DataPDF Available

Somali Boidae (genus Eryx Daudin 1803) and Pythonidae (genus Python Daudin 1803) (Reptilia Serpentes)

P.K. Malonza and G.J. Measey
Somali Boidae (genus Eryx Daudin 1803)
and Pythonidae (genus Python Daudin 1803)
(Reptilia Serpentes)
B. LANZA 1,2 and A. NISTRI 2
1 Dipartimento di Biologia Animale e Genetica, Università degli Studi di Firenze, Via
Romana 17, 50125 Firenze, Italy
2 Museo di Storia Naturale dell’Università degli Studi di Firenze, Sezione di Zoologia
«La Specola», Via Romana 17, 50125 Firenze, Italy
Received 27 October 2004, accepted 25 May 2005
Data on nomenclature, taxonomy, morphology, geographical distribution
and natural history of Somali Boidae [the three monotypic species Eryx colubri-
nus (Linnaeus 1758), Eryx somalicus Scortecci 1939 and Eryx borrii n. sp.], and
Pythonidae [the monotypic species Python sebae (Gmelin 1788)].
KEY WORDS: Boidae, Pythonidae, Eryx, new species, Python, Somalia.
Introduction . . . . . . . . . . . . . . . . . 68
Material and methods . . . . . . . . . . . . . . 68
Superfamily Booidea Gray 1825 . . . . . . . . . . . . 68
Key to the families of Somali Booidea . . . . . . . . . 69
Family Boidae Gray 1825 . . . . . . . . . . . . 69
Genus Eryx Daudin 1803 . . . . . . . . . . . . 69
Key to Somali Eryx . . . . . . . . . . . . . 73
Eryx colubrinus (Linnaeus 1758) . . . . . . . . 73
Eryx somalicus Scortecci 1939 . . . . . . . . . 88
Eryx borrii n. sp. . . . . . . . . . . . . . 95
Family Pythonidae Fitzinger 1826 . . . . . . . . . . . 99
Genus Python Daudin 1803 . . . . . . . . . . . 100
Python sebae (Gmelin 1789) . . . . . . . . . . 101
Acknowledgements . . . . . . . . . . . . . . . 121
References . . . . . . . . . . . . . . . . . 121
Appendix 1 . . . . . . . . . . . . . . . . . 132
Tropical Zoology 18: 67-136, 2005
Corresponding author: Benedetto Lanza (E-mail:
68 B. Lanza and A. Nistri
69Somali Boidae and Pythonidae
The present paper concerns the taxonomy and biology of the python and the
three species of sand boas occurring in the Somali Democratic Republic, one of
them new to science.
Though Python sebae is one of the best known African snakes, we deemed it
essential to give a detailed account of its biology, until now confused with that of
the recently resurrected Python natalensis A. Smith 1840, a species widely distrib-
uted in southern and eastern Africa and locally sympatric with P. sebae in Kenya
and Tanzania. However we also deal rather extensively with the biology of Eryx, for
which the data are scattered in several, not easily found contributions.
We have studied 79 specimens of Eryx colubrinus (Linnaeus 1758), 8 specimens of Eryx
somalicus Scortecci 1939, 1 specimen of Eryx borrii n. sp., and 14 specimens of Python sebae
(Gmelin 1789).
For the study of variability we also utilized data found in the literature or provided by
The drawings were made from photographs under the supervision of the senior author.
The describing the dentition we used the following two useful but poorly known terms
proposed by WALL (1921): anonodont, «An uninterrupted series of teeth», and scaphiodont,
«Decreasing in length from before backwards. This type of dentition when seen in the lower
jaw remains one of the shape of the bows of a ship».
Abbreviations. Institutional acronyms follow LEVITON et al. (1985): ANSP (Academy of
Natural Sciences, Philadelphia, Pennsylvania, USA); BMNH [Natural History Museum, Lon-
don, formerly British Museum (Natural History), London, UK]; CAS (California Academy
of Sciences, San Francisco, California, USA); MCZ (Museum of Comparative Zoology, Cam-
bridge, Massachusetts, USA); MFSN (Museo Friulano di Storia Naturale, Udine, Italy); MNHN
(Muséum National d’Histoire Naturelle, Paris, France); MSNG (Museo Civico di Storia Natu-
rale «Giacomo Doria», Genoa, Italy); MSNM (Museo Civico di Storia Naturale, Milan, Italy);
MSNV (Museo Civico di Storia Naturale, Venice, Italy); MZUF (Museo di Storia Naturale,
Sezione di Zoologia «La Specola», dell’Università degli Studi, Florence, Italy); NMK (National
Museums of Kenya, Nairobi, Kenya, formerly Coryndon Museum); SMF (Natur-Museum
und Forschungs-Institut Senckenberg, Frankfurt am Main, Germany); ZMB (Museum für
Naturkunde, Universität Humboldt, Berlin, Germany).
SBS means «Spedizione Biologica in Somalia (= Biological Expedition to Somalia)
dell’Università degli Studi di Firenze (years 1959 to 1970)» or «Spedizione Biologica in Soma-
lia del Centro di Studio per la Faunistica ed Ecologia Tropicali del Consiglio Nazionale delle
Ricerche e dell’Università degli Studi di Firenze (years 1971 to 1986)».
Superfamily Booidea Gray 1825
According to the different authorities, this alethinophidian superfamily
includes only the family Boidae, with the subfamilies Boinae and Pythoninae (cf.
DOWLING 1959), or the two families Boidae and Pythonidae; some authors also rec-
ognize the family Erycidae, which was quoted with the wrong name Erycinidae by
WALLS (1998a, 1998b).
68 B. Lanza and A. Nistri
69Somali Boidae and Pythonidae
However the problems involved in Booidea phylogenesis appear to be far from
being resolved (see f.i. WALLS 1998a and 1998b). We believe that it will be better to
combine analyses based on morphological characters, such as those by UNDERWOOD
(1976), UNDERWOOD & STIMSON (1990) and KLUGE (1991, 1993a, 1993b), with in-
depth genetic research.
In the present paper we follow MCDOWELLs classication (1987), which treats
Pythonidae and Boidae separately, dividing the Boidae into Boinae Gray 1825 and Ery-
cinae Bonaparte 1831:
(i) Pythonidae Fitzinger 1826. «A supraorbital bone; anterior end of Meckelian carti-
lage free of dentary; palatine not independently erectile, its lateral process under-
lapped by an anteroposteriorly long medial process of maxilla; palatine enters
infraorbital fenestra between maxilla and pterygoid; palatine tooth row (absent in
Calabaria) continuous with that of pterygoid; rectus capitis dorsalis inserted on a
transverse nuchal crest of supraoccipital; oviparou(MCDOWELL 1987: 28).
(ii) Boidae Gray 1825. «No supraorbital bone; tip of Meckelian cartilage enclosed
by dentary; palatine erectile independently of maxilla (immobile in Lichanura
and Charina, where retractor palatini muscle is lost), the medial process of
the maxilla meeting only extreme rear of palatine and nearly or quite meeting
pterygoid to exclude palatine from infraorbital fenestra; palatine dentition
(reduced in Lichanura and Charina) set off from the pterygoid tooth row and
more medially placed; rectus capitis dorsalis insertion converging medially
on a sagittal crest of supraoccipital produced backward dorsal to exoccipitals
(unique); live-bearing» (MACDOWELL 1987: 29).
For recapitulatory data on the taxonomy, ecology, ethology and rearing of
Booidea see f.i. MINTON & RUTHEFORD MINTON (1973), STAFFORD (1986), SCHMIDT D.
(1989), VOGEL (1996), WALLS (1998a, 1998b), and STÖCKL & STÖCKL (2003).
Key to the families of Somali Booidea
1 Top of head covered with small scales; no labial pits .................................. Boidae
Top of head partly covered with symmetrical shields; labial pits
present .................................................................................................... Pythonidae
Family Boidae Gray 1825
According to MCDIARMID et al. (1999), the family includes 5 genera of Boinae
(Boa Linnaeus 1758, Candoia Gray 1842, Corallus Daudin 1803, Epicrates Wagler
1830, and Eunectes Wagler 1830) and 3 of Erycinae (Charina Gray 1849, Eryx Daudin
1803, and Gongylophis Wagler 1830, here tentatively considered a synonym of Eryx).
Only the genus Eryx occurs in Somalia.
Genus Eryx Daudin 1803
Eryx DAUDIN 1803a: 188 [type species: Boa turcica Olivier 1801 (= Anguis Jaculus Linnaeus 1758),
of the Middle East and northern Africa, by monotypy; see also ZHAO & ADLER 1993: 199 for
details on the date of publication].
70 B. Lanza and A. Nistri
71Somali Boidae and Pythonidae
Chlothonia DAUDIN 1803b: 283 [type species: Boa Anguiformis Schneider 1801 (= Boa johnii Russell
1801) by monotypy].
Gongylophis WAGLER 1830 [type species: Boa Conica Schneider 1801 by monotypy and subsequent
designation (FITZINGER 1843: 24)].
Cusoria GRAY 1849: 107 (type species: Cusoria elegans Gray 1849 by monotypy).
Cursoria GÜNTHER 1864: 333 (unjustied emendation of Cusoria).
Erys TESTI 1935: 107 (typographical error).
Erix LIPPARONI 1954: 647 (typographical error).
Neogongylophis TOKAR’ 1989: 54 (type species: Anguis colubrina Linnaeus 1758; described as a
subgenus of Gongylophis).
Pseudogongylophis TOKAR’ 1989: 54 (type species: Eryx jayakari Boulenger 1888; described as a
subgenus of Eryx).
Derivatio nominis. The genus name derives from the Latin Eryx (genitive Ery-
cis), the name of a mountain in the NW corner of Sicily, with a famous temple to
Venus at its top; also an eponymous hero, son of Venus, defeated by Heracles in a
boxing-match. According to GOTCH (1986: 128) «There is probably no special reason
for this generic name» as a number of names from classical mythology has been
used «without thought of any physical signicance». The gender of the genus is
Content and distribution. Thirteen species in SE Europe, N and E Africa, and
SW Asia to India and Sri Lanka, northward to Caucasia and Afghanistan, then east-
ward from Turkestan to S Mongolia and W China; the African species are Eryx col-
ubrinus (Linnaeus 1758), E. jaculus (Linnaeus 1758), E. muelleri (Boulenger 1892),
E. somalicus Scortecci 1939, and E. borrii n. sp.
RAGE (1972) and TOKAR (1989) proposed revalidation of the genus Gongylo-
phis Wagler 1830, an opinion not shared by UNDERWOOD (1976), RIEPPEL (1978) and
SZYNDLAR & SCHLEICH (1994). According to TOKAR (1989) in this genus the medial
crest of the pterygoid, relatively small, merges into its dentigerous process and the
palatine bears 5-6 teeth, while in Eryx the relatively large medial crest ends abruptly
close to the origin of the above-mentioned process and the palatine bears 4-5 teeth.
The Russian author (1989, 1995, 1996) includes in Gongylophis the species colubri-
nus (Linnaeus 1758), conicus (Schneider 1801), muelleri Boulenger 1892, and whita-
keri Das 1991; in Eryx the species elegans (Gray 1849), jaculus (Linnaeus 1758),
jayakari Boulenger 1888, johnii (Russell 1801), miliaris (Pallas 1773), somalicus
Scortecci 1939, tataricus (Lichtenstein 1823), and vittatus Chernov 1959. Instead,
according to SZYNDLAR & SCHLEICH (1994: 235) «as evidenced by their caudal oste-
ology, the living members of Eryx form two distinct groups: (1) the Afro-Arabian
complex including E. jayakari, E. somalicus, and E. muelleri, characterized jointly
by strongly reduced haemapophyses, the condition considered apomorphic; (2) the
Euroasiatic complex including the remaining species (perhaps also the African E.
colubrinus), with well developed haemapophyses (Z. SZYNDLAR unpubl.)». According
to KLUGE (1993a: 293) «Further resolution of Eryx species relationships is required
before Gongylophis (type species conicus) can be recognized». In the present paper
we prefer to follow LARGEN & RASMUSSEN (1993) in adopting the classical view and
retaining the name Eryx for both species of sand boas occurring in NE Africa.
Three species in Somalia: Eryx colubrinus, Eryx somalicus and Eryx borrii n. sp.
Fossils. Subfamily Erycinae appeared perhaps during the late Cretaceous
(Mesozoic era) and was extremely common during the Cenozoic era (RAGE 1987).
According to SZYNDLAR & SCHLEICH (1994: 235) the erycine snakes «were widespread
70 B. Lanza and A. Nistri
71Somali Boidae and Pythonidae
in the Palaeogene [Cenozoic era] of both Old and New World, but fossil nds of the
living genus Eryx have never been reported from pre-Miocene times» perhaps with
the only exception of a skull of Crythiosaurus from the Mongolian Oligocene (Pal-
aeogene period), originally described as an amphisbaenid. True Eryx are known to
have been widely distributed in the Mediterranean and Black Sea areas between the
end of the lower Miocene and the late Pliocene (Neogene period). Messinian (late
Miocene) remnants, probably belonging to Eryx, have been described from near Alba
(Piedmont) by CAVALLO et al. (1993). The available fossil remains suggest that most
representatives of Eryx inhabiting Europe in the past were similar to the recent Asi-
atic members of the genus, except for one or two taxa from S Spain (Eryx primitivus
Szyndlar & Schleich 1994, middle Pliocene; Eryx aff. primitivus, middle/late Pliocene)
belonging to the Afro-Arabian complex, displaying primitive conditions in their cau-
dal osteology and probably representing an early offshoot of a hypothetical ancestral
stock leading to the extant species of Eryx. According to HOLMAN (1998), E. jaculus
is only known from the Pleistocene (Quaternary period) of Europe [Greece: Chios Is.
(Middle Pleistocene) and Kos Is. (“Upper Quaternary”)]. For fossil erycine snakes see
also KLUGE (1993a) and RODRÍGUEZ-ROBLES et al. (1999).
Description. Size small to rather large: greatest adult total length from about
40 cm, e.g. in E. elegans (Gray 1849), to a little more than 1 m as in E. johnii and E
tataricus, according to SEUFER (2001). Head not or scarcely distinct from the neck;
body cylindrical, stout; tail short, thick, conical or terminating in a sharp point, not
or, very slightly prehensile. Eye small or very small, with vertical pupil. No labial
pits. Head covered with small, irregular, smooth scales (except on the snout); dor-
sals small, numerous, smooth or keeled; subcaudals entire, sometimes some of them
divided; anal single, only exceptionally divided. Mental groove present or absent. No
chin shields. Hemipenis extending to ventral 8 to 14 when retracted, feebly or not
bilobate, without papillae, ounced and calyculate, the sulcus spermaticus forked or
undivided. Skull and muscles: prefrontals widely separated; nasals broad posteriorly,
clearly longer than frontals; palatine mobile, strongly toothed, its choanal process
reduced; retractor arcus palatini present (palatine scarcely mobile, weakly toothed,
its choanal process large, and no retractor arcus palatini in the North American
Erycinae Lichanura and Charina). Maxillary and mandibular teeth numerous, solid,
ungrooved, arranged in an anonodont and scaphiodont set (only the rst tooth some-
times a little shorter than the second). ALEXANDER & GANS (1966) found in E. colubri-
nus (n = 2), E. conicus (Schneider 1801) (n = 1) and E. jaculus (n = 2) the following
combined data: 168(?)-188 (ventral) and 254-278 (dorsal) body scale rows, ca 18-28
(ventral) and ca 18-28 (dorsal) caudal scale rows, 174-190 body vertebrae, 19-29 tail
vertebrae; the dermal-vertebral ratios of the three species turned out to be: body (ven-
tral 1:1; dorsal 1.4:1), caudal (ventral 1:1; dorsal 1.0-1.1:1).
Biology (see also Eryx colubrinus, Biology). Terrestrial, fossorial or semifos-
sorial, predominantly crepuscular and nocturnal snakes inhabiting arid and sub-
arid regions, usually not strictly linked to sand substrata, notwithstanding their
common name; they are able to disappear very quickly into the sand where they
creep just beneath the surface. According to VAN WOERKOM (1987: 49) «a furrow
along the middle of the back and, sometimes, strongly keeled scales prevent the
sand from sliding off their backs and showing their presence». Sand boas may be
found above the ground during the night or at twilight, hiding by day under stones,
in soil crevices, in burrows or under the sand; occasionally they are active on the
72 B. Lanza and A. Nistri
73Somali Boidae and Pythonidae
surface also by day, especially with overcast sky. According to STAFFORD (1986: 106),
the sand boas «are able to tolerate wide extremes of temperature and can withstand
long periods of drought»; hibernation and/or estivation takes place in loose sand,
burrows, termitaria, crevices or beneath rocks. The genus occurs from sea level to
the high mountains; e.g., according to SEUFER (2001), E. miliaris and E. elegans are
known to reach about 1800 m and more than 2500 m a.s.l., respectively.
Very voracious animals, the sand boas dispatch their prey by means of con-
striction or squeezing against hard objects, but nestlings of small birds and mam-
mals may be swallowed alive; according to POLS (1986; Kenyan specimens of E.
colubrinus) relatively small prey are grasped very quickly and suffocated not by
constriction but by pulling them under the sand. Sand boas usually hunt by lying
in ambush under the sand, with only the eyes and nostrils above the surface, or in
any case remaining in a hidden position; but when the prey is within striking dis-
tance, they grasp it by projecting themselves from their hiding place. In any case,
according to RODRÍGUEZ-ROBLES et al. (1999), since nestling birds and rodents are
necessarily sedentary, erycines must encounter those prey by wide-foraging rather
than ambush. RODRÍGUEZ-ROBLES et al. (1999: 49), dealing with the diet of Eryci-
nae (a basal clade of macrostomatan snakes, i.e. snakes characterized by increased
mobility of the jaws and increased gape size) stated that «species of Eryx feed
mainly on mammalian prey, but also eat lizards and occasionally birds. Evolution-
arily more basal groups of snakes primarily feed on elongate prey, which suggests
that innovations of feeding apparatus of macrostomatans allowed these snake to
eat heavier and bulkier prey, particularly mammals. Erycines appeared and diversi-
ed at approximately the same geological time as rodents, suggesting that rodents
perhaps constituted an abundant prey resource that favoured the diversication of
early macrostomatans». Exceptionally Eryx feeds even on reptile eggs or inverte-
brates, mainly slugs (KLINGELHÖFFER 1959). The prey may also be relatively large;
e.g., J.N.B. BROWN found in the stomach of a 38 cm long E. jayakary a jerboa with a
total length of 28 cm (JONGBLOED 2000).
While generally sluggish, sand boas are quick to bite, seizing their prey in
sudden sidelong strikes. Towards man they are either docile or aggressive depend-
ing on the circumstances; when handled, their attack may result in deep lacerations
due to a very rapid series of lateral slashing bites; captive specimens are eventually
inclined to become very tame and even to accept food by hand. Although their bite
is innocuous, at least some species of Eryx (E. colubrinus included) have saliva con-
taining powerful poisons, though in minute quantities (CORKILL 1935).
The tail may be used to deect attention away from the head when the snake
is under threat (MATTISON 1995).
Sand boas reproduce by ovoviviparity; captive animals copulate in spring
and early summer; the male tries to dig the posterior half of the female out of the
sand prior to copulation; the female sometimes refuses all food during pregnancy;
4-17 young are born within 4 to 5 months, with a length varying approximately
between 120 and 290 mm according to the species; the rst slough takes place
between 7 and 16 days after birth (LAMERS 1984, VAN WOERKOM 1987, SHARP 1992,
TRACER 1996, SPAWLS et al. 2002). Little is known about the time necessary to
reach sexual maturity; KEND & KEND (1993), dealing with rough-scaled sand boas
(E. conicus), stated, with no further details and perhaps on the basis of informa-
tion gleaned from published reports, that, if fed well, they reach sexual maturity
in about two to three years, although it is possible for 18-month-old specimens
to reproduce (see also E. colubrinus, Biology). Some data on the growth rate of
72 B. Lanza and A. Nistri
73Somali Boidae and Pythonidae
captive E. colubrinus (see below), E. jaculus and E. johnii have been given by VAN
WOERKOM (1987).
Data on the rearing of sand boas, some of them including E. colubrinus, can
be found in KLINGELHÖFFER (1959), MATZ (1974), VAN WOERKOM (1981), MCLAIN
(1982), BARTLETT (1984), GRIFFITHS (1984), LAMERS (1984), DATHE (1985), KELLOUGH
(1986), POLS (1986), SHARP (1992), TRACER (1996), MEER (1997), and WALLS (1998b).
Key to Somali Eryx
1 Lowest number of scales across the head from eye to eye 9-13; scales around the
eye 11-16; nostril surrounded by 3 scales (internasal, prenasal and postnasal);
midbody scales 43-61 .............................................. E. colubrinus (Linnaeus 1758)
Lowest number of scales across the head from eye to eye 4-6; scales around
the eye 9-11; nostril surrounded by 2 scales (internasal-prenasal complex and
postnasal); midbody scales 34-40 ............................................................................... 2
2 156-163 ventrals (ventrals + pseudoventrals) .......... E. somalicus Scortecci 1939
Ventrals (ventrals + pseudoventrals) more numerous, 193 in the only known speci-
men .................................................................................................... E. borrii n. sp.
Eryx colubrinus (Linnaeus 1758) (Figs 1-7)
Anguis colubrina LINNAEUS 1758: 228 (terra typica: Egypt; type: presumed lost).
Eryx colubrinus; DAUDIN 1803b: 261.
Eryx thebaicus REUSS 1834: 134 (terra typica: Egypt; holotype: SMF 16937).
Eryx scutata GRAY 1842: 45 (terra typica: Egypt; holotype: BMNH 1946.1.7.98, formerly BMNH
E[ryx] jaculus var. sennaariensis JAN 1863: 21 [terra typica: Sennaar = Sannaˉ r, 13°33’N 33°38’E,
Sudan; type originally in the MSNM, lost when the museum was partly destroyed during
World War II (cf. KLUGE 1993a) (synonymy de BOULENGER 1893: 125) (see also JAN &
Eryx jaculus var. Surinamensis (ex errore for sennaariensis JAN 1863); PARENTI & PICAGLIA 1886: 61;
Assab (= Aseb, 13°00’N 42°45’E).
E[ryx] tebaicus [sic!]; LEPRI 1911: 320.
E[ryx] t[hebaicus] thebaicus; STULL 1932: 29.
Eryx thebaicus loveridgei STULL 1932: 29 pl. II, Fig. B [terra typica: “Mbunyi” (actually Mbuyuni),
Kenya; holotype: MCZ 18184].
Eryx colubrina; FLOWER 1933: 804.
Eryx rufescens AHL 1933: 324 (terra typica: Dadab, 11°N 43°15’E, Somalia; holotype ZMB 27391).
Eryx colubrinus colubrinus; STULL 1935: 406.
Eryx colubrinus loveridgei; STULL 1935: 406.
Gongylophis (Neogongylophis) colubrinus; TOKAR’ 1989: 54.
Derivatio nominis. The species name derives from the Latin colubrinus (mas-
culine adjective; feminine: colubrina; neuter: colubrinum): having the qualities of a
Ethnozoology. Eryx colubrinus is widely feared in Somalia (SCORTECCI 1955)
and Kenya, where, according to SPAWLS et al. (2002: 312), «it looks dangerous; […]
legend has it that if bitten, you take seven steps and die».
74 B. Lanza and A. Nistri
75Somali Boidae and Pythonidae
Fig. 1. Head and neck of Eryx colubrinus MZUF 5943 from Afmadù; dorsal, ventral and lateral
view (drawing by Manuela Mascherini).
74 B. Lanza and A. Nistri
75Somali Boidae and Pythonidae
Somali records. Afgoi (SCORTECCI 1939a: 267, 3 spec. MSNG, leg. U. Casale,
1910); Balad (SCORTECCI 1939a: 267, 2 spec. MSNG, leg. U. Casale, 1911): Bardera
(BOULENGER 1909b: 311, BMNH 1909.3.29.2 and 1 spec. MSNG 40800, leg. U.
Ferrandi, 1908; TESTI 1935: 107); Belet Amin (SCORTECCI 1939a: 267, 3 spec. MSNG,
leg. S. Patrizi, VI-VIII.1934). Berbera (LARGEN & RASMUSSEN 1993: 326, BMNH
1929.4.29.4); Biji (BOULENGER 1901: 49, Donaldson Smith Expedition, 1899);
Borama District (09°56’N 43°11’E, 1500 m) (LARGEN & RASMUSSEN 1993: 326, BMNH
1955.1.11.38, leg. Ingen); Brava (BOULENGER 1896a: 20, 1 spec. MSNG 29234,
Bottego’s Expedition, X.1895); Bulbar (LOVERIDGE 1936: 235, 1 spec. MCZ R 39010,
leg. D. Brockman); Burao (LARGEN & RASMUSSEN 1993: 326, BMNH 1970.1606);
Coastal zone of British Somaliland (the Guban) (PARKER 1949: 27, 1 spec.); Dadab
(STERNFELD 1908: 239, 1 spec., Erlanger and Neumann’s Expedition 1900-1901;
AHL 1933: 324, 1 spec., Erlanger and Neumann’s Expedition 1900-1901; LARGEN &
RASMUSSEN 1993: 326 and map 7 of p. 409); Dolo (BOULENGER 1912: 331, 2 spec.
MSNG 30002/1-2, leg. C. Citerni, V-VII.1911); Giohàr (SCORTECCI 1929: 269, unsexed
spec. MSNM 635, formerly 2035, leg. U. Fiechter, 23.IV.1929; SCORTECCI 1931: 203,
MSNM 647, formerly 2063, leg. U. Fiechter II.1930; SCORTECCI 1939a: 267, 3 spec.
MSNG 42706/1-2, 3rd spec. not found, leg. A. Andruzzi, 1923; 3 spec. MSNG, leg.
A. Andruzzi, III.1924; and 2 spec. MSNG 46352/1-2, leg. Paoli, 1926, or 1936?);
Guban and regions south of the Shebelli River (BALLETTO 1968: 255 and 267);
Jumbo (BOULENGER 1909a: 309, 2 spec. MSNG 29136 and 30003, leg. G. Ferrari,
VIII-IX.1908); Kismayu (LOVERIDGE 1916: 82, 1 spec. NMK/I. 129; LOVERIDGE 1936:
233, 4 unnumbered specimens NMK); Labadaad and Bulo Yaag (= Yaaq) (between)
(HOEVERS & JOHNSON 1982: 184, and Hoevers’ unpublished notes, 2 ¢¢ MZUF
27046-27047, eld numbers LGH 609 and PMJ 106, respectively, leg. L.G. Hoevers,
X.1977, and P.M. Johnson, 31.X.1977; 2 ™™ CAS 153326-153327, eld numbers LGH
Fig. 2. — Midbody view of Eryx colubrinus MZUF 5943 from Afmadù (drawing by Manuela
76 B. Lanza and A. Nistri
77Somali Boidae and Pythonidae
571 and LGH 572, leg. L.G. Hoevers, XI.1977 and I.1979, respectively; 1 CAS
153328, eld number H 0934, collected at Hargesa (= Hargeysa: 00°27’N 42°43’E),
leg. L.G. Hoevers and P.M. Johnson, 26.V.1980); Meleden (northern Somalia)
(SCORTECCI 1955: 647, 2 spec., one of them is the MZUF 2589, leg. G. Scortecci,
1953). Middle Shebelli River (LIPPARONI 1954: 655); Mogadishu (environs of) (LEPRI
1911: 320 and 328, 1 young spec. MZUR, leg. U. Casale; SCORTECCI 1931: 203, ¢
MSNM 643, formerly 2062/A, and ? MSNM 644, formerly 2062/B, leg. Luppi 1930;
SCORTECCI 1939c: 128, 2 unsexed spec., Missione Biologica nel Paese dei Borana, E.
Zavattari, 25.II.1937); Northern (regions West Galbeed, Togdheer, and Bari), central
(regions Mudug and Galgadud), and southern (regions Middle Shebelli, Mogadishu,
Bakool, Bay, Lower Shebelli, Gedo, Middle Juba, and Lower Juba) Somalia (LANZA
1983: 220; LANZA 1990: 431); Obbia and Magangib (between) (CALABRESI 1927: 31, ,
Stefanini and Puccioni’s Expedition 1924); Somalia (MOCQUARD 1888: 122, 2 spec.
MNHN, Révoil’s Expedition 1882-1883; SCORTECCI 1934a: 337 and unnumbered
g. of p. 338; SCORTECCI 1934b: 53 and unnumbered g. of p. 54.; TESTI 1935: 115;
SCORTECCI 1936: 923; SCORTECCI 1937: 265, 1 unnumbered g.; SCORTECCI 1939b: 103
and 104, g 53; BALLETTO 1968: 214; WALLS 1998b: 163, unnumbered gs of pp. 164-
165, 167-170, map of p. 166).
Material studied. In total, of 79 specimens have been examined from the following local-
ities: Afgoi ( MZUF 1059, donated by A. Sammicheli, XII.1957; MZUF 22257, leg. A.M.
Fig. 3. Eryx colubrinus MZUF 22257 from 6 km E of Afgoi (photo by Benedetto Lanza).
76 B. Lanza and A. Nistri
77Somali Boidae and Pythonidae
Simonetta, IV.1978); Afgoi (ca 6 km E of) (2 ¢¢ MZUF 24194-24195 and 24196, leg. A.M.
Simonetta, 13.III.1979, summer 1979, and III.1979, respectively); Afgoi (environs of) (2 ¢¢
MZUF 1084 and 1086, leg. A. Sammicheli, XI-XII.1959; 2 ¢¢ MZUF 1085 and 1087, leg. SBS,
a native, 5.IX.1959 and 25.IX.1959, respectively); Afmadù ( MSNM 805, formerly 2112, leg.
G. Scortecci, 1931; ? MSNM 634, formerly 2117, leg. L. Boschis, V.1935; 2 ¢¢ MZUF 5944-
5945 and 4 ™™ MZUF 5943, 5946-5948, leg. SBS, VIII.1970); Afma (ca 15 km W, close to
the Lak Dere) (¢ MZUF 2021, leg. SBS, 10.VIII.1962); Alessandra Island (¢ MZUF 2190 and
MZUF 2108, leg. SBS, 13-14.VIII.1962; MZUF 2238, donated by the Catholic Mission, 1962);
Balàd ( MZUF 1060, leg. U. Funaioli, 6.IX.1954; MZUF 1083, donated by U. Funaioli,
VII.1957). Bender Cassìm (? MSNM 641, formerly 2114, leg. I. Zanetti, I.1932; MSNM 642,
formerly 2113, leg. G. Scortecci, 12.IX.1931); Bud Bud ( MZUF 5293, leg. SBS, 12.VIII.1968);
Dinsòr (2 ¢¢ MZUF 1735 and 2245, leg. SBS, 25-26.VII.1962); Dusa Marèb (ca 30 km NE) (
MZUF 23789, leg. A.M. Simonetta and J. Simonetta, XII.1978); Genale (¢ MZUF 1090 and
2 ™™ MZUF 1088-1089, donated by G. Esposito, 1959); Geribàn ( MZUF 39700, leg. SBS,
13.VIII.1964). Giohàr ( MSNM 647, formerly 2063, leg. U. Fiechter, II.1930; 2 MSNM 718-
719, formerly collective number 2115, leg. U. Fiechter, VII-VIII.1933; unsexed spec. MSNM
635, formerly 2035, leg. U. Fiechter, 23.IV.1929); Giohàr (10 km NW) ( MZUF 2237, leg. SBS,
A.M. Simonetta, 19.VIII.1962); Giohàr (8 km SW) ( MZUF 5374, leg. SBS, 12.IX.1968); Jam-
ame ( MZUF 3945, leg. SBS, 1964); Jilib (2 ¢¢ MZUF 21641, 22011 and 2 ™™ MZUF 22008,
27652, donated by G. Tarabini Castellani, 1970, No. 21641, 1977, No. 22008, 22011, and
1980, No. 27652); Kismayu (¢ MZUF 6934, donated by Todaro and Lapucci, 1971); Labadaad
(00°30’N 42°46’E) and Bulo Yaag (00°12’N 42°46’E) (= Yaaq) (between) (2 ¢¢ MZUF 27046-
27047, eld numbers LGH 609 and PMJ 106, respectively, leg. L.G. Hoevers, X.1977, and P.M.
Johnson, 31.X.1977); Lak Badanà (ca 10 km WSW of Mangab) (¢ MZUF 3959, leg. Lucano,
9.IX.1964); Mahaddei Wen and Avai (between) (¢ MZUF 21965 and 2 ™™ 21881-21882, leg.
Fagotto, ca 1975); Meledèn ( MZUF 2589, leg. G. Scortecci, 1953); Mogadishu (unsexed spec.
MSNM 648, leg. Onorio Govinidi?, undecipherable, 1936; 3 ¢¢ MZUF 2804-2806 and 1
MZUF 2803, donated by B. Lanza); Mogadishu (environs) (¢ MSNM 643 and ? MSNM 644,
Fig. 4. Eryx colubrinus ¢ MZUF 28538 from Xayo (ca 80 km SW of Afmadù) (photo by Marco
78 B. Lanza and A. Nistri
79Somali Boidae and Pythonidae
formerly collective number 2062, leg. Luppi 1930; ¢ MSNM 645 and ¢? MSNM 646, formerly
collective number 2116, leg. Confalonieri, 1935; 3 ¢¢ MSNM 636, 638-639 and unsexed spec.
MSNM 637, formerly collective number 2111, leg. G. Scortecci, X.1931); Obbia and Magangìb
(between) (¢ MZUF 979, leg. G. Stefanini and N. Puccioni, 1924); Oddùr (¢ MZUF 23759,
MZUF 23760 and unsexed spec. MZUF 23758, leg. Marescotti, 1956, donated by the Istituto di
Zoologia, University of Genoa). Scialambòd (1 ? MSNM 640, ex 2110, leg. Urbinati brothers,
1930); Somalia (¢ MZUF 33637, leg. SBS, 1964?). Tessenei (extralimital: Eritrea) (2 ™™ MSNM
632-633, formerly collective number 2217, leg. A. Remedelli, donated by the Istituto Sierotera-
pico Milanese, summer 1937). Uagadi (¢ MSNM 631 and 2 ™™? MSNM 629-630, formerly col-
lective number 2109, leg. Urbinati brothers, 1930). Warmahàn ( MZUF 24197, leg. A.M. Sim-
onetta, 27.I.1979); Warshèk (environs of) (¢ MZUF 27529, leg. A.M. Simonetta, 5.IX.1983);
Xayo (ca 80 km SW of Afmadù) (¢ MZUF 28538, leg. SBS, Marco Borri, 30.X.1983).
Description (Figs 1-6; see also Geographical variability).
Morphology. Dorsal head scales small and smooth, imbricate, only the nasals
and internasals enlarged; rostral large and broad, about 2.5 times wider than high,
with a labial angular horizontal edge; nostril surrounded by 3 scales (a large inter-
nasal, a medium-sized prenasal and a slightly smaller postnasal); 9-15 interorbital
(usually 10-12) and 11-16 (usually 13-14) circumorbital scales; eye separated from
the labials by 1-3 rows of scales, usually 2; 4-7 scales between the nasal and the
eye, usually 5; 11-15 supralabials, usually 13-14; approximately 19-21 small lower
labials; 13-23 scales between the mental and the rst pseudoventral, usually 15-
17; mental groove absent or, according to BOULENGER (1893) and ANDERSON (1898),
rarely present (never seen by us). Body scales smooth or more or less keeled ante-
Fig. 5. — Eryx colubrinus MZUF 2108 from Alessandra Island; ventral view (photo by Benedetto
78 B. Lanza and A. Nistri
79Somali Boidae and Pythonidae
riorly, becoming increasingly keeled distad, strongly carinate on the posterior third
to fth of the trunk; supracaudal scales strongly keeled; from 43 (39 according to
MOCQUARD 1888, and CHIPPAUX 2001) to 61 midbody scales; ventrals (wider than
long system) from 162-(159 according to MOCQUARD 1888) to 205; anal entire, very
rarely divided; subcaudals 19-30, normally single, sometimes some of them paired
(each pair counted as one subcaudal); tail short (usually 7-9% of total length), coni-
cal, more or less pointed. Total length approaching 1 m in the females (BARKER &
BARKER 2002) [800 mm (745 + 55 mm) in a female from Hargesa, southern Somalia
(HOEVERS & JOHNSON 1982; Hoevers’ unpublished notes)]. The males, more slender
and considerably smaller than females, have a relatively longer tail (about 8-11% of
total length; ™™ 7-10) and longer spurs which become progressively larger with age;
according to WALLS (1998b), in some females the spurs may be completely hidden
under scales and hard to nd.
The hemipenis, studied by BRANCH (1986: 292-293) in two specimens from
Ethiopia and Kenya, has been described as follows: «Eryx colubrinus colubrinus
(Linnaeus). Voucher specimen: BMNH 1902.12.12.70, Billen, Ethiopia. Description:
Retracted, simple, left hemipenis extends to 12th subcaudal, sulcus dividing at 5th.
Sulcal forks diverge to lateral aspects of organ, which is undivided, although retrac-
tor muscle divided for 1-2 subcaudals, each branch inserting upon organ in region
of ipsilateral sulcal termination. Ornamentation: Flounced; deep, smooth-walled
Fig. 6. — Eryx colubrinus ¢ MZUF 1735 from Dinsòr, killed by its prey, a Quelea quelea (see text),
preserved specimen (photo by Benedetto Lanza).
80 B. Lanza and A. Nistri
81Somali Boidae and Pythonidae
Fig. 7. — Somali distribution of Eryx colubrinus. 1, Dadab; 2, Bender Cassìm; 3, Bulbar; 4, Ber-
bera; 5, Meledèn; 6, Biji; 7: Borama district; 8, Burao; 9, Geribàn; 10, Dusa Marèb; 11, Obbia;
12, Magangìb; 13, Dolo; 14, Oddùr; 15, Bud Bud; 16, Mahaddei Uen; 17, Giohàr; 18, Bardera; 19,
Dinsòr; 20, Warmahàn; 21, Balàd; 22, Warshèk; 23, Afgoi; 24, Mogadishu; 25, Genale; 26, Uagadi;
27, Scialambòd; 28, Avai; 29, Brava; 30, Afmadù; 31, Xayo; 32, Labadaad (= Alessandra Island); 33,
Jilib; 34, Hargesa; 35, Xayo; 36, Bulo Yaag; 37, Belet Amin; 38, Jamame; 39, Jumbo; 40, Kismayu;
41, Lak Badanà.
80 B. Lanza and A. Nistri
81Somali Boidae and Pythonidae
calyces cover organ distally, becoming shallower and narrower proximally, form-
ing shallow, scalloped ounces encircling organ at level of sulcal furcation; basal
section short, nude. Along sulcal fold calyces fused forming two large longitudinal
folds, that in turn form outer folds of sulcal forks, possibly facet-like in everted
organ. Eryx colubrinus loveridgei (‘rufescens phase) Stull. Voucher specimen: BMNH
1952.1.8.89, Muri Hills (2700 m), Northern Kenya. Description: Retracted left hemi-
penis undivided, extending to 11th subcaudal; sulcus dividing at 3rd subcaudal,
forks diverging to lateral aspects of organ. Ornamentation: Flounced; similar to typ-
ical colubrinus, with numerous ne smooth-walled calyces, coalescing proximally to
form 5-6 transverse ounces at, and slightly distad to, sulcal furcation».
About 12-15 maxillary and mandibular teeth, solid, ungrooved, arranged in
an anonodont and scaphiodont set; 5 palatine teeth, of which 3 functional (RIEPPEL
1978: 186 and TOKAR 1989: 52, drawing 22 of g. 5; p. 53, footnote; personal obser-
vations); 6-8 pterygoid teeth (RIEPPEL 1978: 187, TOKAR 1989: 51), of which only 3
are sometimes functional (personal observations).
ALEXANDER & GANS (1966) (n = 2) found ca 182-184 (ventral) and 270-277
(dorsal) body scale rows, ca 26-28 (ventral) and ca 26-28 (dorsal) tail scale rows,
187-190 body vertebrae and 28-29 tail vertebrae (see also the description of Eryx).
For the Somali specimens and two Eritrean females studied by us see Table 1
and Appendix 1.
It is noteworthy that sexual dimorphism signicantly affects not only the
relative tail length (TaL%), the relative spur length (SL), the ventral (V and VD) and
subcaudal (SC) counts, as well as the number of midbody scales (MSR), but also
some characters that seem to be devoid of any functional signicance: number of
interorbitals (IO), circumorbitals (CO), scales between eye and supralabials (OS),
and gulars (GL). TOKAR (1996) studied the following pholidosis characters: TaL%,
V, SC, CO, UL, IO (but excluding the upper circumorbitals), number of postinter-
nasals (in the row of scales immediately behind the internasals and between their
external posterior corners), number of loreals (in the row bordering the upper labi-
als) between nasal and oculars, number of scales in one midhead longitudinal row
from the level of postoculars to postinternasals, and ordinal number (counted from
rostral) of the highest upper labial. Dealing with individuals from the almost entire
range of the species, he found sexual dimorphism only in TaL%, MSR, V and SC.
The greatest total lengths so far known for Somali specimens are 610 mm [¢
MSNM 636 (ex-collective number 2111)] and 800 mm [ held in CAS (HOEVERS &
JOHNSON 1982); see above].
For the negative allometric growth of the eye see Table 2.
Some specimens obtained by VAN DER POLS (1986) and VAN WOERKOM (1987)
weighed 8.1-8.9 g at birth, 65-140 g at the age 0.5 months, and 8.5 g at birth, 25-35
g at the age of 2.5 months, respectively. According to PITMAN (1938), an Egyptian
specimen with a total length of 770 mm weighed 0.6 kg.
Colouration. Fundamentally the species is represented by two phenotypes:
colubrinus, the most widespread, and sennaariensis (of which rufescens is a junior
synonym), usually occurring sporadically in the same localities as colubrinus, but
possibly dominating elsewhere. In the phenotype sennaariensis the dorsal coloura-
tion, uniform or nearly uniform, is light brown (as in the type of E. rufescens, a
preserved specimen; AHL 1933), purplish-brown [ANDERSON 1898, specimens from
Suakin (Sudan; sex not indicated); young MZUF 5293, from Bud Bud, Soma-
lia] or dark brown; lower scale rows and ventrals whitish or yellowish colour to
82 B. Lanza and A. Nistri
83Somali Boidae and Pythonidae
Table 1.
Eryx colubrinus: characters of the Somali specimens studied.
Character Sex n Range Mean cl (99%) t df P
TaL% ¢¢ 31 8.21-11.14 9.64±0.124 9.29-9.98 9.00 57 <0.01
TaL% ™™ 28 6.78-9.84 7.91±0.149 7.49-8.32
SL ¢¢ 18 20.63-50.00 32.52±2.022 26.66-38.38 6.35 30 <0.01
SL ™™ 14 25.86-158.06 87.19±9.444 58.74-115.64
GL ¢¢ 34 14-20 15.97±0.255 15.31-16.63 3.22 65 <0.01
GL ™™ 33 13-23 17.39±0.364 16.46-18.33
MSR ¢¢ 34 47-57 50.77±0.455 49.59-51.94 6.14 63 <0.01
MSR ™™ 31 46-61 55.36±0.604 53.69-57.02
V¢¢ 33 168-190 174.97±0.808 172.89-177.05 5.94 62 <0.01
V™™ 31 170-197 182.58±1.004 179.82-185.34
VD ¢¢ 32 166-188 172.72±0.815 170.62-174.82 6.02 61 <0.01
VD ™™ 31 167-194 180.36±0.976 177.67-183.04
SC ¢¢ 33 22-28 25.73±0.290 24.98-26.47 4.53 63 <0.01
SC ™™ 32 20-30 23.47±0.412 22.41-24.53
Character Sex n Range Mean cl (99%) G df P
IO ¢¢ 33 9-12 10.36±0.143 9.99-10.73 19.46 4 <0.01
IO ™™ 34 9-13 11.29±0.143 10.92-11.66
CO left ¢¢ 33 11-14 12.79±0.161 12.35-13.23 18.02 5 <0.01
CO left ™™ 34 12-16 13.71±0.182 13.21-14.20
CO right ¢¢ 34 11-15 12.85±0.175 12.37-13.33 21.50 5 <0.01
CO right ™™ 34 12-16 13.91±0.208 13.34-14.48
OS left ¢¢ 32 2-3 2.28±0.081 2.06-2.50 7.15 1 <0.01
OS left ™™ 31 2-3 2.61±0.089 2.37-2.86
OS right ¢¢ 33 1-3 2.24±0.087 2.00-2.48 6.04 2 <0.05
OS right ™™ 31 2-3 2.55±0.091 2.30-2.80
NO left ¢¢ 34 4-6 5.06±0.084 4.83-5.29 6.91 3 NS
NO left ™™ 34 4-7 5.41±0.113 5.10-5.72
NO right ¢¢ 34 4-6 5.24±0.085 5.01-5.47 2.18 3 NS
NO right ™™ 34 4-7 5.38±0.104 5.10-5.67
UL left ¢¢ 32 11-14 13.19±0.145 12.79-13.59 15.29 4 <0.01
UL left ™™ 29 12-15 13.76±0.177 13.27-14.25
UL right ¢¢ 33 11-15 13.27±0.152 12.85-13.69 3.84 4 NS
UL right ™™ 30 12-15 13.63±0.140 13.24-14.02
Character Sex n Range Mean cl (99%)
RR ¢¢+™™ 24 2.09-2.92 2.45±0.044 2.32-2.57
NS ¢¢+™™ 11 43-53 47.55±0.790 45.04-50.05
PC ¢¢+™™ 12 25-32 28.50±0.571 26.73-30.27
PV ¢¢+™™ 70 0-7 2.33±0.170 1.89-2.77
H¢¢+™™ 27 9-14 10.82±0.245 10.13-11.50
82 B. Lanza and A. Nistri
83Somali Boidae and Pythonidae
yellow; a ¢ from Dinsor (MZUF 1735) that the senior author saw alive was light
reddish-ochrous dorsally and whitish ventrally; iris was reddish-brown. The pheno-
type colubrinus has: dorsal parts with a grey, light yellow, yellow (sometimes with
apricot nuances), orange or red-orange ground colour (seemingly depending on the
local soil colour), which is more or less largely obscured or almost completely over-
whelmed by irregular, roundish or more or less transverse, grey, purplish-brown to
blackish-brown or black blotches, frequently conuent on the mesial line of back,
and prolonged onto the sides, where there are often small spots of the same colour;
in the animals with a prevalently dark back, the light ground colour persists on
body and tail as small, irregular spots and about 30 continuous or subcontinuous,
mostly transverse stripes, occupying only one side of the body or completely cross-
ing it; the latter are ca 2-4 scales long and 15-20 scales wide; underparts off-white
to light yellow, immaculate or, at least in some specimens from S Somalia, with a
few, sparsely arranged grey to light brown spots, in addition to which there may be
some similarly-coloured transverse bars, each one entirely or nearly entirely affect-
ing a gastrostege; usually a dark, generally ill-dened, oblique streak runs through
the eye from the loreal to the temporal region. TOKAR (1996) distinguished three
colour patterns: 0 “standard” (large rounded or oval dark-brown blotches scattered
on yellowish background colour), 1 “intermediate” (large rounded or oval dark-
brown blotches joined to leave a pattern of scattered yellowish stripes on dark back-
ground), and 2 rufescens[actually sennaariensis”] (back and body sides uniform
dark- or red-brown, only lower scale rows on sides and ventrals are yellow); he
found that “standard” and “intermediate” patterns are distributed between the sexes
in approximately equal proportions, while the “rufescenspattern is more character-
istic of males, and that the rufescenspattern occurs throughout the range of the
species in males, while it is completely absent in females from the northernmost
and southernmost parts of the range. The southernmost rufescensSomali speci-
men seen by us is the ¢ MZUF 1735 from Dinsòr (02°28’N 43°00’E). Amelanistic
[i.e. albino: DOUGLAS 1993] and anerythristic variants have recently become avail-
able to herpetoculturists (TRACER 1996). More specic information on these mutants
has been given by BARKER & BARKER (2002: 2): «In early 1995 a female albino East
African sandboa was imported into the United States from Tanzania. So far as we
know, this is the rst record of albinism in this taxon.» and «[…] unexpected axan-
thic [i.e. anerythristic: DOUGLAS 1993] babies were born to normal adults. Axanthic
snakes are missing yellow, red, and brown pigments. Axanthic baby Kenyan sand
boas are black with bluish-white pale markings, they grow to become black and
white adults with black button eyes. This mutation has been demonstrated to be an
inheritable trait, apparently a simple recessive mutation». Fine colour illustrations
Abbreviations: cl = condence limits; df = degrees of freedom; CO = number of circumorbital scales;
GL = number of gulars between mental and rst pseudoventral; H = subcaudal reached by the
retracted hemipenis; IO = lowest number of interorbital scales (including the upper circumorbitals);
MSR = numbers of longitudinal rows of scales at midbody; NO = lowest number of scales between
postnasal and eye; NS = number of longitudinal rows of scales behind the occiput; OS = lowest
number of scales between eye and upper labials; PC = number of longitudinal rows of scales a head
length anterior to the cloaca; PV = number of pseudoventrals; RR = rostral width/rostral depth; SC
= number of subcaudals; SL= tail length/spur length; TaL% = tail length × 100/total length; UL =
upper labials, supralabials; V = number of ventrals according to the “wider than long system”, i.e. of
pseudoventrals + true ventrals; VD = number of ventrals according to the “Dowling system”, i.e. of
the true ventrals.
84 B. Lanza and A. Nistri
85Somali Boidae and Pythonidae
of the species may be found in ANDERSON (1898), PITMAN (1938), KUNDERT (1974),
DATHE (1985), STAFFORD (1986), POLS (1986), COBORN (1991), SHINE (1992, 1998),
MATTISON (1995), TRACER (1996), WALLS (1998b; also colour mutants), CHIPPAUX
(2001), and SPAWLS et al. (2002).
Distribution (Fig. 7). Northern Africa from Niger (Air) east to Egypt, including
Sudan, Ethiopia, Eritrea, Somalia, and, according to BROADLEY & HOWELL (1991), arid
areas of Kenya and northeastern Tanzania (Msembe, near Ruaha National Park, and
Kahe); about the erroneous records from Morocco and Yemen see ANDERSON (1898:
235, footnote 1) and LANZA (1966: 37, footnote 10; 1983: 233, footnote 1), respectively.
Biogeography. PARKER (1949) hypothesized that E. colubrinus might be a
recent entrant into Somalia, which penetrated the region from the north and from
Table 2.
Negative allometric growth of the eye in Eryx colubrinus.
4.40 1.80 2.44
¢4.50 1.70 2.65
4.50 1.85 2.43
¢4.60 2.35 1.96
4.80 1.90 2.53
¢4.90 2.00 2.45
? 4.90 1.60 3.06
5.00 2.00 2.50
¢5.50 1.80 3.06
? 5.95 2.10 2.83
6.80 2.40 2.83
¢7.00 2.60 2.69
¢7.40 2.30 3.18
¢7.50 2.50 3.00
¢7.60 2.34 3.25
8.00 2.50 3.20
8.30 2.50 3.32
¢8.40 2.90 2.90
8.40 2.40 3.50
8.60 2.80 3.07
9.10 2.30 3.96
¢9.20 2.95 3.12
9.70 2.80 3.46
9.70 2.40 4.04
11.50 2.50 4.60
12.20 2.95 4.14
Measures in mm; SnL= snout length, i.e. distance between snout tip and eye; EHD = horizontal eye
84 B. Lanza and A. Nistri
85Somali Boidae and Pythonidae
the south; TOKAR (1996) disagreed with Parker and, on the basis of an elaborate sta-
tistical research, stated: «I think that Ethiopia was the origin and expansion centre
of G. colubrinus in Africa. As the land rose, this species spread to the coastal plains
of Somalia and later penetrated Kenya, the Sudan and Egypt».
Geographical variability (see also Description). STULL (1932) separated east-
ern African specimens of this species under the name Eryx thebaicus loveridgei,
distinguishing them from those of northern Africa in having a higher number of
midbody scales (53-59 instead of 47-49), a lower average number of ventrals (168-
182, average 173.2), instead of 175-192 (average 184.8), and immaculate belly and
sides. In 1933 AHL, overlooking Eryx jaculus var. sennaariensis Jan 1863 (see above),
described the new species Eryx rufescens.
Three years later LOVERIDGE (1936: 235) conrmed the occurrence in loveridgei
of only a lower average number of ventrals, reduced rufescens to the subspecies
level and arranged the following tentative key: «dorsal coloring uniform, ventrals
181-194, average (2 ex.) 187 = E. c. rufescens; dorsal coloring consisting of heavy
blotches ventrals 175-197, average (36 ex.) 186.3 = E. c. colubrinus; ventrals 162-
182, average (24 ex.) 170.8 = E. c. loveridgei».
Later SCORTECCI (1939a) questioned the validity of rufescens and supported
loveridgei, giving the following gures: 47-49 midbody scales and 175-197 ventrals
in colubrinus, 50-50 and 162-185 in loveridgei, 44-50 and 181-194 in rufescens.
PARKER (1949), dealing with Somali snakes, wrote that «The material available
to the present author is insufcient to throw fresh light on the status of loveridgei
and the single available specimen from Guban (rufescens) can be matched exactly
amongst specimens from Egypt. It is possible that Eryx colubrinus is a recent
entrant into Somaliland from both the north and the south and that further col-
lecting may show that a uniform reddish colouring which appears sporadically in
Egypt, is present in a sufcient majority of the population along the northern inva-
sion route to justify the recognition of a local race under the name rufescens» [actu-
ally sennaariensis; see above].
TOKAR (1996), studying morphometric indices, pholidosis and colour pattern
of 178 specimens (98 ¢¢, 80 ™™) from almost the entire range of the species and
using elaborate statistical methods, conrmed the monotypy of the species, already
supported by WELCH (1982), LANZA (1983, 1990), KLUGE (1993a) and LARGEN & RAS-
MUSSEN (1993), and pointed out «the lack of hiatus in any part of the species range
and that variance has a strong radial trend with a slight tendency towards a south-
east-northwest cline».
Biology (see also under Eryx, Biology). The species, usually associated with arid
and semi-arid habitats, has been recorded at altitudes between sea level and about
1300 m (in Ethiopia; LARGEN & RASMUSSEN 1993) and 1500 m (SPAWLS et al. 2002). In
Somalia it has been found in many different ecosystems which, according to the geo-
botanical map published by PICHI-SERMOLLI (1957), are characterized by the following
types of vegetation: grass, perennial herb and subshrub steppe; shrub steppe, subdes-
ert scrub; broken xerophilous open woodland; xerophilous open woodland; savanna;
lowland dry evergreen forest; coastal formations; riparian formations.
Its habits in nature are poorly known, but are probably very similar to those
of its congeners (see also Eryx, Biology). Eryx colubrinus is a burrowing, semi-
nocturnal species usually living by day in different types of vacuities (crevices,
holes, termite-hills, etc.) or buried in sand, from which it emerges at night for
86 B. Lanza and A. Nistri
87Somali Boidae and Pythonidae
hunting or occasionally by day, especially in the early morning, for hunting or
basking; underground it probably hunts both at night and by day. «In areas where
the soil is mostly hard (such as the low eastern foothills of Mt Kenya) it will live in
sandy riverbeds, which must put it at risk in ash oods» (SPAWLS et al. 2002: 311-
312). It may shelter, partially buried, under rocks, bark, logs, or dumped objects,
sometimes close or inside huts (as, f.i., MZUF 28538). According to SPAWLS et al.
(2002: 311), the strongly keeled scales on the tail and the nal fth of the body
«may aid locomotion or have a defensive purpose» (see also Eryx, Biology).
We know the following dietary records for E. colubrinus (see also PITMAN
1958: 104, whose citation of CANSDALE (1955: 23) is incorrect since the latter refers
to E. muelleri, not to E. colubrinus): mice and small birds (LOVERIDGE 1923); small
birds, lizards and young rodents (CORKILL 1935); small rodents and birds up to the
size of a quail (PITMAN 1938: 63); Charadrius asiaticus Pallas 1773 (Charadriidae)
[LOVERIDGE 1956: 142, writes: «The late Blayney Percival once told me he had sur-
prised a (sand) boa at Voi trying to swallow a Caspian Plover (Charadrius asiaticus),
but the migrant proved too big a mouthful and after a couple of futile attempts the
snake gave it up»]; Motacilla aguimp Dumont 1821 (Motacillidae) and other uniden-
tied birds (PITMAN 1958); mice and Scincidae of the genus Mabuya Fitzinger 1826
[= Trachylepis Tschudi 1845] (HOEVERS & JOHNSON 1982; captive specimen from
southern Somalia); Stenodactylus Fitzinger 1826 sp. (Gekkonidae; GASPERETTI 1988);
the burrowing Bathyergidae naked mole-rat Heterocephalus glaber Rüppell 1842
(BRAUDE 1991); SPAWLS et al. (2002: 312) state that «The juveniles eat mostly lizards
(especially Heliobolus Fitzinger 1843 and Latastia Bedriaga 1884), large adults take
rodents as well, and even birds, indicating how well they are concealed and how
rapidly they strike»; Fig. 6 shows a ¢ 358 mm long, from Dinsòr (MZUF 1735),
found dying by a native, with the body walls pierced at the esophageal level by the
strong bill of a red-billed weaver [Quelea quelea (Linnaeus 1758); total length ca
104 mm, tail length ca 32 mm)], of which the snake had swallowed the head (LANZA
& BARBAGLI in preparation). The juveniles obtained by SCHWEIZER (1972), LAMERS
(1984), POLS (1986), SHARP (1992), and TRACER (1996) accepted, only a few days
after birth, newborn naked mice for food. SHARP (1992), who obtained 10 babies in
captivity, stated that only two of them constricted their food.
According to LOVERIDGE (in litt., quoted by PITMAN 1938: 64) «Though capa-
ble of inicting a painful bite, this sluggish reptile relies for its protection on
the appalling odour which it discharges from its cloaca when molested» (cf. also
SPAWLS et al. 2002).
According to TRACER (1996) E. colubrinus is a highly seasonal breeder, seem-
ing-ly not requiring the pre-breeding cooling period so vital to other boids, and
relatively easy to breed in captivity. The number of neonates per delivery ranges
between 4 and 20 (SPAWLS et al. 2002). LOVERIDGE (1936: 234) reported that at
Voi, Kenya, on April 24, 1934, «a native brought in a female and her seven young
[evidently her recently born litter, according to FITCH 1970] which he had found
altogether in a hole. The length of the mother is given above [634 (584 + 47) mm],
the young ranged from 176 to 189 mm»; SCHWEIZER (1972) obtained some deliver-
ies, each of 10-15 offspring, in captivity; the newborns were 20 cm long. LAMERS
(1984) bred a female about 18 months old with a male about 4 years old (living
together from 7.VI.1983) in a terrarium measuring 100 × 50 × 50 cm, with a layer
of sand about 15 cm deep, and at a temperature of 23-36 °C. He observed copula-
tions between 8.VI.1983 and 10.VII.1983 and obtained 17 offspring on 25.X.1983,
measuring about 19 cm and with an average weight of 8 g; the rst sloughing was
86 B. Lanza and A. Nistri
87Somali Boidae and Pythonidae
after 9 days. POLS (1986: 209-211), dealing with three captive specimens received on
8 March 1983 from Kenya (a ¢ of 40 g, two ™™ of 48 and 59 g), stated: «Early in
April 1984, I observed the male’s rst attempts at copulation. Copulation poses an
interesting problem for a burrowing species such as the Kenyan sandboa. Since the
female is usually always buried underneath a layer of sand, the male must some-
how dig her out in order to mate with her. The male accomplishes this by digging
his head under the (much larger) female, coming up on the other side, and then
lifting the female out of the sand with the fore-part of his body. He then throws the
rst part of his body over the female, making a loop around her. Once in this posi-
tion, the male will slide the loop down the female’s body, moving the position of the
loop down his own body at the same time, until their vents are opposite each other
and he is able to work her with his spurs. In spite of all this, the male did not seem
to be having any success at these rst attempts. The females would only try to bury
themselves deeper in the sand. These mating attempts lasted from about noon until
the next morning, but I do not think that the male had any success at all during
that period because the females just did not seem to want to co-operate. Finally I
decided to put the male back into his own cage. Strangely enough, he did not begin
to eat until August; and then only irregularly. Early in June 1985, I put the male
in with the females again. He began his copulation attempts within a few days. It
seemed that the orange female was more co-operative than the yellow female as on
2 July 1985, I observed an actual copulation with the former. I have not observed
any copulations involving the yellow female in spite of the interest the male has
shown. It is my opinion that this mating behaviour was induced by a lengthened
photo-period and a rise in temperature. [...] The orange female started to refuse all
food on 17 August 1985. The second half of her body had begun to visibly swell.
She also sun-bathed more often. At rst she lay only in the vicinity of the bulb
from the beginning of the day on, but as the state of her pregnancy progressed, she
began to lie more directly under the bulb. Maybe it would have been better if I had
heated the sand from underneath; it just seems to me more natural for a sandboa
to soak up heat while buried in the sand than when lying on the surface. [...] On
4 December 1985, the orange female gave birth to fourteen young snakes, two of
which were still-born. When I discovered them, at about six o’clock in the evening,
most of the juveniles and their mother (who was behaving restlessly) still lay on the
surface of the sand. One hour later they had all burrowed out of the sight. The next
day, I separated the young from their mother. Of the twelve, six were yellow and
six were orange. One yellow and one orange juvenile had died before birth. Their
birth-weights varied from 8.1 to 8.9 g. Within two weeks they had all sloughed
and had all started to feed on new-born (“pinky”) mice. The weight of the orange
female after her pregnancy had dropped from about 500 g to 360 g. Eight of the
young snakes were sold shortly after they had started to eat. The remaining four I
have kept and these are developing well. On 20 October 1986, they weighed 65, 80,
95 and 140 g». WALLS (1998b), whose mating description agrees with that by POLS
(1986), stated that in some cases «the short tails of the couple end up projecting
vertically through the surface of the sand». SHARP (1992), breeding a female about
40 cm long with a male about 30 cm long in a terrarium measuring 90 × 30 × 38
cm, with a substrate of silver sand of a depth of 7.5 cm, observed several attempted
matings from mid-January 1991 to early February; no actual matings were wit-
nessed, but the female began to increase in size in late February; the birth of her
10 babies, weighing 7 g each, was observed on 27 June 1991 between 07:00 pm and
09:30 pm; all of the young had sloughed within 7 days. According to MEER (1997:
88 B. Lanza and A. Nistri
89Somali Boidae and Pythonidae
14) E. colubrinus copulates in June and July and after ca 4 months the female gives
birth to 10-13 young about 20 cm long; «in the wild, females have a litter every two
years, but in captivity they may produce offspring every year»; he kept his animals
in a terrarium measuring 70 × 60 × 35 cm, with a layer of sand about 6 cm deep,
and at a temperature of 22-35 °C. A Ngomeni (Kenya) specimen had 8 young and
7 infertile eggs in mid-November (SPAWLS et al. 2002). The neonates’ length ranges
between 15 (SPAWLS et al. 2002) and 20 cm. According to TRACER (1996: 48), one of
the only references to age at sexual maturity in E. colubrinus «was made by ROSS &
MARZEC […] (1990): “Although sexual maturity seems to be reached in about three
to four years in females, one second generation mating was reported in which the
female was 18 months old”. It is unclear whether or not this mating was fertile as
its outcome was not reported. They added later, “Males reach sexual maturity at
about three years”» (see also WALLS 1998b). Anyway, a pair of E. colubrinus bred
by the same author (TRACER 1996) reached sexual maturity in 14 months (at which
time their rst courtship was observed) and the female produced its rst clutch at
approximately 19 months of age.
Specimens of E. colubrinus often live for 8-10 years in captivity (WALLS
The growth rate of E. colubrinus is illustrated by the following examples from
specimens kept in captivity (VAN WOERKOM 1987): (1) a female: newborn = 200 mm;
5 years old = 510 mm; 8 years old = 630 mm; (2) a male = 390 mm; after 2 years
= 485 mm; (3) six newborn specimens = 200 mm, 8.5 g; after 2 months = 270-310
mm; 25-35 g.
Eryx somalicus Scortecci 1939 (Figs 8-14, 18)
Eryx somalicus SCORTECCI 1939a: 269 [terra typica: environs of Mogadiscio (= Mogadishu), ca
02°01’N 45°20’E, Somalia; holotype MSNM 581 (formerly 2118)].
Derivatio nominis. The species name derives from the Latin somalicus (mascu-
line adjective; feminine: somalica; neuter: somalicum): “of, relating to, characteristic
of Somalia”, or “inhabiting Somalia”.
Ethnozoology. No data available, but this species is surely as feared as E. colu-
brinus by natives.
Somali records. Haud, 44°44’E 8°45’N, 3500 ft (PARKER 1949: 27, 1 BMNH
1949.2.1.46, leg. R.H.R. Taylor, VIII.1932, or 1933; LARGEN & RASMUSSEN 1993: 327,
map 7 of p. 409, 1 BMNH 1949.2.1.46, leg. R.H.R. Taylor, VIII.1932, or 1933); Mahad-
dei Uen (SCORTECCI 1939a: 269, 1 spec. MSNG, leg. L. Saito, 1915); Mogadishu (3 km
north) (GANS & LAURENT 1965: 60, 1 spec. MCZ R 72038, Gans’ Exp., 26.VIII.1961);
Mogadishu (coastal zone) (GANS 1965: 10); Mogadishu (environs of) (SCORTECCI 1939a:
269, 1 MZUF 33636, former 2119, MSNM paratypus, leg. Confalonieri 1935, and 1
MZUF 581, former 2118, MSNM holotypus, leg. G. Scortecci, 1931); northern coastal
belt of Somalia, Guban, central Somalia, Somalia south of Shebelli River (BALLETTO
1968: 267); northern (regions Togdheer and Bari) and southern (regions Mogadishu
and Middle Shebelli) Somalia (LANZA 1983: 221; 1990: 431); Somalia (BALLETTO 1968:
214; WALLS 1998b: 191, unnumbered g. of p. 191, map of p. 192).
88 B. Lanza and A. Nistri
89Somali Boidae and Pythonidae
Material studied. Nine specimens have been examined from the following localities: Bud
Bud (1 spec. MZUF 5250 and 1 MZUF 5251, leg. SBS 1968, 16.VIII.1968); Garbaqabat (1
MZUF 27653, leg. SBS 1980, 28.IV.1980); Gardò (1 MZUF 2588, leg. G. Scortecci, autumn
1953; 1 MZUF 6714, leg. SBS, 19.X.1973): Haud (locality of), 44°44’E 08°45’N, 3500 ft (=
Fig. 8. Head and neck of Eryx somalicus MSNM 581 (holotypus) from the environs of Mogadi-
shu; dorsal, ventral and lateral view (drawing by Manuela Mascherini).
90 B. Lanza and A. Nistri
91Somali Boidae and Pythonidae
ca 1150 m a.s.l.) (1 BMNH 1949.2.1.46, leg. R.H.R. Taylor, VIII.1932, or 1933); Mogadishu
(environs of) (1 MZUF 33636, ex 2119 MSNM, in exchange from the MSNM, paratypus, leg.
Confalonieri, 1935; 1 MSNM 581, ex 2118, holotypus, leg. G. Scortecci, 1931); Mogadishu (3
km N of) (1 spec. MCZ R 72038, leg. C. Gans’ Exp., 26.VIII.1961).
Description (Figs 8-14).
Morphology. Dorsal head scales more or less enlarged, smooth, juxtaposed as
far as eye level, imbricate backwards; scales between the anterior level of eyes and
postnasal larger than in E. borrii, their minimum number along the middle region of
the snout 1 (vs 3 in E. borrii n. sp., see below); rostral large and broad, about 2 times
wider than high, with a labial angular horizontal edge; nostril surrounded mostly by 2
scales only (internasal-prenasal complex and postnasal), since the internasal is usually
fused with prenasal bilaterally [completely fused in 8 of 9 specimens checked: MSNM
581 (formerly 2118; holotypus), MZUF 6714, 27653, 5250, 5251, 2588, 33636, BMNH
1949.2.1.46; fused only on the right side in MCZ R 72038]; 4(5.4)6 interorbital and
9(10)11 circumorbital scales; eye separated from the labials by 1 scale; 2(2.9)3 scales
between the postnasal and the eye; 8(9.71)10 supralabials; 11(11.6)12 scales between
the mental and the rst pseudoventral; mental groove absent; body scales smooth
anteriorly, becoming increasingly keeled distad; tail scales strongly keeled; 34(37.2)40
mid-body scales; ventrals (wider than long system) 156(158.7)163, Dowling system
155(156.3)158; anale entire; 21(22.6)25 single subcaudals; tail short, about 8-10.5%
the total length, conical, pointed, sometimes more or less curved downward at tip.
Size seemingly small; the ratio total length/maximum transverse diameter 24.03-30.00
(n = 4); the largest specimen known thus far, the MZUF 33636, has a total length
of 390 mm. GANS & LAURENT (1965), referring to the unsexed specimen MCZ R 72038
from Mogadishu (probably , VAN WALLACH personal comm.), pointed out that «there
is both increase and reduction of dorsal scale rows, as well as the intercalation of
additional scales within the dorsal rows. The latter occurs commonly in the second and
third lateral rows, and the additional scales are found in all rows dorsal to the row in
question. The 159 ventrals thus correspond to 204 scales along the mid-dorsal line».
Hemipenis unknown.
Fig. 9. — Midbody view of Eryx somalicus MSNM 581 (holotypus) from the environs of Mogadishu
(drawing by Manuela Mascherini).
90 B. Lanza and A. Nistri
91Somali Boidae and Pythonidae
Maxillary and mandibular teeth solid, ungrooved, arranged in an anonodont
and scaphiodont set; RIEPPEL (1978), who studied the skull of the specimen MZUF
5250, found 9 maxillary teeth, 4 functional teeth on the palatine dentigerous process
and only 3 teeth on the shortened pterygoid; TOKAR (1989), who studied the same
skull (cf. p. 47), found 5 palatine teeth (p. 53) and 5 pterygoid teeth (3 teeth and 2
sockets; footnote of p. 51); in 3 dead specimens with open mouth we counted ca 8-9
maxillary and mandibular teeth, 4 palatine functional teeth, and 3-4 pterygoid func-
tional teeth.
Table 3.
Eryx somalicus: characters of the Somali specimens studied.
Character Sex n Range Mean cl (99%)
TaL% ¢1 8.28 —
TaL% ™™ 7 8.07-10.42 8.88±0.306 7.75-10.02
RR ™™ 5 1.93-2.22 2.09±0.049 1.86-2.32
IO ¢¢+™™ 10 4-6 5.40±0.221 4.68-6.12
CO left ¢¢+™™ 10 9-11 9.90±0,233 9.14-10.66
CO right ¢¢+™™ 10 9-11 10.00±0.258 9.16-10.84
OS left ¢¢+™™ 10 1 1.00±0.000 —
OS right ¢¢+™™ 10 1 1.00±0.000 —
NO left ¢¢+™™ 7 2-3 2.86±0.143 2.33-3.39
NO right ¢¢+™™ 7 2-3 2.86±0.143 2.33-3.39
UL left ¢¢+™™ 7 8-10 9.71±0.286 8.66-10.77
UL right ¢¢+™™ 8 9-10 9.88±0.125 9.44-10.31
GL ¢¢+™™ 5 11-12 11.60±0.245 10.47-12.73
NS ¢¢+™™ 7 33-38 35.14±0.634 32.79-37.49
MSR ¢¢+™™ 10 34-40 37.20±0.593 35.27-39.13
PC ¢¢+™™ 8 23-27 25.88±0.515 24.07-27.68
V¢¢+™™ 9 156-163 158.67±0.764 156.10-161.23
VD ™™ 4 155-158 156.25±0.750 151.87-160.63
PV ™™ 4 2-3 2.5±0.289 0.81-4.19
SC ¢¢+™™ 10 21-25 22.60±0.499 20.98-24.22
Abbreviations as in Table 1.
Table 4.
Negative allometric growth of the eye in Eryx somalicus.
3.30 1.40 2.36
3.55 1.56 2.28
5.00 2.00 2.50
5.10 2.00 2.55
5.60 1.80 3.11
Measures in mm; SnL= snout length, i.e. distance between snout tip and eye; EHD = horizontal eye
92 B. Lanza and A. Nistri
93Somali Boidae and Pythonidae
For the characters of the examined specimens (7 ™™, 1 ¢?, 2 unsexed speci-
mens), corresponding to almost all the known ones, see Table 3 and Appendix 1.
The negative allometric growth of the eye is shown in Table 4.
Fig. 10. — Side view of the head of Eryx somalicus ¢? MZUF 5250 from Bud Bud (drawing by
Benedetto Lanza).
Fig. 11. — Eryx somalicus MSNM 581 (holotypus) from the environs of Mogadishu, preserved
specimen (photo by Benedetto Lanza).
92 B. Lanza and A. Nistri
93Somali Boidae and Pythonidae
Fig. 12. — Eryx somalicus MSNM 581 (holotypus) from the environs of Mogadishu, preserved
specimen; head and neck, dorsal view (photo by Benedetto Lanza).
Fig. 13. Eryx somalicus BMNH 1949.2.1.46 from Haud, preserved specimen (photo by Bene-
detto Lanza).
94 B. Lanza and A. Nistri
95Somali Boidae and Pythonidae
Colouration (in alcohol). Pattern rather variable. Dorsal ground colour light to
rather dark brown, crossed by about 30 off-white, sometimes dark-edged, transverse
to more or less oblique stripes, each approximately 1-6 scales long; some of the
light stripes may fuse with each other forming Y or X shaped marks. Furthermore
the back is nely dark-striped longitudinally, since each dark scale is off-white
medially. Lower parts of the anks off-white with dark, irregular small spots and
sometimes with a series of brownish black, larger roundish spots, each usually
underlying a light dorsal stripe.
Distribution. Thus far Eryx somalicus is known only from Somalia but, since
it has also been collected along the border between Somalia and Ethiopia, it «will
undoubtedly be found eventually in neighbouring parts of eastern Ethiopia» (LAR-
GEN & RASMUSSEN 1993: 327). WALLS’ (1998b) statement about its occurrence in
Ethiopia is merely conjecture.
Biology. Practically unknown, but probably very similar to that of its con-
geners (see Eryx, Biology). The species, usually associated with sandy, arid and
semi-arid habitats, has been recorded at altitudes between sea level (Mogadishu)
and about 1150 m (Haud, 3500 ft). It has been found in rather different ecosystems
which, according to the geobotanical map published by PICHI-SERMOLLI (1957), are
characterized by the following types of vegetation: shrub steppe, subdesert scrub,
broken xerophilous open woodland, xerophilous open woodland, and coastal for-
mations. PARKER (1949) stated that the «specimen from the Haud was collected in
August in sandy country with patches of grass and scattered large thorn trees; it
was found during daylight half in and half out of a hole in the ground». The Moga-
Fig. 14.Eryx somalicus MZUF 5251 from Bud Bud, preserved specimen (photo by Saulo
94 B. Lanza and A. Nistri
95Somali Boidae and Pythonidae
dishu specimen mentioned by GANS & LAURENT (1965) «was taken on the beach at
an altitude of 5 m, some 10 m from the breaker zone. It was coiled in a small space
in sand between two rocks. On rst sight its color pattern was reminiscent of the
Dasypeltis found in the same zone, but the individual did not engage in defensive
behavior». The MZUF 5250 from central Somalia (Bud Bud) was collected under a
stone in the broken xerophilous open woodlands.
The stomach of a juvenile (¢? MZUF 5250) 145 mm long contained an Hemi-
dactylus with a snout-vent length of about 45 mm.
The MZUF 6714 from Gardò, total length 254 mm, has an ovary with 5
eggs of comparable size, the largest one 6 mm long; the largest egg still present in
the MZUF 2588 from Gardò, total length about 240 mm and strongly damaged
abdomen, is 5.2 mm long.
Eryx borrii n. sp. (Figs 15-18)
Material studied. 1 holotype, BMNH 1900.11.28.4, Biji (400 m a.s.l.), leg. Donaldson
Description of holotype (Figs 15-17).
Morphology. Dorsal head scales enlarged, smooth, juxtaposed as far as eye level,
imbricate backwards; scales between the anterior level of eyes and postnasal smaller
than in E. somalicus, their minimum number along the middle region of the snout 3;
rostral large and broad, about 2.5 times wider than high, with a labial angular hori-
zontal edge; nostril surrounded by 2 scales only (postnasal and a scale deriving from
the fusion of internasal and prenasal, seemingly only partially fused on the right
side); 5 interorbital and 10/10 circumorbital scales; eye separated from the labials by
1/1 scale; 3/3 scales between the postnasal and the eye; 10/11 supralabials; 12 scales
between the mental and the rst pseudoventral; mental groove absent; body scales
smooth anteriorly, becoming increasingly keeled distad; tail scales strongly keeled;
39 mid-body scales; ventrals (wider than long system) 193, Dowling system 189; anal
entire; 26 single subcaudals; tail short, about 8.25% the total length, conical, pointed,
slightly curved downward at tip. Body slender, the ratio total length/maximum trans-
verse diameter about 39.35; total length of 390 mm.
Maxillary and mandibular teeth solid, ungrooved, arranged in an anonodont
and scaphiodont set; about 8-9 maxillary and mandibular teeth; palatine and ptery-
goid not counted.
Colouration (in alcohol). Dorsal ground colour brown, lighter on head and
sides of body, with longitudinal or more or less oblique off-white fragmented
stripes, sometimes fused to form undulating narrow bands. Lower parts of the
anks off-white with dark, irregular small spots and sometimes with a series of
brownish black, larger roundish spots.
Derivatio nominis. We take pleasure in dedicating this new species to our
friend Dr Marco Borri (Sezione Zoologica «La Specola», Museo di Storia Naturale
dell’Universi degli Studi di Firenze), irreplaceable companion and valuable
collaborator during several expeditions in Italy and abroad.
96 B. Lanza and A. Nistri
97Somali Boidae and Pythonidae
Fig. 15. — Head and neck of Eryx borrii BMNH 1900.11.28.4 (holotypus) from Biji; dorsal, ventral
and lateral view (drawing by Manuela Mascherini).
96 B. Lanza and A. Nistri
97Somali Boidae and Pythonidae
Ethnozoology. No data available, but this species is surely as feared as E.
colubrinus by natives.
Somali records. Biji (PARKER 1949: 27, 1 BMNH 1900.11.28.4, leg. A. Donaldson
Smith; LARGEN & RASMUSSEN 1993: 327, map 7 of p. 409, BMNH 1900.11.28.4);
Northern Somalia (region West Galbeed) (LANZA 1983: 221; LANZA 1990: 431).
Remark on the Somali records. GANS & LAURENT (1965: 61) stated: «It should
be noted that the upland (3,500 ft) record, whose counts are cited by Parker, has
193 rather than 159 ventrals. This suggests the presence of a distinct high altitude
(or northern?) population». GANS & LAURENT (1965) obviously refer to the BMNH
1900.11.28.4 from Biji, but their altitudinal indication is wrong; Biji is only 400
Fig. 16. Eryx borrii BMNH 1900.11.28.4 from Biji, preserved holotypus; head and anterior part
of the body, dorsal view (photo by Benedetto Lanza).
98 B. Lanza and A. Nistri
99Somali Boidae and Pythonidae
m a.s.l. and the altitude of 3,500 ft (ca 1150 m) cited by PARKER (1949) actually
relates to the locality of the Haud (44°44’E 08°45’N) where the of Eryx somalicus
1949.2.1.46 BMNH was collected.
Distribution. Thus far Eryx borrii is known to occur only in the type locality.
Biology. Unknown, but probably very similar to that of its congeners (see Eryx,
Biology). Judging from the geobotanical map published by PICHI-SERMOLLI (1957),
the only known specimen of the species was probably collected in the shrub steppe.
Remarks. The new species is morphologically close to E. somalicus from
which it is distinguished by a more slender habitus (ratio total length/maximum
transverse diameter about 39.35, versus 24.03-30.00 in E. somalicus); a larger num-
ber of ventral scales (193 instead of no more than 163 in E. somalicus); smaller size
of the scales between the anterior level of eyes and postnasal (their minimum num-
ber along the middle region of the snout 3, versus 1 in E. somalicus). The dorsal
pattern shows longitudinal or more or less oblique off-white fragmented stripes that
are transverse to more or less oblique in E. somalicus.
Fig. 17. Eryx borrii BMNH 1900.11.28.4 from Biji, preserved holotypus (photo by Benedetto
98 B. Lanza and A. Nistri
99Somali Boidae and Pythonidae
Family Pythonidae Fitzinger 1826
According to MCDIARMID et al. (1999), the family includes the following 8 gen-
era: Antaresia Wells & Wellington 1984; Apodora Kluge 1993; Aspidites Peters 1877;
Fig. 18. — Somali distribution of Eryx borrii (p, 1, Biji) and Eryx somalicus (ò, 2, Gardò; 3, Gar-
baqabat; 4, Haud; 5, Bud Bud; 6, Mahaddei Uen; 7, Mogadishu).
100 B. Lanza and A. Nistri
101Somali Boidae and Pythonidae
Bothrochilus Fitzinger 1843; Leiopython Hubrecht 1879; Liasis Gray 1842; Morelia
Gray 1842; Python Daudin 1803.
Only the genus Python occurs in Somalia.
Genus Python Daudin 1803
Python DAUDIN 1803a: 187 (type species in RUSSELL 1796: plates 24 “Pedda Poda B” and pl. 39
“Bora” (= Coluber Molurus Linnaeus 1758) by monotypy; see also ZHAO & ADLER 1993: 199,
for details on the date of publication, and 222).
Constrictor (not of Laurenti 1768) WAGLER 1830: 168 (type species: not designated, subsequently
designated as Constrictor Schneideri Wagl[er]” (= Boa Reticulata Schneider 1801) by
FITZINGER 1843: 24).
Enygrus WAGLER 1830: 166-167 (type species: not designated; subsequently designated as “Enygrus
regius Wagl[er] (= Boa Regia Shaw 1802) by FITZINGER 1843: 24).
Engyrus GRAY 1831: 97 (incorrect subsequent spelling of Enygrus, apparently a typographical error).
Enygris GRAY 1842: 42 (unjustied emendation of Enygrus).
Heleionomus GRAY 1842: 42 (type species: Heleionomus variegatus Gray 1842 (= Coluber Sebae
Gmelin 1789) by monotypy).
Hortulia GRAY 1842: 43 (type species: Hortulia Natalensis Gray (= Python natalensis A. Smith 1833)
by monotypy).
Asterophis FITZINGER 1843: 24 (type species: Python Tigris Daudin 1803 (= Coluber Molurus Linnaeus
1758) by original designation).
Hortalia SCUDDER 1882: 166 (emendation of Hortulia).
Aspidoboa SAUVAGE 1884: 143 (type species: Python curtus Schlegel 1872 by monotypy).
Pyton FUNAIOLI 1957: 79 (incorrect subsequent spelling of Python).
Derivatio nominis. The genus name derives from the Latin Python (masculine
noun; genitive Pythonis): the serpent killed at Delphi by Apollo. The gender of the
genus is masculine.
Content and distribution. Following MCDOWELLs view (1975), the genus should
include a dozen species inhabiting: Africa S of the Sahara [P. anchietae Bocage 1887,
P. regius (Shaw 1802), P. sebae (Gmelin 1789), P. saxuloides Miller & Smith 1979 from
Kenya, whose validity is to be conrmed], Palaearctic Region (where P. molurus (Lin-
naeus 1758) occurs in a restricted range), Oriental Region [P. curtus Schlegel 1872,
P. molurus (Linnaeus 1758), P. reticulatus (Schneider 1801)] and Australian Region
[P. amethistinus (Schneider 1801), P. boeleni (Brongersma 1953), P. bredli GOW 1981,
P. carinatus Smith 1981 from W Australia, whose validity is to be conrmed, P. oen-
pelliensis Gow 1977, P. reticulatus (Schneider 1801), P. spilotus (Lacépède 1804), P.
timoriensis (Peters 1876)]. According to UNDERWOOD & STIMSON (1990) and KLUGE
(1993b), only the African and Asiatic species, including reticulatus and timoriensis,
belong to the genus Python, while the remaining species pertain to the genus Morelia.
Here we consider as belonging to the genus only 4 African species, all occur-
ring South of the Sahara [P. anchietae Bocage 1887, P. regius (Shaw 1802), P. sebae
(Gmelin 1789), P. natalensis A. Smith 1840, a full species according to BROADLEY
(1999), of which P. saxuloides Miller & Smith 1979 should be a synonym (BROADLEY
1984)], and 4 Asiatic species living in the Palaearctic (where P. molurus occurs in a
restricted range) and Oriental Regions [P. curtus Schlegel 1872, P. molurus (Linnae-
us 1758), P. reticulatus (Schneider 1801), P. timoriensis (Peters 1876)] (cf. UNDER-
WOOD & STIMSON 1990, KLUGE 1993b, MCDIARMID et al. 1999).
Only one species occurs in Somalia: Python sebae.
100 B. Lanza and A. Nistri
101Somali Boidae and Pythonidae
Fossils. RAGE (1984), in his review of fossil snakes, cited only two species of
the genus Python: Python maurus, from the Miocene of Morocco, which «may not
be a pythoninae» (KLUGE 1993b: 55), and P. sardus, from the Middle Miocene of
Sardinia; KLUGE (1993b: 55) stated that this species is «maybe correctly identied to
genus; however, a more detailed study on the only available material, an articulated
palatine and anterior pterygoid fragment, is required to determine the presence of
diagnostic features». SZYNDLAR & RAGE (2003), in their revision of the European
non-erycine Booidea, assign to the genus three species: Python euboicus Roemer
1870 (early Miocene, Greece), Python europaeus n. sp. (middle Miocene, France)
and the above-mentioned Python sardus.
Description. Large snakes, the largest of which, the SE Asiatic P. reticulatus
has been fairly reliably recorded as attaining a total length of close to 10.5 m; head
nearly as broad as body, depressed, elongate and distinct from neck; snout long,
broadly rounded, without any canthus; rostral touching 6 shields, with a wedge-
shaped pit on each side; mental grooved; 2-5 anterior upper labials deeply pitted;
a few anterior lower labials and sometimes also some of the posterior ones from
deeply to feebly pitted; snout covered with shields; upper surface of head covered
with more or less symmetrical shields or small scales; nasals semi-divided and
separated from each other by a pair of internasals; nostril large, placed high on
the snout, directed upwards or upwards and outwards; eye moderate with verti-
cally elliptic pupil. Body massive, cylindrical or somewhat compressed, covered
with rather small, smooth scales, about 53-99 at midbody; ventrals well developed,
but distinctly narrower than the body; anal entire or divided; terminations of the
concealed rudimentary hind limbs seen as small claw-like processes on each side of
the vent, larger in males than in females. Tail moderately long to short, prehensile;
subcaudals all or mostly paired. Premaxillary toothed; palatine and pterygoid teeth
scaphiodont; maxillary and mandibular teeth numerous, anododont and scaphi-
odont, the anterior ones strongly recurved. Hypapophyses absent over posterior
part of the vertebral column.
Data on the rearing of the typical pythons are given by many authors (e.g.
WELCH 1987: 81, BARKER & BARKER 1994, WALLS 1998a, FAIN 2001, VIANELLO 2002,
Python sebae (Gmelin 1789) (Figs 19-30)
Coluber Sebae GMELIN 1789: 1118.
C[oluber] Speciosus BONNATERRE 1790: 17. Holotype based on pl. 199, g. 2 in SEBA 1735. Type-
locality: “Brazil”, in error, according to LOVERIDGE 1957: 248.
Boa Hieroglyphica SCHNEIDER 1801: 266. Holotype based on pl. 27, g. 1, in SEBA 1735. Type-locality:
“Cairo, Egypt” by inference as taken from Seba, according to LOVERIDGE 1957: 248.
Python Houttuyni DAUDIN 1803b: 254. Type(s) unlocated. Type-locality: unknown. (Possibly synonym
of P. natalensis).
Heleionomus variegatus GRAY 1842: 43. Holotype BMNH, no. IV. 3. 1a, presumed lost. (Possibly
synonym of P. natalensis).
Python sebae; DUMÉRIL & BIBRON 1844: 400.
Python Liberiensis HALLOWELL 1845: 249. Holotype: ANSP 6705. Type-locality not given in original
description but presumed to be “Liberia, West Africa” based on stated locality for other species
described in the same paper. (Considered a synonym of P. sebae by HALLOWELL 1857: 66).
102 B. Lanza and A. Nistri
103Somali Boidae and Pythonidae
Boa Liberiensis HALLOWELL 1854: 100.
Python jubalis PITMAN 1936: 211. Type(s) given as in the museum at Mogadiscio, though I (C.R.S.
Pitman) am unaware of any published description in the original paper. Loveridge 1957: 248,
also stated that no published description was known. Type-locality: “Juba River” (nomen
Boa constrictor (not Linnaeus) LIPPARONI 1954: 655.
Python molurus (not Linnaeus) BELLAIRS 1969: pl. 30.
P[ython] sebae sebae BROADLEY 1983: 68.
Derivatio nominis. The species name derives from the genitive (Sebae) of the
latinized family name Seba, as the species was named by J.F. Gmelin in honour of
the Dutch naturalist Albertus Seba (Etzel, E Friesland, now in NW Germany, May
2, 1665 Amsterdam, May 2, 1736). «A merchant by profession, Seba possessed
the largest collection of natural history objects of his era» and «published a
colossal four-volume work, in folio, to describe his two collections, the famous
“Locupletissimi Rerum Naturalium Thesauri [...]”. The text is mediocre, but the 449
outstanding plates gure literally thousands of specimens in life size, including an
especially large number of reptiles and amphibians, which form the largest part of
volumes 1 and 2, issued in 1734 and 1735.» (ADLER 1989: 9).
Ethnozoology. At least the big specimens of Python sebae are widely feared in
Somalia especially because of their potential to kill by constriction.
Somali records. Alessandra Island (BROADLEY 1984: 362); Arbidin (pond
of), along the lower Shebelli River (TEDESCO ZAMMARANO 1924: 158, tracks,
18.II.1922; TEDESCO ZAMMARANO 1930: 206, leg. V. Tedesco Zammarano, 3.II.1922);
Bardera (BOULENGER 1909b: 311, 1 spec. 30025 MSNG, leg. U. Ferrandi, 1908);
Bardera (north of) (STEFANINI 1922: 210, uncertain sight record); Belet Amin
(SCORTECCI 1939a: 270, 5 spec. MSNG, leg. S. Patrizi, [VI-VIII].1934); Ciula
Island (GROTTANELLI 1955: 5); Comia (environs of), near the Shebelli River
(TEDESCO ZAMMARANO 1924: 125-126, tracks, 30.I.1922, 128, g. p. 129, g. p.
130, 1 spec., leg. V. Tedesco Zammarano, 30.I.1922, and pp. 138-139, 1 spec.,
leg. V. Tedesco Zammarano, 3.II.1922; 1930: 165, tracks, 30.I.1922, 168, g. p.
169, 1 spec., leg. V. Tedesco Zammarano, 30.I.1922, and pp. 180-182, leg. V.
Tedesco Zammarano, 3.II.1922); ex Italian Somalia (ANONYMOUS 1933: 63; TESTI
1935: 106, 115; SCORTECCI 1936: 922); Gelib [= Jilib] (BROADLEY 1984: 362);
Giohàr (LANZA 1968: unnumbered g. of columns 239-240, 1 MZUF 5154,
leg. SBS, 26.IV.1968; BROADLEY 1984: 362); Jilib (= Gelib), near the Juba River
(RALLO 2000: 287-288, 1 unsexed skin MSNV CCC4962, leg. N. Forin, 1956);
Kismayu (RALLO 2000: 287-288, g. 2, 1 unsexed skin MSNV CCC4963, leg. D.E.
Ancilotto, 1952); Labadaad and Bulo Yaag (= Yaaq) (between) (JOHNSON 1981: 1,
1 , 27.VII.1978; HOEVERS & JOHNSON 1982: 184, 1 young ¢ MZUF 27044, eld
number 72112, 14.VII.1978, 1 young MZUF 27045, eld number 72070, 28.V.1978,
1 young CAS 153324, eld number 72299, 8.V.1979, 1 young ¢ CAS 153325, eld
number H 0892, 10.II.1980); Lugh Guddèi peninsula (very common; less common
at Lugh than in the forests along the lower Shebelli River) (FERRANDI 1903: 73);
Manamofa (BROADLEY 1984: 362); Mareri (BROADLEY 1984: 362); Middle Shebelli
River (LIPPARONI 1954: 655); Mogadishu (GESTRO 1932: 10); Nogal (CORONARO
1926: 32, 1 spec. 31899 MSNG, donated by G. Lagorio); Ola Uager (SCORTECCI
1939a: 270, 2 spec. MSNG, leg. S. Patrizi, 18.VIII.1934); Shebelli and Juba
rivers (along) (SCORTECCI 1937: 265, SIMONETTA & MAGNONI 1986: 424); Shebelli
102 B. Lanza and A. Nistri
103Somali Boidae and Pythonidae
River (regions south of the) (BALLETTO 1968: 267); Somalia (FUNAIOLI 1957: 79,
pl. XXIV, g. 4; LAPINI 1984: 48, 1 spec. MFSN 175, donated by S. Costantini);
Somalia (southernmost parts of) (BALLETTO 1968: 225); Somalia (watered districts
in the south of) (PARKER 1949: 28); Southern Somalia (regions Middle Shebelli,
Mogadishu, Bay, Lower Shebelli, Gedo, Middle Juba, and Lower Juba) (LANZA 1983:
221; LANZA 1990: 432); Upper Bubasci (along the), near Ola Uager (PATRIZI 1935: 22,
Material studied. Fourteen specimens have been examined from the following localities:
Alessandra Island (2 young ¢¢ MZUF 2214 and 2216, 3 young ™™ MZUF 2202 and 2215 sent in
exchange to the Universidade Federal de Rio de Janeiro, Brazil, MZUF 2217 sent in exchange
to the Instituto Butantan, São Paulo, Brazil, leg. SBS 1962 and natives, 14.VIII.1962, No. 2202,
other 15.VIII.1962); Giohàr (1 young ¢ MSNM 802, ex 2120, leg. U. Fiechter, 1936; 1 young
MZUF 3972, leg. SBS 1964 and a native; 1 ad. MZUF 5154, leg. SBS and a native, 26.IV.1968;
1 young MZUF 5653, leg. SBS, Ettore Granchi, 3.VIII.1970); Jilib (1 young MZUF 18483,
leg. a native, donated by G. Tarabini Castellani, 22.IV.1971; 2 adult, unsexed specimens, head
only, MZUF 21651 and MZUF 21663, leg. natives, donated by G. Tarabini Castellani, 1970);
Labadaad (00°30’N 42°46’E) and Bulo Yaag (00°12’N 42°46’E) (= Yaaq) (between) (1 young ¢
MZUF 27044, eld number 72112, and 1 young MZUF 27045, eld number 72070, leg. L.G.
Hoevers and P.M. Johnson, 14.VII.1978 and 28.V.1978, respectively).
Description (see also Description of the genus) (Figs 19-29).
Morphology. Head subtriangular, elongate, attened, well distinct from the neck;
snout narrowly rounded. Body slender in juveniles, becoming more and more stout
with age, slightly depressed in the big animals, otherwise cylindrical; in the environs
of Comia, near the Shebelli River (Somalia), TEDESCO ZAMMARANO (1924: 139) killed
a fasting specimen 5.72 m long with a midbody circumference of 41 cm. A 4.45 m
long, found dead between Labadaad and Bulo Yaag (lower Juba Valley), had a girth
of 46 cm (JOHNSON 1981, and HOEVERS unpublished notes). The greatest, seemingly
trustworthy total length known thus far is, 9.80 m (9.81 m in POPE 1961, for the
same specimen), attained by a specimen killed in a garden hedge at Bingerville
(05°21’N 03°54W; Ivory Coast), measured by CH. BÉART in 1932 (in litt. to Villiers:
VILLIERS 1950, 1975); however SPAWLS et al. (2002), agreeing with BRANCH (1984),
considered this record «an anecdotal report (without any tangible evidence) […] and
a slightly more credible record of a 7.5 m specimen from a research station, also
in Ivory coast, but again evidence is lacking». STUCKI-STIRN (1979: 50), according to
whom the pythons found near habitations are usually smaller than snakes taken in
remote areas, regarded as a very important specimen an animal 7.85 m long, found
in Cameroon (Mt Ali). STARIN & BURGHARDT (1992: 53) stated that in Gambia «During
the systematic eld study 52 python sightings were made on 45 different occasions.
The pythons ranged in size from .5 m to about 7.5 m in total length. Based on 50
sightings, the median and modal total length was 3 m with a mean of 3.3 m (sd =
1.3); 20% of the sightings were of animals greater than 4.5 m». However, there is
considerable variation in weight according to condition and sex, the females being
proportionately heavier than the males. Tail moderately long, prehensile, about 11 to
16% of the total length in males and 9 to 14% in females, usually terminating with a
sharp, almost spiny point; its growth seems to be isometric since we found similar
percentage values in two freshly killed Somali females with a total length of 1055
and 3970 mm, 10.43% and 10.93%, respectively. Rostral about as broad as deep,
visible from above, with an elongate oblique pit on either side of its apex. Nostril
104 B. Lanza and A. Nistri
105Somali Boidae and Pythonidae
large, placed high on the snout, directed upwards and outwards; nasal semidivided.
Eye moderate, obliquely positioned as seen in dorsal view and much exposed;
Fig. 19. — Head and neck of Python sebae young MZUF 27044 found between Labadaad and Bulo
Yaag; dorsal, ventral and lateral view (drawing by Manuela Mascherini).
104 B. Lanza and A. Nistri
105Somali Boidae and Pythonidae
pupil vertically elliptic, readly distinguishable in live specimens; iris brownish. Two
subrectangular internasals 1.5 times as long as broad or a little shorter, followed by
a pair of large anterior prefrontals and a pair of smaller posterior prefrontals, the
latter partly or completely separated by 1-3 azygous shields (interprefrontals). Usually
a pair of more or less lengthened frontals (which may be fused). Supraocular large,
irregularly shaped, broken up into two or more shields, usually two. Frontals and
supraoculars followed by a variable number of unequal, irregularly arranged shields.
Posterior third or fourth of the head upper surface covered by scales similar to those
of the back, larger in front and gradually merging backwards into those of the body.
About 8-16 loreals, disposed in 2-3 irregular rows; 5-13 circumorbitals (usually 6-10)
made up of 1-2 supraoculars, 2-3 preoculars (a large upper and 1-2 smaller ones
below), 1-2 suboculars, and 1-3 postoculars (the upper one the largest); 11-17 upper
labials, always separated from the eye, the rst two deeply pitted; 15-25 lower labials,
the anterior and posterior ones respectively with 2-6 and 3-7 small, shallow pits near
their upper margin [in two specimens from central Africa LAURENT (1956) found the
following lower labial pit formulae: 2-3 + 12-15 in a juvenile and a male (another ¢
and another 2 ™™ cited by him proved to be P. natalensis; D.G. BROADLEY pers. comm.,
VI.2004); in ve specimens of both sexes from the Virunga National Park (Democratic
Republic of Congo) DE WITTE (1975) found the following formulae: 2-3 + 16-19, 1-
3 + 16-19, 2-4 + 17-19, 2-5 + 15-17, and 2-4 + 15-18; MADERSONs (1970) data are
incomplete: «There are pits in the rst two supralabials, and the last four infralabials,
although the fourth is much smaller than the others»; for data on Somali specimens
see Table 5]; body covered by rather small, smooth scales, 71-99 at midbody
(respectively about 66-77 and 43-54 rows a head length behind the occiput and
anterior to the cloaca); ventrals 259-294; anal entire, rarely divided; 55-81 subcaudals
all, or for the most part, in two rows. The dorsal scale microdermatoglyphics have
been described by PRICE & KELLY (1989).
Fig. 20. — Midbody view of Python sebae young MZUF 27044 found between Labadaad and Bulo
Yaag (drawing by Manuela Mascherini).
106 B. Lanza and A. Nistri
107Somali Boidae and Pythonidae
ALEXANDER & GANS (1966) found in one specimen 278 (ventral) and 606 (dor-
sal) body scale rows, > 78 (ventral) and > 109 (dorsal) tail scale rows, 283 body and
> 75 tail vertebrae and the following dermal-vertebral ratios: ventral body 1:1, dor-
sal body 2.1:1; ventral tail 1:1, dorsal tail 1.5:1.
The hemipenis is heavy, bilobed with a projection (awn) at the tip of each
lobe, and provided with a strong ounce on the stem and a sulcus spermaticus
bifurcating beyond half the length of the organ; its surface is tomentose but spine-
less (DOUCET 1963: 227, pl. VII, gs E-F [but B, A on plate]; WALLS 1998a).
On each side 2 premaxillary, 13-16 maxillary (16-18, LEESON 1950; 22, EDMUND
1969; 17, CHIPPAUX 2001), 6 palatine (7, EDMUND 1969; 8, LEESON 1950), 8-9 ptery-
goid (10 LEESON 1950; 9-10 EDMUND 1969), and 13-17 mandibular teeth (Edmund’s
and Chippaux’s data obtained from their gs of the skull). The bones of the skull
were studied in detail by FRAZZETTA (1959, 1966).
Fig. 21. Python sebae MZUF 5154, about 4 m long, from Giohàr (photo by Benedetto Lanza).
106 B. Lanza and A. Nistri
107Somali Boidae and Pythonidae
The characters of the Somali specimens studied by us are shown in Table 6
and Appendix 1.
In the young Somali MZUF 27045 (in alcohol; skull not prepared!) we
counted on each side respectively 2 and about 13, 7, 8, 16 premaxillary, maxillary,
palatine, pterygoid and mandibular teeth.
The negative allometric growth of the eye is shown in Table 7.
Colouration. Background greyish to pale brown or tan above, with more or
less sinuous dark brown, black-edged saddles, which are variable in shape and size
and often connected by a discontinuous dark stripe running along each side of the
back. A most unusual specimen with a partially striped dorsum was described by
WALLS (1998a: 171), according to whom «Striping frequently is not genetic but
instead is caused by cool temperatures during incubation»; colour and pattern
mutations, all obtained in captivity, include patternless and hypomelanistic individ-
uals, as well as various forms of aberrant striping that may or may not be genetic
(MCCURLEY & GLASGOW 2003); a patterness animal is portrayed in VIANELLO (2002).
General colour of the sides lighter, nely dark dotted and with irregular, more or
less vertical or subvertical spots, sometimes fused with the dorsal ones and becom-
ing more or less irregularly C- to U-shaped, with the open side anteriorly, about
from midbody cephalad; the lateral spots are more sharply dened behind, becom-
ing gradually less distinct on the anterior part of the body. Upper surface of tail
with a continuous light to golden-yellow stripe between two black ones; this stripe
may extend well forward onto the body as a continuous or interrupted vertebral
light stripe. A large arrowhead-like dark brown blotch occupies the top of the head,
bordered on each side by a light stripe beginning at the end of the snout, above
the nostril, and passing above the eye; head lateral and ventral pattern clearly dif-
Table 5.
Python sebae: number and position of the lower labial pits.
Anterior sublabial pits Posterior sublabial pits
Number Sex No. of pits:
Position of pits:
No. of pits:
Position of pits:
MSNM 802 ¢3/3 2-4/2-4 3/4 14-16/15-18
MZUF 2214 ¢3/3 2-4/2-4 5/3 12-17/15-17
MZUF 2216 ¢4/4 2-5/2-5 6/5 14-19/14-18
MZUF 27044 ¢3/3 2-4/2-4 7/6 14-20/15-20
MZUF 2202 4/3 2-5/2-4 4/4 14-17/14-17
MZUF 3972 4/4 2-5/2-5 6/5 14-19/14-18
MZUF 5653 3/3 2-4/2-4 4/4 14-17/14-17
MZUF 18483 3/3 2-4/2-4 4/? 15-18/ ?
MZUF 27045 3/4 2-4/2-5 5/5 15-20/15-20
MZUF 21651 ? 3/3 2-4/2-4 5/5 13-17/14-18
MZUF 21663 ? 4/4 2-5/2-5 3/4 15-17/15-18
108 B. Lanza and A. Nistri
109Somali Boidae and Pythonidae
Table 6.
Python sebae: characters of the Somali specimens studied
Character Sex n Range Mean cl (99%)
TaL% ¢¢ 4 12.51-12.87 12.65±0.078 12.19-13.10
TaL% ™™ 7 10.43-13.75 12.03±0.431 10.44-13.63
RR ¢¢+™™ 8 1.02-1.30 1.16±0.032 1.05-1.27
IPPF ¢¢+™™ 11 1-3 1.91±0.211 1.24-2.58
CO left ¢¢+™™ 14 5-9 7.71±0.345 6.68-8.75
CO right ¢¢+™™ 12 5-9 7.58±0.398 6.34-8.82
UL left ¢¢+™™ 14 13-16 14.00±0.210 13.37-14.63
UL right ¢¢+™™ 14 12-16 14.00±0.277 13.16-14.83
LL left ¢¢+™™ 12 19-23 21.33±0.414 20.04-22.62
LL right ¢¢+™™ 11 20-23 21.73±0.359 20.59-22.87
MSR ¢¢+™™ 12 82-92 87.00±0.739 84.71-89.29
NS ¢¢+™™ 9 63-71 67.33±0.882 64.37-70.29
PC ¢¢+™™ 9 44-46 44. 78±0.278 43.85-45.71
V¢¢+™™ 11 270-280 275.18±0.923 272.26-278.11
VD ¢¢+™™ 11 270-280 275.18±0.923 272.26-278.11
SC ¢¢+™™ 10 65-74 68.90±0.971 65.74-72.06
SC ¢¢ 4 67-74 71.50±1.555 62.42-80.58
SC ™™ 6 65-69 67.17±0.601 64.74-69.59
Abbreviations: cl = condence limits; CO = number of circumorbital scales; IPPF = number of
interprefrontals (i.e. number of scales between posterior prefrontals); LL = lower labials; MSR =
numbers of longitudinal rows of scales at midbody; NS = number of longitudinal rows of scales
behind the occiput; PC = number of longitudinal rows of scales a head length anterior to the cloaca;
RR = rostral width/rostral depth; SC = number of subcaudals; TaL% = tail length x 100/total length;
UL = upper labials, supralabials; V = number of ventrals according to the “wider than long system”,
i.e. of pseudoventrals + true ventrals; VD = number of ventrals according to the “Dowling system”,
i.e. of the true ventrals.
Table 7.
Negative allometric growth of the eye in Python sebae.
¢13.50 5.00 2.70
13.60 4.90 2.78
13.60 4.80 2.83
13.80 4.80 2.88
¢14.00 5.10 2.75
¢14.10 5.10 2.76
? 29.40 7.40 3.97
? 46.10 8.90 5.18
Measures (in mm); SnL = snout length, i.e. distance between snout tip and eye; EHD = horizontal
eye diameter.
108 B. Lanza and A. Nistri
109Somali Boidae and Pythonidae
Fig. 22. Python sebae unsexed specimen MZUF 21651 from Jilib;
preserved head; head, dorsal view: r, rostral; in, internasal; apf,
ante-rior prefrontal; ppf, posterior prefrontal; f, frontal; aso, anterior
supraocular; pso, posterior supraocular (photo by Benedetto Lanza).
Fig. 23. — Python sebae young MZUF 5653 from Giohàr; preserved
specimen; head and neck, dorsal view (photo by Benedetto Lanza).
Fig. 24. — Python sebae young MZUF 2202 from Alessandra Island; pre-
served specimen; head and neck, dorsal view (photo by Benedetto Lanza). Fig. 25. — Python sebae young MZUF 27045 found between Labadaad
and Bulo Yaag; preserved specimen; head and neck, dorsal view (photo
by Benedetto Lanza).
110 B. Lanza and A. Nistri
111Somali Boidae and Pythonidae
ferent from that of P. natalensis (BROADLEY 1984): a diagonal pale stripe between
nostril and 2nd upper labial, followed by an extensive preocular dark patch; a large
pyriform dark postocular patch, broader than the eye diameter posteriorly; a large
subocular subtriangular or irregular dark patch, occasionally conuent with the
preocular and/or the postocular dark patch; head always immaculate white ven-
trally, usually with solid dark markings on the lower labials and large dark blotches
on the throat. Underparts white, profusely spotted and dotted with dark brown to
black posteriorly to the throat, the light colour predominant on the median line;
subcaudal surface irregularly dark-mottled and, according to BROADLEY (1984),
without the continuous light median line occurring in P. natalensis (but such a line
exists, at least in the basal part of the tail of the individual gured in DOUCET (1963:
226, plate V); as its provenance is unspecied, it could not be from Ivory Coast,
therefore possibly a P. natalensis). Good colour illustrations of the species may be
found in PITMAN (1938, 1974), KUNDERT (1974) and WALLS (1998a; also an unusual
pattern variant).
The eggs are large: on average they measure approximately 10 cm in length
and weight about 150 g; when rst laid, they are soft, oval to sub-elliptic in shape
and ashy grey to ivory white in colour; on exposure they soon become almost spher-
ical, rm and clear white in colour; later they become leathery and turn to a dirty
creamy white shade. However, the egg size may be smaller: measurements of 13
eggs made by BARKER & BARKER (1994) ranged from 7.4-7.9 cm in length (average
7.6 cm) and 5.8-6.3 cm in diameter (average 6.1 cm). Their shell is pergameneous,
nely granulous and, as in the other pythons, its calcareous material is formed
into columns that penetrate deeply into the membrane; the calcium carbonate usu-
ally consists of calcite crystals, but in some species, e.g. in Python regius, aragonite
crystals are interspersed among the calcite ones (PACKARD & PACKARD 1994). Soon
after the eggs are laid, they tightly adhere to each other and become very difcult
to separate without tearing the shells (STAFFORD 1986).
Distribution. Africa south of the Sahara from Senegal east to southern Sudan,
Eritrea, Ethiopia and Somalia, including Guinea-Bissau, southern Mali, Guinea,
Sierra Leone, Liberia, Ivory Coast, Burkina Faso, Ghana, Togo, southern Niger,
Nigeria, Cameroon, Equatorial Guinea, southern Chad, Central African Republic,
Congo Republic, Democratic Republic of Congo, Uganda, northern Angola, Rwan-
da, Kenya, and Tanzania (see also BROADLEY 1984, and BROADLEY & HOWELL 1991).
In Kenya, P. sebae occurs at relatively low altitudes in the Lake Victoria basin to
the west and in the eastern coastal lowlands, being replaced by P. natalensis on the
Kenya highlands (BROADLEY 1999).
According to KÁDÁR (1977: 44), Diodorus Siculus (1st century BC) «relates of
the Libyan amazones; speaking of the armour of the belligerent people he mentions
the fact that they use as defensive armour the skins of great snakes (most prob-
ably shields made of the skins), which occur in great number in Libya; most likely
he thinks here of the assala [the usual Arabic name of the python] (Python sebae)
[…]». Also the giant snake, whose capture is vividly described by Diodorus Siculus,
who derived his data from Agatharchides of Cnidus (2nd century BC; cf. BODSON
2003), was almost surely a Python sebae (cf. VOGEL 1996: 86). It had a length of 30
cubits, i.e. of 13.2 m, almost surely an overestimated gure; this animal, collected
in “Aithiopia” (at present corresponding to an area situated between most of Sudan
and northern part of modern Ethiopia) was donated to Ptolemy II, King of Egypt,
and lived in his Alexandria palace. KÁDÁR (1978), after a zoological evaluation of
110 B. Lanza and A. Nistri
111Somali Boidae and Pythonidae
Fig. 26. Python sebae young MZUF 2202 from Alessandra Island; preserved specimen; head and
neck, side view (photo by Benedetto Lanza).
Fig. 27. Python sebae young MZUF 5653 from Giohàr; preserved specimen; head and neck, side
view (photo by Benedetto Lanza).
112 B. Lanza and A. Nistri
113Somali Boidae and Pythonidae
the north-western African Roman mosaics (ca 2nd to 4th century AD), points out
the details of zoogeographical changes, like the extinction of numerous Afrotropical
elements, e.g. large mammals, ostrich and Python sebae, having taken place in his-
torical times; a mosaic found near Carthage represents a python struggling with an
African elephant. Another two big snakes, presumably P. sebae, are portrayed in the
upper level of the famous Nile mosaic of Palestrina (near Rome; 2nd century BC);
one is coiling around a rocky outcrop, the other, in ambush on the Nile bank, has
just caught a bird in its mouth (BODSON 2003). However, the species still occurred
in the Saharian S Algeria at least at the beginning of the 20th century (ANGEL &
LOTHE 1938).
Seba’s python rarely extends to altitudes much above 2000 m (LARGEN & RAS-
MUSSEN 1993).
Geographical variability. The species seems to be only moderately variable, but
large-scale research on the topic is lacking; some anomalous colourations and pat-
terns obtained in captivity have been mentioned under Colouration.
Biology (see also under Python, Biology). BROADLEY validated P. natalensis at
the subspecic level in 1983 and 1984, but at the specic level only in 1999. Con-
sequently, before 1983-1999 the natural history of P. sebae was not distinguished
from that of P. natalensis and at present, it is possible to discriminate among the
available data only when the provenance or other relevant details of the studied
Fig. 28. Python sebae young ¢ MZUF 2216 from Alessandra Island; preserved specimen (photo by
Benedetto Lanza).
112 B. Lanza and A. Nistri
113Somali Boidae and Pythonidae
material are indicated. Unless otherwise stated, information given in this paragraph
usually refers to or may apply also to the “true” P. sebae. As pointed out by STARIN
& BURGHARDT (1992), despite their notoriety, the natural history, ecology, and behav-
iour of large free-ranging snakes in native habitats are so poorly known that, the
most “sustained” treatment of African rock python natural history is to be found in
a popular book by F.W. FITZSIMONS (1930), but this refers to P. natalensis.
Python sebae is a rather sedentary, euryokous, good tree-climbing and mark-
edly hygrophilous species widespread in a great variety of sub-Saharan habitats but
seldom if ever occurring far from permanent water. In Somalia it has been found
in many different ecosystems which, according to the geobotanical map published
by PICHI-SERMOLLI (1957), are characterized by the following types of vegetation:
broken xerophilous open woodland; xerophilous open woodland; lowland ever-
green thicket and scrub; lowland dry evergreen forest; lowland moist evergreen
forest; riparian formations. Though not yet conrmed, its occurrence in Nogal (cf.
LIPPARONI 1954, under Somali records) is very likely since this particularly arid
region of northern Somalia is crossed by the Wadi Nogal, a watercourse almost
always dry but with numerous perennial pools, some of which are relatively large
and deep, surrounded by thick riparian formations, inhabited by helmeted turtles
[Pelomedusa subrufa (Lacépède 1788)] and assiduously frequented by potential prey
like ocks of sheep and goats. STARIN & BURGHARDT (1992: 53-54, table 2), who car-
ried out a study in Gambia, wrote: «Neglecting relative area, most animals were
found in the clearing, with the tree and shrub savanna the second most popular
Fig. 29. Python sebae young MZUF 2202 from Alessandra Island; preserved specimen (photo by
Benedetto Lanza).
114 B. Lanza and A. Nistri
115Somali Boidae and Pythonidae
area, and the woodland savanna the least popular […]. If the number of snakes is
adjusted by the relative area of the habitats in the reserve, the relative popularity of
habitats is even more marked […]. The preference for the clearing was thus much
stronger than even the table suggests, since the clearings make up less than 1% of
the total habitat yet contributed 40% of all python observations». Python sebae is
also rather frequent near or even within residential areas, where its presence is
often completely overlooked because of its unobtrusive and largely nocturnal hab-
its. Being excellent swimmers, Seba’s pythons spend a lot of time in water, in which
they may lie submerged with only the nostril and eyes exposed above the surface,
often waiting for prey close to the bank. Because of their swimming ability they are
probably able to cross sea stretches of some kilometers, which could explain pos-
sibly allochthonous occurrence on the small coastal Ciula Island (Bajuni Islands;
S Somalia). PAKENHAM (1983: 22), dealing with the herpetofauna of Zanzibar and
Pemba islands, wrote: «Once one was killed near Pemba’s east coast but I never
heard of another wild python in Pemba and suspect this one was escaped […] or
else had crossed the strait from mainland Africa». Juveniles are perhaps more ter-
restrial (STAFFORD 1986). They are usually nocturnal, but will bask and hunt oppor-
tunistically during the day; when inactive, they shelter in holes, especially aardvark
[Orycteropus afer (Pallas 1766)], warthog [Phacochoerus aethiopicus (Pallas 1766)]
and porcupine [Hystrix Linnaeus 1758] burrows, in thickets, reedbeds, up a tree, in
a rock ssure or under water, often curled up in a heap of coils, with the head rest-
ing on top (SPAWLS et al. 2002).
Python sebaes diet consists mainly of warm-blooded animals, which it may also
prey upon at night, guided by the heat receptors situated in its rostral and labial sen-
sory pits; reptiles, amphibians and shes are also occasionally eaten. Some dietary
records are quoted hereunder. SCHMIDT (1923; Democratic Republic of Congo): a
specimen had eaten a rat, and another one, whose size is not reported, a female ante-
lope measuring 1040 mm in length and 480 mm high at the shoulder (personal com-
munication by HERBERT LANG). LIPPARONI (1954; Somalia): sight record of a specimen
caught swallowing a big Naja Laurenti 1768 sp. [probably a Naja nigricollis Reinhardt
1843]. PITMAN (1958: 103): «On an island in Lake Victoria, near Kisumu (Kenya),
Meinertzhagen and a companion watched a huge python, which he eventually shot,
high up in the trees swallowing the nestlings of Anhinga and Phalacrocorax […] Mr
E. Wilson (in litt.) caught a small python at a large, dry-season Quelea roost in the
Sudan which disgorged about six Quelea» […] in the permanent freshwater swamps
in the ood area of Sierra Leone the python preys on various waterfowl as the Afri-
can jacana [Actophilornis africanus (Gmelin 1789)] and the pigmy goose [Nettapus
auritus (Boddaert 1783)]. CANSDALE (1961: 19): «Pythons occasionally swallow por-
cupines. In 1948 the Dublin Zoo received an African Python 12 ft [= ca 3.65 m] long
which had been in captivity for three months. A year after arrival it passed several
long quills among its dropping, showing that it had swallowed a porcupine over 15
months earlier». PITMAN (1962) reports as prey: the Guinea-fowl [Numida meleagris
(Linnaeus 1758)], sparrow weaver [Ploceopasser superciliosus (Cretzschmar 1827)],
rock bantam [Ptilopachus petrosus (Gmelin 1789)], and an unidentied Ploceidae in
Sudan; Egyptian goose [Alopochen aegyptiacus (Linnaeus 1766)], in Kenya. GRUS-
CHWITZ et al. (1991) mention as prey in Gambia the bushbuck [Tragelaphus scriptus
(Pallas 1776)]; in the same country, near or within the small zoo of the Abuko Nature
Reserve, STARIN & BURGHARDT (1992) ascertained predation on an unidentied caged
Gazella Blainville, 1816 sp., bushbuck, patas [Erythrocebus patas (Schreber 1775)],
Temmincks western red colobus [Procolobus badius temmincki (Kuhl 1820)], and
114 B. Lanza and A. Nistri
115Somali Boidae and Pythonidae
Fig. 30. Somali distribution of Python sebae. The empty circle refers to an undened locality and
indicate the approximative geographical centre of a more or less wide area. 1, Nogal [according to
the natives (CORONARO 1926), but not seen by the author]; 2, Lugh; 3, Giohàr; 4, Bardera; 5, Mogadi-
shu; 6, Comia; 7, Arbidin; 8, Labadaad (= Alessandra Island); 9, Jilib; 10, Mareri; 11, Bulo Yaag; 12,
Belet Amin; 13, Manamofa; 14, Kismayu; 15, Ciula Island; 16, Ola Uager.
116 B. Lanza and A. Nistri
117Somali Boidae and Pythonidae
green monkey [Cercopithecus aethiops (Linnaeus 1758) sensu lato]. GREENE & MIN-
TON (2002: 185, caption of the lower g.): «[…] (Python sebae), which habitually feeds
on ungulates, including Thomson’s gazelle [Gazella thomsoni Günther 1884] and even
larger antelopes like the kob [Kobus kob (Erxleben 1777)] and impala». SPAWLS et al.
(2002: 310): Seba’s pythons «[…] eat a range of prey. Juveniles will take small mam-
mals, birds and frogs, moving to larger mammals; adults in game areas are signi-
cant predators on antelope such as impala, duiker, kob and gazelle and the young of
larger species such as waterbuck. Other known items include sh, lizards, including
monitors [Varanus], warthogs, monkeys, crocodiles, baboons, hyrax, spring hares
[Pedetes capensis (Forster ....)], porcupines, various game and waterbirds, including
pelicans; they will raid chicken runs and rabbit hutches and in farming areas they
will take goats and sheep». When hunting, pythons usually sneak up on the prey
but when near it stop and lie in ambush, violently projecting forward only when the
intended victim enters within their striking distance. The prey is usually seized by
the neck and, as it is pulled back, the snake forms coils that snap shut around the
victim, squeezing it until it asphyxiates. The swallowing process, which almost invari-
ably starts from the head, is greatly facilitated by the extreme expansion of the jaws,
but also by the emission of a great deal of saliva and the prey’s elongation due to the
previous constriction. Serious injuries, occasionally produced by horns or porcupine
quills piercing the oesophagus or stomach and even the skin, usually heal without
any ill effects since the very active digestive juices rapidly dissolve the internal part of
the prey while its protruding part eventually drops off. A disturbed or frightened ani-
mal invariably disgorges its meal. Pythons can fast for very long periods.
Especially small pythons have a lot of enemies, such as many species of mam-
mals, raptors, and reptiles, but medium-sized and big adults are preyed upon only
by a few non-human species: lions [Panthera leo (Linnaeus 1758)], leopards [Pan-
thera pardus (Linnaeus 1758)], and crocodiles (cf. GREENE 1994). Crocodylus niloti-
cus Laurenti 1768 attacks swimming pythons, killing and consuming them. STARIN
& BURGHARDT (1992: 58), who studied P. sebae in Gambia, wrote: «On one occasion
a 3 m python moved through a puddle containing a group of 6 dwarf crocodiles
[Osteolaemus tetraspis Cope] (ranging in size from .2-1.2 m). No interaction was
observed. After this study was completed, reliable sources stated that a python was
attacked and killed, but not eaten, by a dwarf crocodile in a puddle in the savanna.
[…] it certainly seems possible that middle to large size dwarf crocodiles could prey
on neonate and young pythons». However the number one enemy of the python is
surely man who acts both indirectly [habitat destruction or alteration; use of shing
nets and shing traps (for the latter see CANSDALE 1955, GRUSCHWITZ et al. 1991)]
and directly, killing them for their valuable skin or because they threaten poultry
and livestock in general, or collecting them for zoos, scientists and an ever-increas-
ing number of amateurs. Some peoples eat them extensively or use parts of their
body for medical purposes; BUTLER & SHITU (1985) reported that in Nigeria their
meat is used as food, whereas the fat of the abdominal cavity is used as food and
also as a rubbing ointment said to soothe muscle aches and sprains, arthritis, and
rheumatism. SIMONETTA & MAGNONI (1986), dealing with the status and conservation
problems of lower vertebrates in Somalia, pointed out that in this country Python
sebae has been decimated by the combined effect of hunting for an underground
skin market and destruction of wet habitats, particularly of riverine forests, which
have almost completely disappeared. SCHMIDT (1923: 55) reported that pythons in
the Democratic Republic of Congo «[…] are often caught by means of traps set at
their holes» and that «During the rainy season the natives follow their tracks, and if
116 B. Lanza and A. Nistri
117Somali Boidae and Pythonidae
they nd the retreat a noose is at once set in front of the hole. The snake is usually
caught behind the neck. The natives also spear them (H. Lang)».
Unlike other species, e.g. P. regius and P. molurus, Seba’s pythons are bad-tem-
pered, very aggressive animals, and only a few specimens tame down in captivity. In
the wild they usually try to escape if confronted, but if cornered they strike vicious-
ly and can inict severe traumas, such as bruises, deep cuts and wounds, with their
snout and powerful dentition. SPAWLS et al. (2002: 310) reported that «A herpe-
tologist in South Africa required 57 stitches after a python bite and a well known
Kenyan vet lost his left eye after a python struck him in the face». Specimens about
4 m in length are already capable of constricting a well-built man; however P. sebae
and P. natalensis rarely if ever — eat humans; reported of a fatal attack by a
python and cases where humans were the prey have been critically summarised by
BRANCH & HAACKE (1980) and HAACKE (1981). SPAWLS et al. (2002: 310) refer what
follows: «ARTHUR LOVERIDGE (1933) records a possible fatal attack on a woman on
Ukerewe Island in Lake Victoria (P. sebae); the victim, who was in ill health, had
disappeared and was found dead in the coils of a 4.5 m python, which might have
killed her. In general, humans are too big a prey for pythons, although unattended
small chldren in python country are at risk. A 4.8 m python swallowed a 59 kg
impala; many humans weigh less than that». STARIN & BURGHARDT (1992) observed
in Gambia that pythons often allowed humans or animals to approach closely,
either freezing, remaining motionless, continuing ongoing behaviour, or just mov-
ing away slowly; moreover loud noises, vigorous movements, and other attempts to
encourage the snake to depart often had little effect. GREENE (1994) listed the fol-
lowing antipredator mechanisms in pythons: crypsis, hiss, S-coil (the head and neck
are retracted in a striking posture), strike, bite, disgorge meal (this might be simply
a side effect of fear), and tail display. P. sebae employs two types of defensive caudal
displays, coiling the tail into a at disk (STEMMLER 1971, STARIN & BURGHARDT 1992)
or, according to observations on a captive Liberian specimen, vibrating it violently
in front of its snout in the manner of a rattlesnake (SENTER 2000). Finally, unpro-
voked pythons are not dangerous to man and deserve to be strictly protected, as, by
eating many rodents, they are strongly benecial to agriculture.
Pythons are often associated with superstition by members of many indig-
enous native peoples and, consequently, are not interfered with or killed in many
parts of Africa (FITZSIMONS 1962). In some areas they are venerated; a common
belief is that if they are killed, rain will not fall (SPAWLS et al. 2002). A very long
series of Cameroonian beliefs and tales concerning P. sebae has been listed by
STUCKI-STIRN (1979: 129-131). In the Sudan it is a belief that the claw-like vestigial
hind-limbs are thrust into the nostrils of the prey to suffocate it (CORKILL 1935: 12).
Data on the reproductive biology, difcult to obtain in the eld, are relatively
scarce as Seba’s pythons «are not common in captivity, perhaps because of their
large size, bad tempers, and general resemblance to the much calmer P. molurus»
(WALLS 1998a: 169). On the other hand, in addition to the possibility that some of
the available data could refer to P. natalensis, it is necessary to take into account
that all phases of python reproduction and growth are largely dependent on envi-
ronmental (especially temperature and relative humidity) and individual (e.g. age,
size and diet) conditions, which at least partly explain some discordant information
in the literature.
STARIN & BURGHARDT (1992), studying the social behaviour of P. sebae in
Gambia, observed (only in winter) an apparently unreported type of behaviour in
groups of two or more snakes (up to six) that were largely oblivious to the observer.
118 B. Lanza and A. Nistri
119Somali Boidae and Pythonidae
On March 1, at 14:15 in the sun in a clearing, two pythons (about 3 and 3.5 m long)
were found coiled around each other; the following day, at 12:30 and in the same
conditions, six pythons (2.1 to 6 m long) were observed while actively sliding over
and around each other, and afterwards moving away one after another within about
20 min. On March 3, in the same conditions, two to four animals were observed
tightly wrapped around each other at least between 07:30 and 12:15 hr (none present
at 15:30). On January 23, at 11:00, in the woodland savanna, two 3.5 m pythons were
observed entwined in an S shape; when approached to within a meter, one of them
moved off but the other remained stretched out for at least 10 min. STARIN &
BURGHARDT supposed that these incidents during the dry season could best be
interpreted as courtship, but at the same time they pointed out that it was impossible
to sex any specimen during their study, as the presence of spurs in males is hard to
detect in uncaught animals in the eld. This allows one to speculate that the animals
seen by STARIN & BURGHARDT were actually males engaged in premating ritualized
struggles. But this hypothesis can be excluded, as the ghting behaviour of P. sebae
appears to be completely different; BARKER & BARKER (1994: 20) described it as
follows: «No biting was observed, but there was head raising, topping and throwing,
body wrapping and constricting, tail wrapping and spurring, lots of pushing and
shoving, all the components of a good wrestling match». Copulation, by a 3.75 m
long and a meter shorter ¢, owned respectively for 6 and 12 years, was seemingly
rst described by MUNNIG SCHMIDT (1971), who observed it between March 18 and 24,
in a terrarium with temperature of 25-27 °C. Further research has conrmed the
observations by MUNNIG SCHMIDT (1971) on captive animals and those by STUCKI-
STIRN (1979: 50) and STARIN & BURGHARDT (1992) in the eld: the breeding season
takes place in winter (see also SCHMIDT 1989: in February-March and November;
WALLS 1998a: in February), so that mating can be induced in bred animals by
reproducing winter conditions, i.e. by reducing the day-time photoperiod to 8-10 hr
and dropping the day-time and night-time temperatures respectively to 27-30 and 24-
27 °C, but never under 20 °C (BARKER & BARKER 1994, FAIN 2001, VIANELLO 2002,
MCCURLEY & GLASGOW 2003). However, exceptions occur, since, P. sebae is reported to
mate and lay eggs throughout the year around Lake Victoria (SPAWLS et al. 2002). In
any case, such discordances could be related, at least partly, to the fact that the
female can also reproduce long after the last copulation, being able to store fertile
spermatozoa in her body for longer than 4 years (VISSER 1985). In captivity gestation
lasts 30-40 days (FAIN 2001), about 44-50 days (MCCURLEY & GLASGOW 2003), 51-54
days (BARKER & BARKER 1994), about 70 days (MUNNIG SCHMIDT 1971), or even 4
months (VIANELLO 2002); its duration depends on many factors, especially
temperature and humidity (VIANELLO 2002). The same applies to incubation, lasting
60-80 days at a temperature of 28-32 °C (SLAVENS 1984, BARKER & BARKER 1994, FAIN
2001, VIANELLO 2002, MCCURLEY & GLASGOW 2003); according to STAFFORD (1986) the
optimum temperature range for incubation is 28-32 °C at a relative humidity of 90-
100%, and under these conditions the eggs can be expected to hatch some 50 to 75
days after they are laid. POPE (1961) reported that at 25-34 °C a clutch required
almost 100 days of incubation. DOUCET (1963) mentioned an incubation period of 8
weeks. SCHMIDT (1989) gave different and seemingly questionable gures: 75-109 days
at temperatures of 30-32 °C. In the wild, in Cameroon, mating occurred in December
and the eggs were found in February and March (STUCKI-STIRN 1979: 50). WALLS
(1998a) reported that incubation lasts 52-79 days in captivity and about 2 months in
nature. Places selected by the female to oviposit, usually moist ones, include secluded
cavities beneath objects on the ground, deep crevices and other natural underground
118 B. Lanza and A. Nistri
119Somali Boidae and Pythonidae
cavities, abandoned burrows (especially of warthogs, aardvarks, and porcupines) or
hollows in termite mounds, but in the absence of such sites she may lay her eggs on
the surface of the ground in tangled undergrowth, dense grass or pockets of dead
leaves (POPE, 1961, STAFFORD 1986, PACKARD & PACKARD 1994, WALLS 1998a). The eggs
are deposited in a single mass up to 30 cm in height and covering an area of about
40-60 cm in diameter; their number, usually related to the femal’s size, ranges
between 16 and 100, but more than 80 can be regarded as a very high number of
eggs from one female (BENEDICT 1932, DOUCET 1963, MUNNIG SCHMIDT 1973a,
STAFFORD 1986, WALLS 1998a, CHIPPAUX 2001, VIANELLO 2002). The following gures
refer to 8 Cameroonian specimens, e.g., have been found: a 305 cm long laid 27
eggs; 312 cm, 32 eggs; 420 cm, 43 eggs; 456 cm, 54 eggs; 543 cm, 34 eggs; 568 cm, 42
eggs; 617 cm, 65 eggs; 678 cm, 73 eggs (STUCKI-STIRN 1979: 50). Immediately after
oviposition the female coils round her clutch and remains so, except for almost daily
visits to the nearest water, until a few hours before the eggs are due to hatch; at least
in captivity, she may leave the eggs for a few hours, more and more frequently as
hatching gets nearer, to bask in the sun or near a heater (MUNNIG SCHMIDT 1973a).
This type of parental care, practically occurring only in pythons, serves both to
protect the eggs from predators (mongooses, rats, some reptiles, birds of prey, etc.)
and to aid in incubation of the eggs. In support of the latter purpose, it has been
observed in some species of pythons that the temperature of the brooding female
rises by as much as 6.5 °C above the ambient air temperature, to reach as high as 36
°C. This is due to metabolic heat production resulting, at least in some species (e.g.
in Python molurus), from intermittent, spasmodic contractions of the body
musculature (shivering thermogenesis); the shivering frequency is inversely related to
the ambient temperature and directly related to oxygen consuption (BARTHOLOMEW
1982), which increases as much as tenfold above resting levels (BENNETT 1982). In
short, whereas males and nonbrooding females act like other ectotherms, the females
brooding eggs or preparing to oviposit behave like endotherms, since their metabolic
response to temperature is qualitatively similar to that of birds and mammals
(BARTHOLOMEW 1982; see also SCHIERECK 1989 concerning thermoregulation in
snakes); moreover BARTHOLOMEW (1982: 201) considers that «Although only
fragmentary and anecdotal data are available for boids other than P. molurus
(VINEGAR et al. 1970), it would be surprising if endothermic incubation did not occur
in other species of large size […]». However, this does not seem to be the case for P.
sebae (e.g., cf. VINEGAR et al. 1970, PITMAN 1974, SCHMIDT 1989, BARKER & BARKER
1994, and WALLS 1998a), despite the statement to the contrary by SHINE (1994: 289,
table III); in fact, accord-ing to our audit, in SHINE’s statement appears to be based
on a wrong interpretation of BENEDICT (1932) and NOBLE (1935). Other authors found
that the body of the brooding female was warmer than the ambient temperature (e.g.,
MUNNIG SCHMIDT 1973a: seemingly about 3 °C warmer at an ambient temperature of
25-27 °C) but did not give any indication of muscular contractions. SPAWLS et al.
(2002: 310) stat-ed that «the rock python is not known to shiver but a Cameroonian
Central African Rock Python [i.e. a “true” P. sebae] was found to have elevated its
temperature by 6.5 °C while brooding»; if this statement refers to STUCKI-STIRN (1979:
154) it does not concern a «Cameroonian» specimens as Stucki-Stirn’s text is nothing
but the almost word-for-word quotation of clearly very general statements by
FITZSIMONS (1962: 67, right column)! Usually, if not always, each clutch contains a
variable number of sterile eggs, up to 50% according to FITZSIMONS (1962). However,
even fertile eggs, may fail to develop due to different types of stress; BRANCH &
PATTERSON (1975) found, in captive P. natalensis from Transvaal, that a particularly