Litocranius walleri ( Brooke, 1878 )

Bärmann, Eva V, Klappert, Ariana N & Chen, Augustin, 2021, Litocranius walleri (Artiodactyla: Bovidae), Mammalian Species 53 (1005), pp. 65-77 : 65-70

Litocranius walleri ( Brooke, 1878)

Gerenuk

Gazella walleri Brooke, 1878:929 , pl. LVI. Type locality “mainland of Africa, north of the island of Zanzibar, about lat. 3° S. and long. 38° E ”; restricted to the vicinity of Chisimaio by Grubb (2002; see “Nomenclatural Notes”).

Litocranius (Gazella) walleri: Kohl, 1886:79 . First use of current name combination.

Lithocranius sclateri Neumann, 1899:21 . Type locality “das nördliche Somali-Land [= northern Somalia]” (see “Nomenclatural Notes”).

Lithocranius walleri sclateri: Rothschild 1902:473 . Name combination.

CONTEXT AND CONTENT. Context as for genus. Litocranius walleri has two subspecies ( Grubb 2002).

L. w. sclateri Neumann, 1899 . See above.

L. w. walleri ( Brooke, 1878) . See above.

NOMENCLATURAL NOTES. In the original description of the genus, Kohl (1886) did not comment on the etymology of the name. However, he emphasized the extraordinary elongated and flattened appearance of the skull compared with gazelles that justified generic distinction. Later authors assumed that the name was based on the “solid stony character of the cranium (λιθοs = lapis [gemstone])” (Sclater and Thomas 1898:227). Therefore, they concluded that there was a misprint in the original description and changed the name to Lithocranius . In 1933, Krumbiegel clarified the situation by demonstrating that the species was named according to its long and flattened skull (λιτός = plain, modest, simple, frugal— Krumbiegel 1933). The name appears twice in the original description, so a misprint seems unlikely.

In the original species description, Brooke (1878:929) gave the type locality as “mainland of Africa, north of the island of Zanzibar, about lat. 3° S. and long. 38° E, ” therefore in Kenya. Sclater and Thomas (1898) refer to their correspondence with Sir John Kirk, who had employed the hunters who collected the type specimens (one of the three skulls is preserved in the Natural History Museum in London, BM [British Museum of Natural History] 1897.7.29.1— Lydekker 1914). Kirk stated that they were collected “on the coast near the River Juba in Southern Somaliland” (Sclater and Thomas 1898:230) and that he later shot other specimens at the same location. One of these specimens collected by Kirk is also preserved in the Natural History Museum, with the locality Kismayu (now Kismayo), a town located near the mouth of the Jubba River. Grubb (2002:539) therefore restricted the type locality of Litocranius walleri to “the vicinity of Chisimaio” in southern Somalia. Nevertheless, not all authors followed the argument of Sclater and Thomas and still cite the original type locality in Kenya (e.g., Neumann 1899; Schomber 1963).

Sclater (1884) states that the type specimens of L. w. sclateri (two skins) came from “ Somali -land, vicinity of Berberah.” The name gerenuk derives from the Somali word “garanuug.”

DIAGNOSIS

Litocranius walleri ( Fig. 1 ) differs from the closely related gazelles and springbok by its very elongated legs, neck, ears, and skull (especially in the part behind the horns— Brooke 1878; Kohl 1886; Sclater and Thomas 1898). The skull ( Fig. 2 ) has a flattened skull-roof, hardly curving down toward the back as in gazelles. The extremely elongated supraoccipital bone reaches the length of the parietals. It forms a cone-shaped hump that extends beyond the sloping part of the occiput ( Kohl 1886). The braincase of males is heavily ossified and carries extremely heavy horns ( Kingdon 1982). L. walleri is the only bovid with lowcrowned teeth (hypsodonty index = 1.32— Janis 1986). Contrary to all other gazelles (exception: mountain gazelle [ Gazella gazella ]), the premaxillae of L. walleri do not reach the nasals ( Kohl 1886). The closely related dibatag ( Ammodorcas clarkei ) can be regarded as showing an intermediated morphology between gazelles and L. walleri , in terms of skull elongation and tooth reduction ( Thomas 1891; Lydekker 1914). The dibatag can, however, be easily distinguished from L. walleri by its very different horn shape (horns curve backward and then upward and forward, similar to the horns of reedbuck [ Redunca ]). L. walleri is unique in its feeding behavior, as it frequently adapts a bipedal stance for reaching browse at up to 2 m off the ground ( Menges 1887; Sclater and Thomas 1898).

GENERAL CHARACTERS

Litocranius walleri has characteristic facial markings ( Fig. 1 ): a broad white ring around the eyes; black preorbital glands anterior to the eyes; white lips, chin, and throat; and long ears with black-and-white markings on the inside. A dark-brown stripe extends from between the ears and over the neck; the stripe broadens to a saddle along the entire rump, before narrowing again toward the tail. The flanks and outer parts of the legs are light brown, and the belly and inner parts of the legs are white. A thin light stripe can often be found between the saddle and the flanks. L. walleri has a narrow, white rump patch and dark brown to black knee-tufts. The tail is white near the base and has a black tip, and the underside is black and hairless ( Sclater 1884; Schomber 1963; Estes 1992). Grubb (2002:541) notes that seven skins from Tanzania, Kenya, and southern Somalia have a “band of reversed hairs directed towards the head on the dorsal midline of the neck,” and that he never observed this character in specimens from northern Somalia, which could indicate a potential differentiation of the two subspecies.

There is a marked sexual dimorphism in L. walleri . Males are larger and more massive, especially in the anterior body region. The neck of males has the same width throughout the entire length, whereas the neck of females is generally more slender and decreases in width toward the head ( Ullrich 1963). Males have heavy lyre-shaped horns (32–44 cm) with strong annulations that start to grow relatively late during ontogeny (after about 5 months); females are hornless (Sclater and Thomas 1898).

External measurements (cm) of L. walleri were: total length (snout to tail), 165–195; head–body length, males 155–160 and females 140–155; tail length, 22–35; height at withers, males 89–105 and females 80–100. Weight varies from 31.2 to 52 kg in males and 28.5 to 45 kg in females ( Schomber 1963; Kingdon 1982). The northern subspecies, L. w. sclateri , is noticeably larger than the nominate subspecies (Sclater and Thomas 1898; Neumann 1899; Schomber 1963). L. w. walleri has shorter nasals and premaxillae, a more reddish coat color, and black knee-tufts, as opposed to brown knee-tufts in L. w. sclateri . A geographic cline from north to south may exist in these traits ( Neumann 1899; Schomber 1963).

Measurements for 44 skulls from various museum collections (London, Florence, New York, Nairobi) were provided by Schomber (1963). Grubb (2002) reused these data together with measurements from additional specimens from the Natural History Museum, London and provided the mean ± SD (mm) for the two subspecies according to his definition of their geographic boundaries (see “Distribution”). Measurements for L. w. walleri (24 males and seven females) and L. w. sclateri (10 males and five females), respectively, were: greatest length of skull males, 231.58 ± 6.68 and 251.25 ± 4.35; greatest length of skull females, 210.86 ± 5.27 and 227.30 ± 3.38; nasal length males, 58.57 ± 6.40 and 69.45 ± 3.21; nasal length females, 53.29 ± 4.57 and 61.54 ± 6.71; premaxillary length males, 58.06 ± 3.11 and 65.58 ± 2.91; premaxillary length females, 53.00 ± 3.00 and 59.92 ± 2.99. Neumann (1899) also postulated a difference in toothrow lengths between the two subspecies, but Schomber (1963) did not confirm this. Mean tooth measurements are thus presented for all specimens combined (mm; range in parenthesis): length of upper molar row, 30.8 (28–34.3); length of upper premolar row, 21.2 (19–23); length of lower molar row, 33.2 (31–37.7); length of lower premolar row, 19.3 (17–22.3).

Limb bone lengths (mm) were provided by McMahon (1975; mean ± SD) and Christiansen (1999; mean values in parentheses): humerus, (165.0); ulna, 175 ± 19; radius, (221.0); femur, 213 ± 20.5 (207.0); tibia, 282 ± 18.7 (266.0); and metatarsus, 279 ± 14.8. McMahon (1975) also provided a measurement for the length of the humerus (276 ± 16.2 mm); however, this was probably a mistake (could be 176?) because the humerus is markedly shorter than the radius in Litocranius .

DISTRIBUTION

Litocranius walleri is native to the Somali-Masai arid zone from southern Djibouti, eastern Ethiopia, Somalia, and Kenya to northwestern Tanzania ( Leuthold 1971; IUCN SSC Antelope Specialist Group 2016) at elevations below 1,200 m ( Fig. 3 ). It presumably still occurs throughout its historical distribution, although recent information is not available from all countries inhabited by L. walleri . It is debated whether a gap in the species’ distribution exists. Schomber (1963) stated that a gap exists in the distribution of the species between the Tana and Sabaki rivers in eastern Kenya, and that this gap separates the two subspecies of L. walleri , L. w. sclateri north of the Tana, and the smaller L. w. walleri south of the Sabaki River. Leuthold (1971) clarified that the area of the potential species distribution gap is simply understudied due to its difficult accessibility. He cited personal communications with game wardens that confirm the presence of L. walleri throughout the area. Leuthold also studied L. walleri in Tsavo East National Park that lies north of the Sabaki River. Grubb (2002) postulated that L. w. sclateri only occurs in northern Somalia and Ethiopia, separated from L. w. walleri by the mountain range extending from the Churcher Highlands of Ethiopia to the Goolis Mountains of northern Somalia.

A petrograph exists at Hôsh (Wâdi Sab’er Rigâl) near Silwah on the eastern bank of the Nile River in Upper Egypt. It shows a man holding a L. walleri on a leash and was dated to 4,000 –2,900 BCE (6,000 –4,900 years ago— Winkler 1939). This coincides with an African Humid Period (11,700 –4,400 years ago) when subtropical North Africa was much greener than today ( Claussen et al. 2017). L. walleri , in association with other elements of the tropical fauna from Sudan, was able to extend its distribution into the Sahara, until the subsequent aridification lead to a shrinking of its distribution to its present size ( Schomber 1963). No fossils of L. walleri are known.

FORM AND FUNCTION

Form. — Litocranius walleri has the typical glands of gazelline antelopes: preorbital glands (very prominent in males, excreting a thick oily secretion), carpal glands with knee-tufts, and interdigital glands ( Estes 1992; Sokolov et al. 1994). Primary and secondary hair follicles are present in the integument at a ratio of 1:1 and are interspersed. The secondary hairs are very short, and their bulbs do not extend below the level of the bilobular sebaceous glands associated with the primary hair follicles. The sweat glands of L. walleri have convoluted shapes and are similar in size to those of oryx ( Oryx beisa ) and lesser kudu ( Ammelaphus —Jenkinson and Nay 1975). A detailed histological description of the integument of L. walleri in different body areas is provided by Sokolov et al. (1994).

Females have four mammae ( Estes 1992). The content of the milk of one captive female was reported by Grzimek (1959) as: 2.1% fat, 4.0% lactose, 0.78% albumin fraction, 5.63–7.7% casein, but these values should be taken with caution because the female died next day from a deficiency of vitamin C and K.

The dental formula of L. walleri is i 0/3, c 0/1, p 3/3, m 3/3, total 32. Although L. walleri is classified as a browser (hypsodonty index = 1.32, mesowear score = 0.87— Kaiser et al. 2013), its occlusal morphology is more similar to frugivores due to the thickened enamel (measured on the M2— Heywood 2010). L. walleri has relatively small frontal sinuses that terminate anterior to the supraorbital canal. The sinuses are restricted to the lateral position of the frontal bone and are separated from the lacrimal sinuses by a thin bony lamina ( Farke 2010).

The erect stance in L. walleri is possible due to a pronounced lumbar lordosis permitted by a reduction in the spinous process of the lumbar vertebrae ( Richter 1970; Cartmill and Brown 2017). Apart from this, other morphological features usually attributed to bipedalism (e.g., expanded cranial sector of the acetabular semilunar surface) are not present in L. walleri (Cartmill and Brown 2017) .

Richter (1970) compared the proportions of the limb bones in various gazelline antelopes and concluded that the elongation of the limbs in Litocranius primarily stems from an elongation of the metapodials; however, the humerus and radius are also relatively longer than in other antelope species. He also noted that Litocranius has the longest forelegs of all gazelline antelopes (90.5% the length of the hind legs) and is the only genus where the metacarpals are longer than the metatarsals. Compared with other antelopes, the limb bones of Litocranius are extremely thin-walled. Currey and Alexander (1985) calculated the ratio K between the outer and the inner diameter of the bone for various mammalian species (K = 0 when bone is solid, K approaches 1 when bone is very thin-walled). The values for L. walleri were: femur 0.73; tibia 0.54; metatarsus 0.58; humerus 0.67; and metacarpus 0.63. The values for the femur, metacarpus, and metatarsus of L. walleri are the highest among mammals in general, even including small mammals that tend to have thinner bone walls.

The vertebral formula of L. walleri is 7 C, 13 T, 6 L, 4 S, 7 Ca, total 37 (number of caudal vertebrae according to a drawing in Kingdon 1982:428). The seventh cervical vertebra has a robust, comb-like processus spinosus that is rather thin and spiky in other antelopes ( Richter 1970). The surfaces of the vertebral joints in the neck are very large and arranged almost horizontally, allowing for extensive movement in the sagittal plane.

Despite its facultative bipedalism, the musculature of the hind leg of Litocranius only shows minor differences compared with Thomson’s gazelle ( Eudorcas thomsonii ) or domestic bovids ( Richter 1970). Values for the distribution and diameter of red, intermediate, and white muscle fibers in the longissimus dorsi muscle of L. walleri and other antelopes are provided by Stickland (1979). In comparison to other bovids, L. walleri has the highest percent distribution of red muscle fibers (50.5% versus 11.1–28.2% in other bovids) and the lowest distribution of white muscle fibers (27.5% versus 34.3–44.8%) and intermediate muscle fibers (22.0% versus 28.0–54.6%). The diameters of the respective fibers (red, 40.6 µm; white, 70.3 µm; intermediate, 55.3 µm) lie within the range of those in other bovid species.

Tongue measurements for L. walleri were provided by Meier et al. (2016). Mean values (cm) for three individuals were: tongue length, 13.2; torus length, 5.5; torus width, 2.5; corpus length, 7.7; maximum width of corpus, 2.2; minimum width of corpus, 2.0; free portion, 3.9 (29%). L. walleri therefore has the smallest relative tongue length (in terms of total length, torus length, and corpus length in relation to body mass) of all analyzed species that have 0% grass in their diet ( Meier et al. 2016).

The aorta of L. walleri shows paired ostia shaped like narrow slits for all intercostal arteries. The renal arteries have a common origin in the aorta ( Sikes 1967). The ostial collars have a characteristic cup-like shape, that can also be found in other artiodactyls with a sloping back, adapting a more vertical stance when feeding, for example, klipspringer ( Oreotragus ), giraffe ( Giraffa ), and steenbok ( Raphicerus campestris — Sikes 1969).

Brain weights of two adult L. walleri were reported to be 105 and 130 g; body weights of the respective animals were 38 and 42 kg (Oboussier and Schliemann 1966). Friant (1938) described the fetal brain of L. walleri and made the following observations: posterior groove of rhinal sulcus more pronounced than the anterior; ectosylvia anterior more visible than the posterior; lateral sulcus and suprasylvia clearly visible; coronal sulcus noticeable, and presylvia slightly visible. For the splenial sulcus, only the median part was visible. Friant (1938) suggested that, similar to Gazella , the crucial sulcus would be the only part of the splenial that extends to the external part of the hemispheres in the adult brain of Litocranius .

Average weights of salivary glands of L. walleri are 59.2 g for the parotid gland (n = 4) and 27.8 g for the mandibular gland (n = 2— Hofmann et al. 2008). Maloiy et al. (1982) described the digestive system of 10 wild L. walleri (mean % ± SE): weight of the gastrointestinal contents relative to the body mass, 12.1 ± 0.8%; weight of each gut region (as % of total gut weight) was reticulorumen, 78.1 ± 0.9%; small intestine, 5.6 ± 0.8%; and cecum plus colon, 16.3 ± 0.7%. The analyzed L. walleri had a mean blood glucose concentration of 67.3 ± 1.8 mg /100 ml and a serum urea nitrogen concentration of 48.0 ± 1.6 mg /100 ml. The rumen of L. walleri is relatively evenly papillated (Hofmann and Stewart 1972), with surface enlargement factors of 7.5 in the dorsal rumen, 12.0 in the ventral rumen, and 19.0 in the atrium. These values are typical for browsing ruminants ( Clauss et al. 2009).

Litocranius walleri has a bicornuate uterus, and the placenta occupies both uterine horns. It has a polycotyledonary placenta (typical for most ungulates), with an epitheliochorial relation to the maternal tissues. Benirschke (2008) examined two placentas: one belonged to an immature fetus and had> 80 cotyledons and an umbilical cord of 5 cm length, and the other belonged to a full-term fetus and had 48 cotyledons arranged in four rows and an umbilical cord of 15 cm length.

Function. —Clemens and Maloiy (1983, 1984) studied the concentration, absorption, and secretion of fluids, electrolytes, and organic acids in the digestive system of African wild ruminants. They found that the percentage of dry matter in the distal colon (where fecal pellets are formed) of L. walleri was 65%, indicating a water resorption within the colon of 65.1%. This was the highest value of all analyzed species in these studies (average percentage of dry matter in the colon in other bovid species ranged from 15.0 to 42.2). The resorption and secretion of electrolytes was found to be quite variable among the species, and the sample size for most species was low (e.g., n = 3 for L. walleri ), so results have to be viewed with caution. L. walleri absorbed 38 mg /100 ml of sodium, 4.5 mg /100 ml of potassium, 11.1 mg /100 ml of chloride, and 19.1 mmol/l of volatile fatty acids (Clemens and Maloiy 1983).

In an adrenocorticotropic hormone (ATCH) challenge test, the concentration of fecal glucocorticoid metabolites (stresshormone derivatives) of L. walleri was measured using the ICN 07-120102 corticosterone antibody. An excretion peak of about 230 ng /g occurred about 12 h after the ATCH injection, and a second prolonged peak of 115– 95 ng /g about 38–60 h after the injection ( Wasser et al. 2000).

ONTOGENY AND REPRODUCTION

Litocranius walleri shows nonseasonal reproduction ( Piening Schuler et al. 2009a), with births occurring year-round. Only singletons are born, but one female can have two births in 1 year ( Schomber 1963), with an average calving interval of 248 days (222–321) in captivity ( Houston et al. 1988). Body weights of neonatal L. walleri are 2.8–3.4 kg ( Piening Schuler et al. 2009b). Weights recorded 24 h after birth were 2.3–3.8 kg (mean 3.1 kg — Houston et al 1988). Two hand-reared L. walleri accepted milk until 5 and 9 months old ( Schomber 1963). Weaning of two other hand-reared young at the St. Louis Zoo occurred at 49 and 89 days ( Houston et al. 1988). Minimum age of conception (recorded in a zoo) was 11 months, but females usually reach sexual maturity at 1.5–2 years ( Kirchshofer 1963; Houston et al. 1988). Male offspring start to produce semen at 12 months of age and are pubertal at 18 months ( Penfold et al. 2005); they part with their mothers at 1–1.5 years and become territorial at 2.75–3.5 years ( Leuthold 1978b).

In a study on artificial insemination, Penfold et al. (2005) studied the reproductive physiology of L. walleri . The testosterone level of captive adult males did not show seasonal variation. Mean testes volume was 22.2 ± 1.7 cm 3. Semen was collected from anesthetized males using electroejaculation, and no seasonal variation was found in total sperm number (463.0 ± 88.0 × 106), sperm motility (60 ± 4.4%), and normal sperm morphology (40.0 ± 4.7%). Interestingly, this high number of malformed spermatozoa (60%) was also found in wild L. walleri ( Penfold et al. 2005) . Mean length of the ovarian cycle is 18.7 ± 0.9 days ( Penfold et al. 2005). Pregnancy lasts for about 210 days (range: 194–221 days— Houston et al. 1988). One female treated with artificial insemination using cryopreserved spermatozoa delivered a stillborn young after 213 days ( Penfold et al. 2005). There is one reported case of an ectopic pregnancy carried to full-term in a captive L. walleri with fetal viability confirmed through day 213. Fetal membranes were attached to the omentum (Helmick and Citino 2001).

Figures

Fig. 1.—Adult male Litocranius walleri in Buffalo Springs National Reserve, Kenya.Photograph by Dr. Frank Zachos used with permission.

Fig. 2.—Skulls of adult Litocranius walleri housed in the mammal collection at Zoological Research Museum Alexander Koenig in Bonn. (A–D) Female (ZFMK_MAM_1936.0055 from northern Tanzania, west of Kilimanjaro, condyle-basal length [CBL] = 209 mm); (E, F) male (ZFMK_ MAM_1957.0211 from Kenya, Garba Tula, north of the Tana river, CBL = 208 mm). (A, E) Lateral view; (B, F) dorsal view; (C) ventral view; (D) mandible.

Fig. 3.—Geographic range of Litocranius walleri (based on IUCN 2016). Map is based on a customizable map of the Horn of Africa available on d-maps.com (https://d-maps.com/carte.php?num_car=25596&lang=de). The Sabaki River was added using the map “Tsavo national park map en.png” by Lencer, available on commons.wikimedia.org. Coordinates were added using the map of Kenya available on www.ducksters.com.