Redescription of Apostolepis Longicaudata (Serpentes: Xenodontinae) with Comments on Its Hemipenial Morphology and Natural History

Home , Apostolepis, Elapomorphus, Phalotris, Xenodontinae

Herpetologica, 67(3), 2011, 318–331 E 2011 by The Herpetologists’ League, Inc.

REDESCRIPTION OF APOSTOLEPIS LONGICAUDATA (SERPENTES: XENODONTINAE) WITH COMMENTS ON ITS HEMIPENIAL MORPHOLOGY AND NATURAL HISTORY

1,3 1 2 FELIPE FRANCO CURCIO ,PEDRO M. SALES NUNES ,MICHAEL B. HARVEY , AND 1 MIGUEL T. RODRIGUES 1Departamento de Zoologia, Instituto de Biocieˆncias, Universidade de Sa˜o Paulo, Caixa Postal 11.461 CEP: 05422-970, Sa˜o Paulo, Brazil 2Department of Biological Sciences, Broward College, 3501 S.W. Davie Road, Davie, FL 33314, USA

ABSTRACT: We redescribe the rare elapomorphine snake Apostolepis longicaudata Gomes in Amaral on the basis of a detailed examination of the holotype and comparisons with new specimens from the Cerrado of Central Brazil. The new data include information on coloration in life, hemipenial morphology, and natural history. Two specimens identified as A. quinquelineata by previous authors are now attributed to A. nelsonjorgei, which seems to be the only species in the genus with subcaudal counts overlapping those of A. longicaudata.

RESUMO: Apresentamos a redescriçaõ da rara espećie de serpente Apostolepis longicaudata Gomes in Amaral com base no exame do holo´tipo e em comparaço˜es com treˆs novos exemplares do Cerrado do Brasil Central. O trabalho inclui dados ine´ditos sobre a coloraçaõ em vida, morfologia hemipeniana e histo´ria natural. Dois espećimes lineados previamente identificados como A. quinquelineata por outros autores saõ redeterminados aqui como A. nelsonjorgei, que parece ser a u´ nica espećie do geˆnero cujas contagens de subcaudais se sobrepo˜em a`sdeA. longicaudata. Key words: Apostolepis nelsonjorgei; Apostolepis nigrolineata; Apostolepis quinquelineata; Cerrado; Elapomorphini; Snakes; Taxonomy

THE SNAKE genus Apostolepis includes more coloration, complemented by illustrations of than 30 small to medium-size fossorial spe- the head, fore trunk, and ventral surface of the cies widely occurring in cis-Andean South tail (Fig. 1). As suggested by the specific epithet, America (Ferrarezzi et al., 2005; Harvey, the high number of subcaudals is the most 1999; Lema, 2001). Along with the genera remarkable diagnostic feature of this species Elapomorphus and Phalotris, it is allocated to when compared with the lower counts of most the tribe Elapomorphini and is characterized congeners. by lack of distinct internasals (possibly fused After the original description, A. longicau- to their respective frontal shields) and a black data has only been mentioned in passing (e.g., band on the tip of the tail (Boulenger, 1896; Lema, 2001; Peters and Orejas-Miranda, 1970) Ferrarezzi, 1994; Harvey, 1999; Zaher et al., and, besides the holotype, no new specimens 2009). were reported. During surveys preceding con- Because of their cryptozoic nature mostly struction of a railroad in the Rio Tocantins associated with fossoriality, several species of Depression, municipality of Guaraı´, state of Apostolepis are rare and known from few Tocantins, we obtained three specimens of specimens (Harvey, 1999). Small sample sizes A. longicaudata. On the basis of these new hinder assessment of geographic variations specimens, as well as on direct examination and make it difficult to diagnose species in this of the holotype, we provide a redescription genus. containing data on hemipenial morphology, dis- Apostolepis longicaudata is an example of tribution, and natural history. In the context such a rare species. Its brief description (Gomes of this study, we also discuss some confusing in Amaral, 1921a) was solely based on the information from the literature regarding other holotype and included data on scalation and striped Apostolepis, as well as the taxonom- ic significance of two specimens of uncertain identity that are possibly related to A. long- 3 CORRESPONDENCE: e-mail, [email protected] icaudata.

318 September 2011] HERPETOLOGICA 319

FIG. 1.—Original illustration of the holotype of Apostolepis longicaudata (modified from Amaral, 1921a).

MATERIALS AND METHODS wider-than-long system (Peters, 1964). Tooth Specimens examined and data on localities counts were taken on the right maxillary bone are listed in Appendix I. Institutional acro- only. The diagnosis of A. longicaudata was re- nyms are IBSP (Instituto Butantan, Saõ Paulo), constructed according to the standardized num- CHUNB (Coleçaõ Herpetolo´gica da Universi- bered system introduced by Harvey (1999). The hemipenis of MZUSP 12694 was pre- dade de Brası´lia, Distrito Federal), MPEG pared following procedures described by (Museu Paraense Emı´lio Goeldi, Bele´m, Para´), Manzani and Abe (1988), subsequently mod- and MZUSP (Museu de Zoologia da Universi- ified by Pesantes (1994) and Zaher (1999). dade de Saõ Paulo, Saõ Paulo). The collec- Hemipenial calcareous structures were stained tion permit was provided by the Instituto in an alcoholic solution of alizarin red (adapted Nacional do Meio Ambiente e Recursos Naturais from Harvey and Embert, 2008; Harvey et al., Renova´veis (193/2001-DIFAS; Processo IBAMA 2008; Uzzell, 1973). Nomenclature regarding o n 02001.003451/93-13). hemipenial morphology follows Dowling and To the nearest 1 mm, we measured snout– Savage (1960) and Zaher (1999). vent length (SVL) and tail length with a To test for sexual dimorphism in numbers graduated ruler. To the nearest 0.01 mm with of subcaudals, ventrals, and relative tail lengths, a digital caliper, we measured head length we tested for normality (Shapiro–Wilk test) and (from the rostral to the quadrate-mandibular homogeneity of variance (Levene’s test). Since articulation), head width (at widest point), our data for the first two characters did not sat- eye–nostril distance (on the right side of the isfy these assumptions, we used Welch’s non- head), mid-body diameter, mid-tail diameter, parametric t-test to compare means for the and eye diameter (measured only on the right sexes. For relative tail lengths, we used analysis side of head in all specimens, except that we of covariance, treating SVL as a covariate. An measured both eyes on the holotype of A. F-test for equality of the regression slopes was longicaudata). Nomenclature of scalation and used to test the assumption of parallelism. We color features are those of Harvey (1999). Ven- used the PAST (Hammer et al., 2001) software tral scale counts were performed using the program for these statistical analyses. 320 HERPETOLOGICA [Vol. 67, No. 3

Holotype.—IBSP 1684 (not labeled in field, destroyed by fire in 2010), an adult male from Engenheiro Dodt, municipality o Santa Filo- mena, state of Piauı´, northeastern Brazil, collected by Francisco de Assis I[n]glesias between 1916 and 1918 (the ‘‘n’’ in the collector’s surname is a typographic error; the correct spelling is ‘‘Iglesias’’). Originally described in 1919, by J. Florencio Gomes, formally published by Amaral (1921a). Referred specimens.—MZUSP 12692–12694 (field numbers MRT 7119, 7118, and 7531 respectively) collected 25–28 October 2001, in the ‘‘Fazenda Marupiara,’’ municipality of Guaraı´, state of Tocantins, Central Brazil, by D. Pavan and F. F. Curcio. Diagnosis.—A small species of Apostolepis, SVL up to 254 mm, differing from all congeners by the following combination of characters: (1) snout rounded or slightly acuminate, length of rostral visible from above slightly more than one-third its distance to frontal; (2) six supralabials; (3) preocular– nasal contact present (i.e., nasal and preocular not separated by prefrontal); (4) temporals absent (0+0), fifth and sixth supralabials in contact with parietals; (5) four infralabials contacting first pair of chin shields on each side of head; (6) ventrals 234–244 and subcaudals 49–52 in males (females unknown);

FIG. 2.—Dorasl (A), lateral (B), and ventral (C) views of (7) dorsal head plates uniformly dark brown, the head of the holotype of Apostolepis longicaudata except for small irregular white blotches on (IBSP 1684). Scale bar: 2.0 mm. anteriomedial surface of prefrontals; (8) light supralabial blotch large, covering posterior Apostolepis longicaudata Gomes in Amaral margin of third and almost whole area of (Figs. 2–4) fourth supralabial; (9) white and black nuchal collars completely absent; (10) five dorsal Apostolepis longicaudata:Gomesin Amaral, stripes similar in width (as wide as, or slightly 1921a. Santa Filomena, Piauı´, Brasil. wider than, one dorsal scale row); (11) dorsum

FIG. 3.—Apostolepis longicaudata from Guaraı´, Tocantins (MZUSP 12693), showing general dorsal coloration and detail of head in life (Photo: Ota´vio A. V. Marques). September 2011] HERPETOLOGICA 321

FIG. 4.—Sulcate (A), assulcate (B) and lateral (C) faces of the hemipenis of Apostolepis longicaudata (MZSP 12694); scale bar 5 5.0 mm. background light tan (orange-tan in life); (12) traces of such a structure were present even in terminal scale directed backward, mostly white, the smaller individual (e.g., MPEG 17817; SVL possibly invaded by black pigment from the 5 226 mm, tail length 5 15 mm). (2) Condition tail band at least on its dorsal surface; (13) of the black nuchal collar: A distinctive black chin and gular region immaculate cream. nuchal collar up to two dorsal scales long is Species with five dorsal stripes that are most present in A. quinquelineata (Ferrarezzi, 1993; likely to be confused with A. longicaudata Harvey, 1999; Lema and Renner, 1998) and are the predominantly Amazonian species A. absent in A. longicaudata. (3) Temporal for- nigrolineata and A. quinquelineata and the mulae: Usually 0+1, rarely 0+0inA. nigroli- Cerrado species A. nelsonjorgei, but over- neata, (Ferrarezzi, 1993; Harvey, 1999; Lema lapping values of subcaudal counts are only and Renner, 1998), contrasting with 0+0 in all present in A. nelsonjorgei. Nonetheless, A. known specimens of A. longicaudata. Boulen- longicaudata is unambiguously distinguishable ger’s (1896) original description mentions no from these three species by a suite of char- temporals in the holotype of A. quinquelineata acters: (1) condition of the white nuchal collar: (fifth and sixth supralabials contact parietals), A well-developed white nuchal collar is pres- but this character is apparently variable in this ent in A. nelsonjorgei (Lema and Renner, species, because there are specimens with 0+1 2004a) and entirely lacking in A. longicaudata. temporals. Thus, we do not recognize this Most specimens of A. nigrolineata have rudi- character as informative for distinguishing A. ments of a white nuchal collar, peculiarly ex- quinquelineata from A. longicaudata. (4) Black pressed as two well-defined pale blotches on caudal band: In A. quinquelineata the black each side of the nape (Lema and Renner, caudal band is restricted to the dorsal and dor- 1998:107,118). However, a few specimens have solateral surface of the tail (Ferrarezzi, 1993; complete white collars (e.g., MPEG 16137, Lema and Renner, 1998), contrasting with the 17671). Lema and Renner (1998) suggest that complete black band of A. longicaudata. (5) juveniles of A. quinquelineata may have rudi- Width of the lateral stripe: The lateral stripes of ments of a white nuchal collar. In our com- A. quinquelineata (covering first–fourth dorsal parative material, none of the specimens of this rows) are distinctively wider than the paraver- species can be considered a juvenile, but no tebral and vertebral ones (Lema and Renner, 322 HERPETOLOGICA [Vol. 67, No. 3

1998), contrasting with a lateral stripe (restrict- nuchal collar that is lacking in A. longicaudata;the ed to fourth dorsal row) similar in width to the labial blotch covers a larger area in A. striata— paravertebral and vertebral ones in A. long- third–sixth supralabials—than in A. longicaudata icaudata. (third and fourth supralabials), and the only known Five dorsal stripes are also present in Apo- specimen has 202 ventrals (234–244) and 26 stolepis arenaria, A. christineae, A. intermedia, subcaudals (Lema, 2004). Apostolepis philli- A. lineata, A. serrana, A. striata, A. phillipsi psi has narrow but distinctive white and black and A. vittata (Harvey, 1999; Lema, 2002, 2004; nuchal collars (lacking), a pointed snout, and 24 Lema and Renner, 2004c, 2006; Rodrigues, subcaudals (Harvey, 1999). Apostolepis vittata 1993), but all these taxa have considerably has up to 28 subcaudals, a pointed and strongly lower subcaudal counts and differ from A. projecting snout, five supralabials (six), and only longicaudata (characters in parentheses) in sca- three infralabials contacting the first pair of chin lation and coloration. Apostolepis arenaria has shields (four; Harvey, 1999). 21–31 subcaudals and distinctive white and Redescription of holotype.—A slender Apo- black nuchal collars (lacking). Furthermore, stolepis (mid-body diameter/SVL 5 0.012), this species has 168–181 ventrals, the lowest small in size; tail relatively long, 15.9% of total counts in the genus (Rodrigues, 1993). Aposto- length; tail cylindrical, its diameter roughly lepis christineae has up to 29 subcaudals, a uniform throughout its length, slightly smaller narrow black nuchal collar attaining the gular than mid-body diameter (mid-tail/mid-body region (lacking), a strongly projecting, flattened, diameter 5 0.79); head subcylindrical, longer and acuminate snout (rounded or slightly acu- than wide (head width/head length 5 0.58), minate), the rostral largely visible from above not distinct from neck; eyes small, eye diam- and mostly white on its dorsal surface (a small eter approximately one-tenth of head length portion visible from above and mostly black and slightly shorter than distance from its an- on its dorsal surface), all stripes dark and cohe- terior margin to nostril (eye diameter/eye– sive (at least the paravertebral stripe diffuse), a nostril distance 5 0.75); pupil subelliptical; black caudal band restricted to the dorsum, rostral rounded, not projecting over lower barely reaching the lateral margins of the sub- jaw, length of its portion visible from above caudals (complete caudal band), and the white slightly more than one-third its distance to portion of the terminal scale restricted to its anterior border of frontal; internasals absent; ventral surface (white tip and ventral surface of prefrontals paired, longer than wide, their the termimal scale; Lema, 2002). Apostolepis largest width two-thirds of largest length; fron- intermedia is only known from the holotype, tal roughly hexagonal, longer than wide, its which is apparently lost (Harvey, 1999; Lema, largest width three-quarters of largest length, 1993: 37). However, the original description anterior border contacting posteromedial bor- mentions three infralabials contacting the first ders of prefrontals, lateral borders contacting pair of chin shields (four), a pointed snout medial margins of supraoculars and lateropos- (rounded or slightly pointed), 217 ventrals terior borders contacting anteromedial borders (234–244), and 37 subcaudals (Koslowsky, 1898). of parietals; supraoculars large, roughly trape- Regarding A. lineata, the only objective data zoidal, slightly longer than wide, contacting allowing distinction from A. longicaudata are lateroposterior border of prefrontal and upper provided by Harvey (1999: 393), who reported posterior border of preocular anteriorly, lateral 24 subcaudals, and Lema and Renner (2004b), border of frontal medially, upper border of who mention a distinctly projecting snout (not postocular lateroposteriorly, and anterior bor- projecting). Apostolepis serrana has up to 33 der of parietal posteriorly; parietals paired, subcaudals, besides a distinctive black nuchal longer than wide, largest width of each ap- collar attaining the margins of the gular region proximately two-thirds the length of their mu- (lacking) and an entirely white termimal scale tual suture, anterior margin contacting poste- (marked by black on the dorsal and lateral rior margin of supraocular and upper posterior surfaces). Apostolepis striata has considerable margin of postocular, anteromedial margins posterior—up to three dorsal scales—and ventral contacting lateroposterior borders of frontal, extensions of the head cap, configuring a black lateral margins contacting upper borders of September 2011] HERPETOLOGICA 323 fifth and sixth supralabials; two enlarged occip- mental immaculate; dorsal ground color light itals, one on each flank of nape, clearly dis- tan, with five brown stripes similar in width (one tinguishable in size and shape from dorsals; vertebral, one paravertebral pair, and one occipitals roughly rectangular, obliquely ori- lateral pair); all but vertebral stripe extending ented, twice as long as wide, largely separated to posterior margin of brown cap; vertebral by three interoccipitals (dorsal scales of verte- stripe separated from brown cap by almost bral and both paravertebral rows); occipitals whole length of one dorsal scale; lateral stripes anteriorly contacting lateroposterior margins of light brown, mostly cohesive, more distinctive in parietals and upper posterior margin of sixth first third of trunk, covering central area of supralabial; nasal entire, roughly triangular, fourth row; paravertebral stripes light brown, posterior vertex broadly contacting preocular; diffuse, distinctively lighter than lateral stripes, nostril positioned in anterior third of nasal, covering upper margin of fourth, virtually whole partly visible from above and slightly directed area of fifth and lower margin of sixth rows; anteriorly; preocular roughly pentagonal, in vertebral stripe cohesive, dark brown, covering wide contact with nasal and forming an or- central area of eighth (vertebral) row, also thogonal directed suture with it; postocular marking upper vertexes of contiguous (seventh) roughly ‘‘C’’ shaped, with feeble posterior ver- rows; ventrals and subcaudals immaculate tex; postocular contacting third and fourth su- cream; terminal caudal band dark brown, 10/ pralabials ventrally, fifth supralabial posterior- 11 subcaudals in length; terminal scale almost ly, parietal dorsoposteriorly, and supraocular entirely white, with dark brown pigment dorsally; temporals 0+0; six supralabials, sec- extending slightly above its dorsal surface. ond and third contacting orbit, sixth largest, Measurements of the holotype.—SVL 227 mm, approximately as high as fifth; fifth and sixth tail length 43 mm, head length 4.90 mm, head contacting parietals; mental subtriangular, twice width 2.85 mm, mid-body diameter 2.81 mm, as wide as long; two pairs of chin shields, first mid-tail diameter 2.22 mm, eye-diameter 0.54/ pair distinctively longer; seven infralabials, first 0.57 mm, eye–nostril distance 0.80 mm (only pair in contact immediately behind mental, right side measured). widely separating it from chin shields; first– Variation and coloration in life.—Descrip- fourth infralabials contacting first pair of chin tive statistics on scale counts and measure- shields, four and five contacting second pair; ments for the four known specimens of A. fifth infralabial largest, roughly trapezoidal; 15 longicaudata are shown in Table 1. The dorsal scale rows without reductions near vent; largest specimen is a male (MZUSP 12692; anal plate divided; 241 ventrals; 52/52 subcau- SVL 5 254 mm, tail length 5 43); ventral and dals; trunk broken at level of 84th ventral shield; subcaudal counts range respectively from 235 terminal scale acuminate, directed ventrally; to 244 and 49 to 52; no females are known. maxillary teeth 5+2, fourth tooth missing on The snouts of MZUSP 12692–12694 are right maxillary bone; head light brown; brown slightly pointed (Fig. 5A) in comparison with cap covering whole dorsal and lateral surfaces of the holotype (Fig. 2A). Head scalation shows head, sporadically extending posteriorly over no variation in the four specimens. Coloration one or two dorsal scales and downward to upper of the holotype has considerably faded be- half of second dorsal scale row, slightly below cause of preservation, but MZUSP 12692– level of seventh infralabial; snout mostly brown, 12694 show minor differences between their with small irregular light spots on anteromedial coloration in life (Fig. 3) and after preserva- surface of prefrontals; two pale supralabial tion in 70% ethanol. Their brown head caps blotches; first supralabial blotch covering almost are considerably darker when compared with whole area of first supralabial; second suprala- the tonality present in the holotype. The caps bial blotch covering posterior margin of third extend downward on both sides of the head, and almost whole area of fourth supralabial; achieving the level of the second row of white and black nuchal collars absent; chin and dorsals (MZUSP 12693 and 12694) or further, gular region immaculate cream, first–sixth almost to the midventral line of the neck infralabials also immaculate; up to two-thirds (MZUSP 12692). The connections of the body of seventh infralabial marked by brown cap; stripes to the head cap are also variable: the 2 EPTLGC [ HERPETOLOGICA 324

TABLE 1.—Variation in some selected characters of striped-patterned Apostolepis examined herein (means 6 standard deviation and sample size follow ranges in parentheses).

Apostolepis aff. Apostolepis aff. Apostolepis longicaudata longicaudata1 longicaudata2 Apostolepis nelsonjorgei Apostolepis nigrolineata Apostolepis quinquelineata Tail length/snout–vent length Males 0.15–0.16 0.17 (n 5 1) 0.14–0.16 0.13–0.16 0.09 (n 5 1) (0.15 6 0.01, n 5 4) (0.15 6 0.01, n 5 3) (0.14 6 0.01, n 5 20) Females — 0.13 (n 5 1) 0.12–0.14 0.08–0.16 0.07–0.08 (0.13 6 0.02, n 5 2) (0.11 6 0.02, n 5 15) (0.08 6 0.00, n 5 3) Light snout spot Absent, or small when Large, well Large, well Large, well defined, Usually small Large, well defined present defined defined possibly paired Ventrals Males 234–244 229 (n 5 1) — 221–256 208–227 228 (n 5 1) (238.5 6 4.80) (238 6 17.52, n 5 3) (213.5 6 4.65, n 5 21) Females — — 237 (n 5 1) 260–261 214–259 231–243 (260.5 6 0.71, n 5 2) (230.7 6 10.10, n 5 15) (236.7 6 6.03, n 5 3) Subcaudals Males 49–52 54 (n 5 1) — 41–53 33–41 31 (n 5 1) (50.5 6 1.29) (37.5 6 0.48, n 5 0.48) Females — — 41 (n 5 1) 37–46 27–43 21–25 (30.7 6 3.97, n 5 3.97) (23.3 6 2.08, n 5 3) Preocular nasal contact Present Lacking Lacking Variable Present Present White nuchal collar Lacking Lacking Vestigial in adults Present Present (usually vestigial; Vestigial (only adults rarely complete) examined) Black nuchal collar Lacking Lacking Lacking Present Lacking Present 1 MZUSP 16615. 2 MZUSP 2405. o.6,N.3 No. 67, Vol. September 2011] HERPETOLOGICA 325

FIG. 5.—Comparisons of head shape and color patterns in dorsal perspective: (A) Apostolepis longicaudata (MZUSP 12693), (B) Apostolepis aff. longicaudata (MZUSP 2405), (C) Apostolepis aff. longicaudata (MZUSP 17614) and (D) A. nelsonjorgei (CHUNB 16182). Notice the variation regarding the white nuchal collar (absent in A and B, vestigial 583 in C, present and well developed in D). Scale bar 5 5.0 mm. vertebral and lateral stripes of MZUSP 12692 caudal band, and stripes are dark brown to and 12694 are not in contact with the brown almost black; the ground color of the trunk is head cap, but the same stripes are connected orange-tan; the venter is cream. to its posterior border in MZUSP 12693. Hemipenial morphology.—Description on Paravertebral stripes contact the brown head thebasis of the fully everted and expanded cap in all specimens; snout spots are virtually right organ of MZUSP 12964 (Fig. 4). Hemi- lacking in MZUSP 12693, restricted to sparse penis unilobed; distal end of retractor muscle light brown marks on the anterolateral surface divided; body corresponding to two-thirds of of the prefrontals in MZUSP 12692 and total length of organ; small spines uniformly represented by paired well-defined small light distributed on proximal one-third of organ’s tan spots on the anterior area of the prefrontals base, entire sulcate face, and lateral and apical in MZUSP 12694. The first light supralabial portions of capitulum; base ornamented with blotch is cream and occupies most of the first two pairs of rows of enlarged spines, one supralabial; the second light supralabial blotch lateral and composed of about five moderate- covers almost the whole area of the fourth ly-sized spines, second one parasagittal to supralabial and, in MZUSP 12694, it invades asulcate face and composed of four distinc- more prominently the posterior and lower tively larger spines on each row; latter pair of portions of the third and the anterior portion spine rows delimit an almost nude median of the fifth supralabials. The dorsal ground color area on asulcate face sparsely ornamented by of MZUSP 12692–12694 is considerably darker small spinules; lobe differentiated as unicaly- than in the holotype; likewise, the lateral brown culated capitulum, noncapitate, covered with stripes are cohesive, mostly restricted to the papilate calyces on sulcate and lateral faces; fourth dorsal row, flanked in some regions of the asulcate face of lobe restricted to concave trunk by weakly defined interrupted lines area, surrounded by small spines, forming formed by brown pigment on the apexes of inverted U-shaped and completely nude area; scales of the third and fifth rows; paravertebral sulcus spermaticus forked and divided ap- and vertebral stripes are similar to the ones of proximately in middle of organ; branches of the holotype. sulcus centrolineal on body, acquiring slightly Live specimens have considerably darker centrifugal direction on capitulum. The aliz- colors than preserved specimens (MZUSP arin red stained the spines and spinules, but 12692–12694). In life, the cephalic cap, terminal no other calcified structures were detected. 326 HERPETOLOGICA [Vol. 67, No. 3

FIG. 6.— White triangles 5 distributional records of Apostolepis longicaudata [(1) Santa Filomena, Piauı´ 5 type locality, (2) Guaraı´, Tocantins 5 locality of the three new records reported herein]; black triangles 5 records of A. aff. longicaudata [(1) Linhares, Espı´rito Santo, (2) Cabaceiras, Paraı´ba].

Distribution and ecology.—Apostolepis lon- assembled in 10 Y-shaped stations (Jones, gicaudata is reported from northeastern Brazil 1981). Each station consisted of one central (type locality) and the Rio Tocantins Depres- and three peripheral buckets, all connected sion. The type locality is in an ecotonal area to the first by plastic fences 4 m long and between the Cerrado and Caatinga biomes 50 cm high. (Fig. 6). Nonetheless, despite the presence MZUSP 12693 and 12694 were collected in of sparse xeric elements, the type locality is two small forest fragments, both on sandy soils generally dominated by cerrado vegetation. and covered by thick layers of leaf litter. No data on habitat or field notes are availa- MZUSP 12694 was collected in a gallery forest ble for the holotype. at an elevation of about 260 m (‘‘point 1’’ of Specimens MZUSP 12692–12694 were col- Fazenda Marupiara; see Appendix for precise lected in the municipality of Guaraı´, Tocan- coordinates); the canopy was mostly closed, tins, about 270 km NW of the type locality although sparse openings allowed sunlight to (Fig. 6). Quartzitic sandy soils predominate in touch the ground in some small clearings. the region (SEPLAN, 1999), and the land- Other reptile species recorded in the area scape is basically composed of typical open were Colobosaura modesta (Gymnophthalmi- cerrado vegetation interrupted by interfluvial dae), Kentropyx calcarata (Teiidae), and Typ- and gallery forests. These specimens were col- hlops brongersmianus (Typhlopidae; collec- lected in pitfall traps, always checked in the tion effort 480 bucket-days; sample period morning. The pitfall trap system at each of the 18–29 October 2001, 12 d). MZUSP 12693 collection sites included 40 buckets (height 5 was found in an interfluvial forest about 14 km 40 cm; diameter 5 30 cm; volume 5 18 L) southwest of the collection site of MZUSP September 2011] HERPETOLOGICA 327

12694 (‘‘point 2’’ of Fazenda Marupiara; see longicaudata may have a similar dis- Appendix for precise coordinates) at a slightly tribution. higher elevation. Large trees were still present The specimens from the Tocantins Depres- in the area, though presence of lianas, small sion have slightly acuminate snouts (Fig. 5A), shrubs, and several clearings were obvious contrasting with the rounded snout of the evidence of disturbance. Bachia micromela, holotype (Fig 2A). This difference cannot be C. modesta (Gymnophthalmidae), Anolis chry- attributed to allometric variation, because all solepis (Polychrotidae), Ameiva ameiva (Teii- specimens are adults. Alternatively, the slightly dae), and T. brongersmianus (Typhlopidae) acuminate snouts may be a local adaptation to were other reptile species collected at this the particular texture of quartzitic sandy soils site (collection effort 440 bucket-days; sample present in Guaraı´. period 18–29 October 2001, 11 d). Like other fossorial vertebrates, most Apos- In contrast to both previous individuals, tolepis species have short tails (subcaudal MZUSP 12692 was obtained in an open cer- counts lower than 40 scales). A relatively long rado area on quartzitic sandy soils (‘‘point 3’’ tail and high subcaudal counts for the genus of Fazenda Marupiara; see Appendix I for are the most important diagnostic features precise coordinates). Short and twisted trees of A. longicaudata. The only species that were sparsely distributed and small bushes undisputedly has subcaudal counts overlap- of grass were frequent. Narrow tracks of bare ping with those of this species is A. nelsonjor- sand remained directly exposed to sun (col- gei. Nonetheless, this fact is contradicted by lection effort: 480 bucket-days; sample period some recent publications where reports of 18–29 October 2001, 12 d). However, it is high subcaudal counts in other species were important to mention that, despite the striking based on misidentified specimens of A. differences in overall structure among the nelsonjorgei. forest areas of MZUSP 12693 and 12694 and On the basis of Lema and Renner (1998), the open cerrado of MZUSP 12692, the lat- Lema (2001:36) claimed that the high sub- ter specimen was collected in a pitfall station caudal counts of A. longicaudata would be placed contiguous to a gallery forest, quite ‘‘equal’’ to those of A. quinquelineata and A. similar in structure to the area of collection nigrolineata. According to Boulenger (1896), of MZUSP 12694. Thus, it would not be the holotype of A. quinquelineata is a male unreasonable to consider the record of specimen with 28 subcaudals (Ferrarezzi, 1993; MZUSP 12692 in an open cerrado area as a Hoge, 1958). Lema and Renner (1998:103) re- possible effect of the contact with forest port 22–30 subcaudals in females and 28–53 sub- habitats to which the species might show caudals in males of this species, suggesting exis- some degree of preference. tence of an enormous discrepancy between the ranges of both sexes. However, subcaudal counts DISCUSSION higher than 40 scales occur in only three spec- Until additional specimens are collected, imensexaminedbyLemaandRenner(1998; we can only speculate about the distribution MPEG 11487, 16954, and 17051). We re- of this rare elapomorphine. Colli et al. (2002) examined MPEG 16954 and 17051, both with considered A. longicaudata endemic to the 52/53 subcaudals, and identified them as A. Cerrado biome. Since the species occurs in nelsonjorgei. The third specimen (MPEG 11487, forested sandy soils of the Rio Tocantins with 48 subcaudals according to Lema and Depression, it may be widespread in similar Renner, 1998) was not available for study. Ex- substrates throughout the Cerrado of north- cluding these three specimens, male A. quinque- ern and northeastern Brazil. Several squamate lineata in the sample of Lema and Renner (1998) taxa such as Parapostolepis polylepis (Amaral, have 28–37 subcaudals, which is in better agree- 1921b, 1930; has not yet been formally ment with the variation observed in females. synonymized with Apostolepis)andAmphis- Regarding A. nigrolineata, nomenclatural baena ibijara (Rodrigues et al., 2003) appear confusion explains high subcaudal counts re- to be largely restricted to the sandy formations ported by Lema (1997, 2001) for this spe- of the Parnaı´ba sedimentary basin, and A. cies. Lema (1997) considered A. nigrolineata 328 HERPETOLOGICA [Vol. 67, No. 3

FIG. 7.—Relative tail lengths of two species of stripe-patterned Apostolepis from eastern Brazil.

(Peters, 1869) to be a senior synonym of A. females. Moreover, the relatively shorter tails of pymi Boulenger, 1903. Nonetheless, 1 yr male A. nigrolineata immediately distinguish this later, Lema and Renner (1998) used the species from A. longicaudata (F20,4 5 10.42, P 5 name A. pymi when referring to populations 0.004, Fig. 7). of the eastern Amazon. This inconsistent use In summary, there are no reliable records of of names is apparently due to timing of the more than 46 subcaudals in A. nigrolineata publications involved, because Lema (1997: and 37 subcaudals in A. quinquelineata.On 194, 197, 199) explicitly cites Lema and the other hand, considering the data from the Renner (1998). Presumably, Lema and Re- literature (Lema and Renner, 2004a), along nner (1998) was submitted before Lema with additional specimens examined herein (1997), and the features of A. pymi provided (Table 1), males of A. nelsonjorgei may have by Lema and Renner (1998) must be up to 53 subcaudals (MPEG 16954), clearly interpreted as features of A. nigrolineata overlapping the range of A. longicaudata. (see also Lema, 2001). In view of this During this study, we examined two striped- information, since Lema and Renner (1998) patterned specimens of doubtful identity (MZ- claim that subcaudals of A. nigrolineata range USP 2045 and MZUSP 17614) that deserve from 21 to 40 in females and from 33 to 46 in attention. Ferrarezzi (1993) considered these males, we believe that the report of 52 specimens related to A. longicaudata because subcaudals in A. nigrolineata by Lema (1997) they have 0+0 temporals, fifth and sixth supra- resulted from confusion with Lema and Renner’s labials in contact with parietals, five dorsal (1998) data for A. quinquelineata (MPEG 16954 stripes similar in width, a white termimal scale, and 17051, herein identified as A. nelsonjorgei). and subcaudal counts high for the genus (54/54 Among specimens examined by us (Table 1), and 40/41, respectively). Nevertheless, the male A. nigrolineata have fewer ventrals same author pointed out that both specimens (t21,15 5 6.17, P , 0.000), more subcaudals were distinct from each other and from A. (t20,15 5 6.01, P , 0.000), and relatively longer longicaudata with respect to features of head tails (F20,15 5 97.62, P , 0.000, Fig. 7) than scalation and color pattern. September 2011] HERPETOLOGICA 329

Herein, we opted for a cautious posture in the collection ‘‘Alphonse Richard Hoge’’ with respect to these specimens, avoiding a at the Instituto Butantan. This tragic loss final decision regarding their identification; of Neotropical snakes included the largest instead, we refer to them as Apostolepis aff. samples of Brazilian taxa, assembled through longicaudata. Both individuals differ from typ- more than a century of work. Along with sev- ical A. longicaudata in lacking preocular-nasal eral type specimens, Gomes’ holotype of A. contact, as well as by presence of a large light longicaudata has apparently disappeared and snout spot (vs. preocular-nasal contact present is probably lost forever. Our measurements and snout pattern mostly black with small and and recent photographs of IBSP 1684 are the irregular light blotches). In addition, MZUSP last information obtained through direct ex- 17614 has rudiments of a white nuchal collar, amination of this specimen. which is completely lacking in MZUSP 2045 Note added to proofs.— While our paper and in all four specimens of A. longicaudata was submitted, Lema and Albuquerque (compare head patterns on Fig. 5). A specimen (2010) resurrected A. pymi and synonymized illustrated by Freitas (2003:pl. 157), from A. quinquelineata with A. nigrolineata.The Itapebi, state of Bahia (15u579S, 39u329W, authors provided a list of characters that 177 m; sensu www.glosk.com), which we were distinguish A. nigrolineata from A. pymi; unable to locate, seems to fit the characterization however, interspecific variation in these of MZUSP 17614 at least with respect to color characters is greater than their brief note pattern. indicates. For example, Lema and Albuquer- Since the character states related to snout que (2010) claimed that only juvenile A. color and preocular-nasal contact are poly- quinquelineata have five dorsal stripes. In morphic in other striped-patterned Apostole- our sample, three adult specimens (MPEG pis (Harvey, 1999; Lema and Renner, 2004a), 17453, 17982–83, each more than 300 mm MZUSP 2045 and 17614 could be attributed SVL) retained five distinct and cohesive to A. longicaudata and these features would dorsal stripes, whereas Lema and Albuquer- be implicitly regarded as polymorphic also for que (2010) claimed that at least the paraver- this species. However, both specimens were tebral stripes tend to fade throughout devel- collected far from the known range of A. opment and become indiscernible in fully longicaudata. MZUSP 17614 is a female col- grown individuals. Herein, we continue to lected in the vicinities of the municipality recognize A. nigrolineata (A. pymi included) of Cabaceiras, state of Paraı´ba, northeastern and A. quinquelineata as defined respec- Brazil, in the core area of the Caatinga biome tively by Lema (1997) and Lema and Renner (Fig. 6). The specimen was found in a mesic (1998). Nonetheless, the status of these forest sheltered by a large block of crystalline names needs clarification and must be rocks and the soil was composed of moder- assessed elsewhere. ately compacted clay, densely covered with Acknowledgments.—We are indebted to Francisco L. leaf litter. On the other hand, MZUSP 2045 is Franco, Valdir J. Germano (IBSP), Ana L. Prudente, a male from Linhares, state of Espı´rito Santo, Paula C. R. de Almeida (MPEG), Hussam Zaher, and in the Atlantic Rainforest of southeastern Carolina Mello (MZUSP) for allowing access to specimens Brazil (Fig. 6); this specimen has obvious under their care. Dante Pavan coordinated fieldwork and provided important suggestions regarding general content anomalies in head scalation (Table 1). of the manuscript. Hebert Ferrarezzi, Giovanna Mon- Further study is required to determine wheth- tingelli, Ana L. Prudente, and anonymous reviewers er these specimens from eastern Brazil are provided constructive comments and criticisms to early conspecific with A. longicaudata. The eastern versions of the paper. Glaucia Pontes and Thales de Lema helped with literature. Ota´vio A. V. Marques provided localities appear ecologically quite different from photographs of Apostolepis longicaudata in life; Cristina the Cerrado, but resolution of this problem must A. Rheims and Antoˆnio D. Brescovit helped with the await collection of additional specimens. photographs of the holotype of A. longicaudata. FFC is Finally, we mention that this paper has a supported by a grant from the Conselho de Desenvolvi- mento Cientı´fico e Tecnolo´gico (CNPq; FFC–grant # sad peculiarity. On 15 May 2010, a tragic 152193/2008-1). PMSN (grant # 07/00811-8) and MTR accident caused a fire that consumed most of are grateful to Fundaçaõ de Amparo a` Pesquisa do Estado the more than 80,000 snake specimens housed de Saõ Paulo (FAPESP) for support. 330 HERPETOLOGICA [Vol. 67, No. 3

´ LITERATURE CITED HOGE, A. R. 1958. Etude sur Apostolepis coronata (Sauvage 1877) et Apostolepis quinquelineata Boulen- MARAL ´ ´ A , A. 1921a.Ultimos trabalhos ineditos de J. Flo- ger 1896 (Serpentes). Memo´rias do Instituto Butantan ´ rencio Gomes: Duas novas especies de Colubrideos opi- 28:73–76. stoglyphos brasileiros (Philodryas oligolepis e Apostole- JONES, K. B. 1981. Effects of grazing on lizard abundance pis longicaudata). Annaes Paulistas de Medicina e Cirurgia and diversity in western Arizona. Southwestern Natu- 9:157–163. ralist 26:107–115. AMARAL, A. 1921b. Contribuicao para o conhecimento dos KOSLOWSKY, J. 1898. Ofı´dios de Mato-Grosso (Brasil). ophidios do Brasil—A. Parte 1. Quatro novas especies Revista del Museo de La Plata 8:25–32. de serpentes brasileiras. Anexos das Memo´rias do Instituto Butantan (Ofiologia) 1:1–88. LEMA, T. 1993. Polimorfismo em Apostolepis dimidiata (Jan, 1862) com a invalidaçaõdeApostolepis villaricae AMARAL, A. 1930. Estudos sobre ophidios dos Neotropicos. Lema, 1978 e Apostolepis barrioi Lema, 1978 (Ser- XVII. Valor systematico de varias formas de ophidios pentes: Xenodontinae: Elapomorphini). Acta Biologica dos Neotropicos. Memo´rias do Instituto Butantan 4:3–68. Leopoldensia 15:25–52. LEMA, T. 1997. A redescription of the tropical Brazilian BOULENGER, G. A. 1896. Catalogue of the snakes in the British Museum (Natural History), Volume 3. British serpent Apostolepis nigrolineata (Peters 1869), (Colu- Museum, London, UK. bridae, Elapomorphinae), synonymous with A. pymi Boulenger. Studies on Neotropical Fauna and Envi- BOULENGER, G. A. 1903. Descriptions of new snakes in the collection of the British Museum. Annals and Magazine ronment 32:193–199. of Natural History, London 7:350–354. LEMA, T. 2001. Fossorial snake genus Apostolepis from South America (Serpentes: Colubridae: Elapomorphi- COLLI, G. R., R. B. BASTOS, AND A. F. B. ARAU´ JO. 2002. The ´ character and dynamics of the Cerrado herpetofauna. nae). Cuadernos de Herpetologıa 15:29–43. Pp. 223–241. In P. S. Oliveira and R. J. Marquis (Eds.), LEMA, T. 2002. Nova espećie de Apostolepis do grupo The Cerrados of Brazil: Ecology and Natural History lineata do sudoeste do Brasil (Serpentes, Elapomor- of a Neotropical Savana. Columbia University Press, phinae). Facena 18:41–52. New York, New York, USA. LEMA, T. 2004. Nova espećie de Apostolepis Cope do ˆ DOWLING, H. G., AND J. A. SAVAGE. 1960. A guide to snake estado de Rondonia, Brasil (Serpentes: Elapomorphi- hemipenis: Survey of basic structure and systematic nae). Comunicaço˜es do Museu de Cieˆncias e Tecnologia, characteristics. Zoologica 45:17–28. PUCRS, Se´rie Zoologia 17:81–89. EMA AND ENNER FERRAREZZI, H. 1993. Sistema´tica filogene´tica de Elapo- L , T., M. F. R . 1998. O status de morphus, Phalotris e Apostolepis (Serpentes: Colubri- Apostolepis quinquilineata Boulenger, 1896, A. pymi dae: Xenodontinae). 1993. Masters Thesis, Instituto de Boulenger, 1903 e A. rondoni Amaral, 1925 (Serpentes, Biocieˆncias, Universidade de Saõ Paulo, Saõ Paulo, Colubridae, Elapomorphini). Biocieˆncias, Porto Alegre Brazil. 6:99–121. FERRAREZZI, H. 1994. Uma sinopse dos geˆneros e LEMA, T., AND M. F. RENNER. 2004a. New species of classificaçaõ das serpentes (Squamata) II. Famı´lia Apostolepis from Alto Tocantins, Brazil, with comments Colubridae. Pp. 81–91. In L. B. Nascimento, A. T. on the stripe pattern species (Serpentes, Elapomorphi- Bernardes, and G. A. Cotta (Eds.), Herpetologia no nae). Biocieˆncias, Porto Alegre 12:139–145. Brasil, I. PUC-MG, Fundaçaõ Biodiversitas e Fundaçaõ LEMA, T., AND M. F. RENNER. 2004b. New specimens of Ezequiel Dias, Belo Horizonte, Brazil. Apostolepis vittata (Cope 1887) (Serpentes, Elapomor- FERRAREZZI, H., F. E. BARBO, AND C. E. ALBUQUERQUE. phinae). Caderno de Pesquisa, Se´rie Biologia, Santa 2005. Phylogenetic relationships of a new species of Cruz do Sul 16:51–56. Apostolepis from Brazilian Cerrado with notes on the LEMA, T., AND M. F. RENNER. 2004c. Description of a new assimilis group. Papeís Avulsos de Zoologia 45:215–229. species of Apostolepis Cope 1861 (Serpentes, Elapo- FREITAS, M. A. 2003. Serpentes Brasileiras. Malha-de- morphinae) from the Brazilian Cerrado. Acta Biologica Sapo Publicaço˜es e Consultoria Ambiental, Proquigel Leopoldensia 26:155–160. Quı´mica, Bahia, Brazil. LEMA, T., AND M. F. RENNER. 2006. A new species of HAMMER, Ø., D. A. T. HARPER, AND P. D. RYAN. 2001. Apostolepis with striped pattern, from Mato Grosso, PAST: Paleontological statistics software package for Brazil (Serpentes, Elapomorphinae). Cieˆncia em Mo- education and data analysis. Palaeontologia Electronica vimento 15:13–18. 4:1–9. LEMA, T., AND N. R. ALBUQUERQUE. 2010. The identity of HARVEY, M. B. 1999. Revision of Bolivian Apostolepis Apostolepis pymi and placement of A. quinquelineata in (Squamata: Colubridae). Copeia 1999:388–409. the synonymy of A. nigrolineata (Serpentes, Xenodon- HARVEY, M. B., AND D. EMBERT. 2008. Review of Bolivian tinae). Biota Neotropica 10:343–366. Dipsas (Serpentes: Colubridae) with comments on MANZANI, P. R., AND A. S. ABE. 1988. Sobre dois novos other South American species. Herpetological Mono- me´todos de preparo de hemipenis de serpentes. graphs 22:54–105. Memo´rias do Instituto Butantan 50:15–20. HARVEY, M. B., G. R. FUENMAYOR,J.R.C.PORTILLA, AND PAYNTER, R. A., JR., AND M. A. TRAYLOR,JR. 1991. J. V. RUEDA-ALMONACID. 2008. Systematics of the Ornithological Gazetteer of Brazil, 2nd Edition. Bird enigmatic dipsadine snake Tropidodipsas perijanensis Department, Museum of Comparative Zoology, Har- Alemań (Serpentes: Colubridae) and review of mor- vard College, Cambridge, Massachusetts, USA. phological characters of Dipsadini. Herpetological PESANTES, O. 1994. A method for preparing the hemipenis Monographs 22:106–132. of preserved snakes. Journal of Herpetology 28:93–95. September 2011] HERPETOLOGICA 331

PETERS, J. A. 1964. Dictionary of Herpetology. Hafner 1991; www.glosk.com) and then refined with the aid of the Publishing Company, New York, New York, USA. software Google Earth 5.0.11733.9347; coordinates of PETERS, J. A., AND B. R. OREJAS-MIRANDA. 1970. Catalogue localities marked with ‘‘**’’ were directly taken in the field of the Neotropical Squamata. Part 1. Snakes. United using a hand-held global positioning system receiver. States National Museum Bulletin 297:1–347. Apostolepis longicaudata.—BRAZIL: PIAUI´: Santa PETERS, W. 1869. U¨ eber neue Gattungen und neue oder Filomena* (09u079S, 45u559W, 277 m): IBSP 1684 weniger bekannte Arten von Amphibien (Eremias, (holotype); TOCANTINS: Guaraı´* (08u509S, 48u309W, Dicrodon, Euprepes, Lygosoma, Typhlops, Eryx, 259 m), ‘‘Fazenda Marupiara’’** (8u389S, 48u279W), point Rhynchonyx, Elapomorphus, Achalinus, Coronella, 1** (8u369S, 48u189W, 258 m): MZUSP 12964; point 2** Dromicus, Xenopholis, Anoplodipsas, Spilotes, Tropi- (8u389S, 48u259W, 296 m): MZUSP 12963; point 3** donotus). Monatsbericht der Ko¨niglich-Preussischen (08u159S, 48u199W, 250 m): MZUSP 12962. Akademie der Wissenschaften zu Berlin 1869:432–447. Apostolepis aff. longicaudata.—BRAZIL: ESPI´RITO RODRIGUES, M. T. 1993. Herpetofauna das dunas inter- SANTO: Linhares*: (19u239S, 40u049W, 33 m): MZUSP iores do Rio Saõ Francisco: Bahia: Brasil V. Duas novas 2405; PARAI´BA: Cabaceiras**: (07u299S, 36u179W, espećies de Apostolepis (Ophidia, Colubridae). Mem- 388 m): MZUSP 17614. o´rias do Instituto Butantan 54:53–59. Apostolepis nelsonjorgei.—BRAZIL: GOIA´ S: Serra da RODRIGUES, M. T., G. V. ANDRADE, AND J. D. LIMA. 2003. A Mesa dam, Point 1** (14u019S; 48u199W, 460 m): MZUSP new species of Amphisbaena (Squamata, Amphisbae- 10997; Conceiçaõ do Araguaia* (08u159S, 49u159W, nidae) from state of Maranhaõ, Brazil. Phyllomedusa 165 m) MPEG 16954; 17051; TOCANTINS: Estaçaõ 2:21–26. Ecolo´gica Serra Geral do Tocantins** (11u079S, 46u469W, SEPLAN (SECRETARIA DO PLANEJAMENTO E MEIO AMBIENTE, 560 m): MZUSP 17615; Porto Nacional*, (10u429S, TOCANTINS). 1999, Atlas do Tocantins: subsı´dios ao 48u259W, 212 m): CHUNB 16182. planejamento da gestaõ territorial. Diretoria de Zonea- Apostolepis nigrolineata.—BRAZIL: PARA´ : Ananin- ´ ˆ mento Ecologico-Economico (DEZ), Palmas, Tocantins, deua* (01u279S, 48u229W, 17 m): MPEG 2368, 6943, Brazil. 9459; Benevides, Santa Ba´rbara* (01u229S, 48u159W, UZZEL, T. 1973. A revision of lizards of the genus 28 m): MPEG 18513; Bragança* (01u039S, 46u469W, Prionodactylus, with a new genus for P. leucostictus 19 m): MPEG 13009; Capitaõ Poço* (01u459S, 47u049W, and notes on the genus Euspondylus (Sauria, Teiidae). 73 m): MPEG 10461, 12170; Castanhal, Macapazinho* Postilla 159:1–67. (01u189S, 47u569W, 41 m): MPEG 11794, 12693; Coloˆnia ZAHER, H. 1999. Hemipenial morphology of the South Nova, BR-316 road, near Rio Gurupi* (approximate American Xenodontine snakes, with a proposal for a coordinates: 01u299S, 46u229W, 50 m): MPEG 15063, monophyletic Xenodontinae and a reappraisal of 15863; Conceiçaõ do Araguaia, Rio Maria* (08u159S, Colubroid hemipenes. Bulletin of the American Muse- 49u169W, 165 m): MPEG 17671; Curupati (not located): um of Natural History 240:1–168. MPEG 10887, 13260; Igarape´-Açu´ * (01u089S, 47u379W, ZAHER, H., F. G. GRAZZIOTIN,J.E.CADLE,R.W.MURPHY, 37 m): MPEG 925; Ilha de Mosqueiro* (01u079S, J. C. MOURA-LEITE, AND S. L. BONATTO. 2009. Molecular 9 * 9 phylogeny of advanced snakes (Serpentes, Caenophi- 48u24 W, 10 m): MPEG 12771; Inhangapi (01u26 S, 9 dia) with an emphasis on South American Xenodon- 47u55 W, 35 m): MPEG 1571; Nova Vida, BR-316, 25 Km * tines: A revised classification and descriptions of new do Rio Gurupa´ (not located): MPEG 16227–29; Oure´m 9 9 taxa. Papeís Avulsos de Zoologia 49:115–153. (01u33 S, 47u07 W, 12 m): MPEG 7016, 7019; Parua´, BR- 316 road (not located): MPEG 10835, 14352; Acara´ road, Km 16 (not located): MPEG 12803–04; Santa Maria, .Accepted: 11 May 2011 rodovia BR-226 (not located): MPEG 16137; Maracana˜ .Associate Editor: Christopher Austin road, km 23 (not located): MPEG 8187; Santa Rosa (not located): MPEG 11835–836, 12590; Santare´m Novo, Trombetinha* (00u569S, 47u249W, 30 m): MPEG 7081; Santo Antoˆnio do Taua´* (01u099S, 48u089W, 51 m): APPENDIX I MPEG 7557; Tome´-Açu* (02u259S, 48u099W, 20 m): MPEG 11723; Viseu* (01u129S, 46u089W, 6 m): MPEG Specimens Examined 15126–27. Specimens are listed according to their respective Apostolepis quinquelineata.—BRAZIL: AMAZONAS: localities, organized as ‘‘COUNTRY: STATE: locality Presidente Figueiredo, Balbina Hydroelectric Power name’’; geographic coordinates are based on the WGS84 Plant* (02u029S, 60u019W, 122 m): MPEG 17453; RON- datum. Coordinates of localities marked with an ‘‘*’’ were DOˆ NIA: Porto Velho, Samuel Hydroelectric Power Plant* obtained primarily from gazetteers (Paynter and Traylor, (08u469S, 63u549W, 145 m): MPEG 17453, 17982–83.