayish brown	Black
Oblique black band over eyes	Attaining first supralabials, with remainder cream	None present on eyes and last supralabial
Posterior end of head	Cream attaining nuchal light collar	Not black with pair of narrow cream blotches
Black cervical collar	Evident, only vertebral, three vertebral scales long	Not narrow, feeble or almost indistinct dark collar

Based on our analysis of the diagnosis for Tantilla marcovani , there are no features included within it that can act to distinguish (= unique state or combination of characters) this putative taxon from the widespread and highly variable T. melanocephala in its current concept. Although the number of ventral scales in the male holotype of T. marcovani was not used by Lema (2004) in his diagnosis, it is worthwhile to compare it to the data supplied by Wilson & Mena (1980, Table 3). This segmental count is the only one usable for such a comparison, since the tail of the holotype is incomplete. The ventral count given by Lema for this specimen is 145. This number is reasonably close to the mean number of ventrals given by Wilson & Mena (1980) for male T. melanocephala from Brazil , which is 143.4 (range 136–154). Thus, this feature would not have served either to differentiate T. marcovani from T. melanocephala . In conclusion, while attempting to construct an identification key for the 66 members of the Tantilla clade (sensu Holm 2008) we discovered that the diagnosis proposed by Lema (2004) for Tantilla marcovani would not serve to distinguish it from the widespread and extensively variable T. melanocephala and, thus, would not allow for the design of a couplet in the key to accomplish this. We undertook, therefore, to examine Lema’s diagnosis in detail. This analysis demonstrated that no single feature of this diagnosis would work to distinguish T. marcovani from T. melanocephala . Therefore, we conclude that Tantilla marcovani Lema, 2004 , should be placed into the synonymy of Tantilla melanocephala (Linnaeus, 1758) .