Naja (Boulengerina) peroescobari, Bauer, 2017

Bauer, Aaron M., 2017, The “ Cobra-preta ” of São Tomé Island, Gulf of Guinea, is a new species of Naja Laurenti, 1768 (Squamata: Elapidae), Zootaxa 4324 (1), pp. 121-141 : 130-134

publication ID

publication LSID


persistent identifier

treatment provided by


scientific name

Naja (Boulengerina) peroescobari

sp. nov.

Naja (Boulengerina) peroescobari sp. nov.

Figs 4–5 View FIGURE 4 View FIGURE 5

Naja haje ( Greef 1884: 47; Bocage 1886: 69; Bocage 1895a: 24)

Naja haje var. nigra ( Bocage 1866: 51; 1879: 87; 1889: 34; Vieira 1886: 237)

Naja haje var. melanoleuca (Bedriaga 1982: 299, 440)

Naja melanoleuca ( Bocage 1905: 94; Manaças 1973: 228; Nill 1993: 71; Hofer 2002: 78, 89; Schatti & Loumont 1992: 31; Themido 1941: 6)

Naja melanoleuca melanoleuca ( Capocaccia 1961: 299; Manaças 1958: 190)

Naia melanoleuca . ( Boulenger 1906: 215; Ferreira 1902: 133)

Holotype. MUHNAC /MB03–001065 ( Fig. 4 View FIGURE 4 ), adult male collected in the vicinity of Praia Inhame (0.028636° N, 6.523203° E, WGS-84; 17 m above sea level), São Tomé Island, Republic of São Tomé e Príncipe, by Luis Ceríaco, Mariana Marques and Ana Carolina Sousa on 24 February 2016. The specimen was chopped in half by a local man with a machete, and collected minutes after this event. The choice of this particular specimen as holotype is justified as it is the source of the comparative molecular data.

Paratypes. All specimens from the Island of São Tomé, Republic of São Tomé and Príncipe. Eight specimens: IICT 18-1972 View Materials , adult male collected in Ribeira Peixe (0.090278° N, 6.615278° E, this and all following coordinates use map datum WGS-84; 17 m above sea level) by an unknown collector on 24 December 1972 View Materials GoogleMaps ; IICT 20-1967 View Materials , adult female collected in Santa Josefina (0.247882° N, 6.738315° E; 69 m above sea level) by an unknown collector on October 1967 View Materials GoogleMaps ; IICT 2-1966 View Materials , adult female, without specific locality ( São Tomé island) by an unknown collector on 20 February 1966 ; ZMH R10526 View Materials , unsexed adult, without specific locality ( São Tomé island) collected by Carl Weiss in the late 1840s ; ZMH R10527 View Materials , unsexed adult, without specific locality ( São Tomé island) collected by Richard Greef in 1879/1880 ; MHNG 2518.41 View Materials and MHNG 2518.42 View Materials , both unsexed adults collected in Rio Angra Toldo (0.156881 N, 6.668792° E; 75 m above sea level) by Tilman Nill in 3 April 1991 GoogleMaps ; MHNG 2462.43 View Materials , adult male collected in the environs of Neves (0.353192 N, 6.634972° E; 281 m above sea level) by Heft and Fahr on 8 August 1989 View Materials GoogleMaps .

Diagnosis. Naja peroescobari is placed in the subgenus Boulengerina (as defined by Wallach et al. 2009) on the basis of having 2–4 maxillary teeth, the penultimate supralabial height reaching eye level the combination of one preocular and one anterior temporal, rostral much broader than deep, internasals shorter than the prefrontals, dorsal scales smooth and fangs not modified for spitting. Morphologically, N. peroescobari differs from other members of the subgenus Boulengerina by a combination of coloration and scalation patterns. It differs from N. (Boulengerina) chrysti in having 19 dorsal scale rows (17 in chrysti ), three infralabials contacting the chinshields (four in chrysti ), and by its homogeneous dark coloration (brownish in chrysti ). Naja (Boulengerina) chrysti is considered to be strictly aquatic ( Chippaux 2006), whereas N. peroescobari is mainly terrestrial. It also differs from N. (Boulengerina) annulata by the lower number of dorsal scales (19 in peroescobari versus 21 to 25 in annulata ), and by its coloration, as annulata has between 21–23 dark bands (the first 3–5 simple, subsequent bands double) over a yellow or brown dorsum (contrasting with the homogeneous black of peroescobari ), and the ventral coloration (uniformly dark brown in annulata , versus white banded throat, followed by homogeneous black of peroescobari ). Naja peroescobari differs from N. (Boulengerina) multifasciata in its size (260 cm maximum in peroescobari versus maximum total length reported for multifasciata 76 cm; Chippaux 2006), by having two postoculars (three in multifasciata ), in having 19 dorsal scale rows (15–17 in multifasciata ), a higher number of subcaudals (30 to 39 in multifasciata versus 52 to 70 in peroescobari ), a higher number of ventral (153 to 175 in multifasciata versus 208 to 215 in peroescobari ), and by its coloration ( multifasciata is uniformly light yellow on the ventral side, and the dorsal scales are yellow in the anterior part and black in the posterior, with a dark head, separated by the rest of the body by a white collar; Chippaux 2006).

Naja peroescobari differs from the more closely related West African N. melanoleuca , with which it was formerly confused, having a combination of white and black bands, but the white bands are always confined to the first 22 ventral scales, whereas in N. melanoleuca these white bands extend throughout the first two thirds of the venter, usually between the 80th to the 100th scale (max. 125th scale, min. 50th scale; in the type series the last white ventral scale of syntype ANSP 6878 corresponds to the 81 st scale , the other having it on the 83th ( ANSP 6876), 86th ( ANSP 6875) and 90th ( ANSP 6879). This is even more evident when comparing the new species with N. subvulva , which has whitish scales mottled with dark spots extending all across the venter. Also the dorsal coloration of N.

peroescobari is uniformly shiny black, never exhibiting the whithish/yellow markings of the dorsal neck area common in N. melanoleuca . In terms of scalation, the most striking difference between N. peroescobari and N. melanoleuca is the contact of the posterior sublinguals. In N. peroescobari the posterior chin shields are mostly never in contact due to the intrusion of a large scale between them, whereas in N. melanoleuca these are always in contact, either fully across their extent or at least in the anterior part ( Meirte 1992 considered this character [souslinguales postérieures se touchent] as the way to differentiate N. melanoleuca in the key to African snakes). These differences were already mentioned by Capocaccia (1961) and are here confirmed with a larger sample of both taxa.

Description of holotype. Body dimensions: Snout-vent length 1700 mm, tail length 380 mm, dorsal head length (snout to end of parietal suture) 46.2 mm, lateral head length (snout to posterior end of lower jaw articulation) 61 mm. Head width across supraocular 24.8 mm, maximum overall width of head 45.5 mm. Head broad, hearth-shaped from above. Eye moderate, diameter half the distance from mouth or from nostril. Body scalation: 208 ventrals, 66 subcaudals, all paired, anal scale entire, the intact tail terminates in a spine. Dorsal scale rows: 24 on the neck, 19 at midbody, 14 one head length ahead of vent. Head scalation: preoculars 1/1, postoculars 3/3, supralabials 7/7 the third and fourth of which contact the orbital rim, infralabials 8/8 of which the first three contact the anterior chin shields. Cuneate scale absent. Anterior temporals 1/1, posterior temporal 2/2. Two temporals and nuchals contact the lateral and posterior edges of parietals. Rostral 1.3 times wider than high, visible from above. Anterior chin shields adjacent to each other, posterior chin shields separated by one elongate scale. Nasal scale entirely divided into a prenasal and a postnasal scale by the large, vertically elongate nostril. Frontal slightly longer than wide (11.7 x 11 mm), longer than distance from rostral (11.4 mm), shorter than supraoculars (14.2 mm), widest along anterior edge; shape pentagonal, anterior edge slightly concave, posterior edge ends in obtuse angle, border with supraoculars straight. Color and pattern: Head uniformly black dorsally, lateral side cream with large black marking on the anterior/posterior borders of the supralabials and infralabials. Underside of head uniformly pearly white. Dorsal color of the total length of body uniformly shiny black. Interstitial skin black. Throat scales and ventral pattern ( Fig. 4 View FIGURE 4 ): first ten scales uniformly white, except the seventh scale, which presents a small dark blotch at the left side of the scale. Eleventh ventral scale black on the half left, and white with small scattered black dots on the right. Twelfth ventral scale uniformly black. Ventrals 13–15 white, except the fourteenth scale, which has a small dark blotch at the left side of the scale. Ventrals 16–17 uniformly black. Ventrals 18–19 white, but the latter with some black intrusion on the posterior edge. The remainder of the ventral and subcaudal scales are uniformly black, and the distal lateral edges of these are continuous in color with the lower dorsal scale rows.

Variation. Variation in scale counts and body measurements of the paratypes of N. peroescobari are reported in Table 3. Midbody scale rows 19–21; Neck scale rows 21–25; Tail scale rows 13–16. Due to the limited number of N. peroescobari specimens for which sex was confirmed, no sexually dimorphic differences could be found in the subcaudal scales of the species, which range from 52 to 70 (52 to 69 in females, n= 4; 57 to 70 in males, n= 3). The same applies to ventral scales, which range from 208 to 215 (208 to 214 in females; 208 to 211 in males). The posterior sublinguals are always separated by the intrusion of a scale, however in paratype MHNG 2518.42 the intrusion of this scale between the posterior sublinguals is not evident, as in the other specimens, and both scales contact narrowly in the anterior part. All the paratypes have one anterior temporal followed by two posterior temporals. However in MCHNG 31002a, Capocaccia (1961) reported three posterior temporal instead of two. The anal scale is usually entire in all the type series, with the exception of the paratype IICT 18-1972, where it is divided. Supralabials always 7, whereas the number of infralabials varies between 7 and 9. Largest male examined ( MUHNAC /MB03–001065, Praia Inhame, São Tomé) 1700 + 380 = 2080 mm; largest female ( IICT 20-1967, Santa Josefina, São Tomé) 2180 + 420 = 2600 cm. All specimens examined were homogenous black on the dorsal areas, entirely agreeing with the holotype, while in the ventral pattern, despite small variations on the arrangement of the white scales (see Fig. 2 View FIGURE 2 ), no specimen had any white scales distal to the 22th ventral.

Etymology. The species is named after Pêro (also known as Pedro) Escobar (?–?), one of the 15th -century Portuguese navigators who discovered São Tomé Island on December 21, 1471, Annobon Island on January 1, 1472, and Príncipe island on January 17, 1472, and participated in the discovery of the ocean route from Portugal to India under the command of Vasco da Gama (circa 1460–1524) in 1497, as well as in the expedition of Pedro Álvares Cabral’s (circa 1467– circa 1520) that “discovered” Brazil in 1500. We adopt the Portuguese local common name “Cobra-preta” and suggest the English common name “ São Tomé cobra” for this species.

Distribution. The species appears to be widely distributed through the island, except the drier savanna areas in the northeast ( Fig. 6 View FIGURE 6 ). Greef (1884) noted that the species was confined to high areas, from 300 to up 900 m in the northeast “Cordilheira”, and that it was much more common on the southeastern part of the island, near Angra de São João (0.1167° N, 6.6500° E; 134 m above sea level), but it has subsequently been found elsewhere: Roça Minho (0.266667° N, 6.633333° E; 712 m above sea level; Bocage 1905), Roça Saudade (0.288319° N, 6.636209° E; 815 m above sea level; Bocage 1905), Água-Izé (0.218727° N, 6.727969° E; 19 m above seal level; Boulenger 1906, Capocaccia 1961), Roça Porto Alegre (0.033333° N, 6.533333° E; 16 m above sea level; Manaças 1958), Roça Santa Josefina (0.247882° N, 6.738315° E; 69 m above sea level; Manaças 1973), “Cascata” and surroundings (0.2854722° N, 6.6254416° E; 929 m above sea level; Nill 1993); Angra Toldo, also referred as Rio Angra Toldo (0.156881° N, 6.668792° E; 75 m above sea level; Nill 1993), Rio Quicha (location unknown; Nill 1993) and Generosa (0.348305° N, 6.551019° E; 81 m above sea level; Nill 1993). Two specimens from the California Academy of Sciences collections ( CAS 219402 and CAS 219403) are from Bom Sucesso (0.288394° N, 6.612155° E; 1154 m above sea level) and Morro Esperança (0.299743° N, 6.616667° E; 998 m above sea level), respectively. The species is also known from Roça Bombaim (0.246481° N, 6.632732° E; 465 m above sea level; LMPC & MPM pers. obs.). The type locality represents the southernmost known record of the species, and also the lowest elevation recorded (21 m). The species is not known from the neighboring Rolas Islet.

Medical relevance and folklore. There is currently no available information on the venom of this snake. Local people (as well as historical reports) refer to the snake as highly venomous, causing death in a few hours, but no credible reports about the symptomology were recorded during our expeditions to the island (L.M.P.C & M.P.M pers. obs.). Vieira (1886) mentions that the species is highly venomous (even recording the local saying: “ homem mordido, homem perdido ”, meaning “man bitten, man lost”), but only bites when attacked or stepped on. Local informants claim that when bitten, patients usually cut the afflicted limb off, and there are also reports of people who are on the top of palm trees to extract palm wine who simply toss themselves to the ground to avoid being bitten whenever they encounter a snake. The manuscript report of José Gonçalves Correia (1888–1954), Portuguese naturalist who collected birds in São Tomé for the American Museum of Natural History in 1928/1929, mentions that after a bite, the afflicted area should be burnt with a hot iron in the first 24 hours ( Correia 1928). Further studies on the venom and also on the symptomology and bite incident records are needed to understand the potential medical relevance of this species to the human population. A possible island effect in the venom should also not be discounted, as snake venom is known to be highly affected by ecological and geographic conditions ( Chippaux et al. 1991). In some parts of the island, namely around Roça Bombaim, the species is captured by locals to serve as a local delicacy, and the meat is sold in the capital city market (L.M.P.C & M.P.M pers. obs.). The recent development of political ties between São Tomé and Príncipe and China, and the expected arrival of Chinese workers to the country, is a potential cause of concern, as snakes from the genus Naja are commonly used in the Chinese Traditional Medicine ( Zheng & Zhang 2000) and gastronomy ( Zhou & Jiang 2005), and conservation problems related to the exploitation of wildlife by Chinese workers have been reported for other parts in Africa ( Munson & Ronghui 2012; Bersacola et al. 2014).

Habitat and Natural History notes. Bocage (1895) reports a skull of a Least Weasel ( Mustela nivalis ) in the stomach contents of one specimen. Specimen IICT 18-1972 had an invasive Black rat (Rattus rattus) in its stomach, but knowledge on the trophic ecology of the species is limited. Vieira (1886) mentions that the species prefer shaded and moist habitats. The snake is mostly seen in forested areas in the southern half of the island, being absent in the northeast regions. It is commonly seen on roads basking during the day.


Museu Nacional de Historia Natural e de CI�ncia


Zoologisches Museum Hamburg


Academy of Natural Sciences of Philadelphia


Museum d'Histoire Naturelle


Centro de Zoologia do I.I.C.T.


California Academy of Sciences


Milwaukee Public Museum














Naja (Boulengerina) peroescobari

Bauer, Aaron M. 2017

Naja melanoleuca melanoleuca

Capocaccia 1961: 299
Manacas 1958: 190

Naia melanoleuca

Boulenger 1906: 215
Ferreira 1902: 133

Naja melanoleuca

Hofer 2002: 78
Nill 1993: 71
Schatti 1992: 31
Manacas 1973: 228
Themido 1941: 6
Bocage 1905: 94

Naja haje

Bocage 1895: 24
Bocage 1886: 69
Greef 1884: 47

Naja haje var. nigra

Vieira 1886: 237
Bocage 1866: 51
GBIF Dataset (for parent article) Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF