Choeropsis liberiensis ( Morton, 1849 )

Flacke, Gabriella L & Decher, Jan, 2019, Choeropsis liberiensis (Artiodactyla: Hippopotamidae), Mammalian Species 51 (982), pp. 100-118 : 100-108

Choeropsis liberiensis ( Morton, 1849)

Pygmy Hippopotamus or Pygmy Hippo

Hippopotamus minor Morton, 1844:15 . Type locality “…from the river St. Paul’s.” Liberia; preoccupied by Hippopotamus minor Desmarest, 1822 (fossil hippo species from Cyprus). Hippopotamus (Tetraprotodon) liberiensis Morton, 1849:4 . Replacement name for Hippopotamus minor Morton, 1844 . Diprotodon liberiensis: Duvernoy, 1849:277 . Name combination.

Choerodes liberiensis: Leidy, 1852:52 . Name combination.

Choeropsis liberiensis: Leidy, 1853:213 . First use of current name combination.

Ditomeodon Liberiensis: Gratiolet, 1867:202 . Name combination.

Hexaprotodon liberiensis: Coryndon, 1977:69 . Name combination.

CONTEXT AND CONTENT. Context as for genus. Two subspecies of Choeropsis liberiensis are currently recognized ( Lewison 2011):

C. l. heslopi Corbet, 1969:388 . Type locality “Omoku (=

Omoko), Owerri Province, Nigeria (5º 19´N, 6º 40´E).” C. l. liberiensis ( Morton, 1849:4) . See above.

NOMENCLATURAL NOTES. Historically, both Hippopotamus amphibius and Choeropsis liberiensis have been classified as Suidae , and more recently, as Hippopotamidae . Recent molecular phylogenetic studies and comparative morphological analyses have supported Hippopotamidae and Cetacea as forming a monophyletic group ( Shimamura et al. 1997; Nomura et al. 1998; Ursing and Arnason 1998; Nomura and Yasue 1999; Geisler and Uhen 2003; Boisserie 2005; Boisserie et al. 2005; Fisher et al. 2007); Cetartiodactyla was proposed as the taxon name for the group ( Montgelard et al. 1997) and subsequently more clearly defined (O’Leary and Gatesy 2008).

The generic name for the pygmy hippo has been debated based on phylogeny, ontogeny, and morphological characteristics. The species was first described by Morton (1844), who labelled the animal Hippopotamus minor based on differences in cranial morphology and dental formula between common and pygmy hippos. Leidy (1852), focusing on striking differences in cranial anatomy between the two hippo species, subsequently proposed Choerodes liberiensis . However, after learning that Choerodes had already been appropriated to an insect, he subsequently proposed Choeropsis , from the Greek for “pig-like” and liberiensis to describe geographic origin ( Leidy 1853). Early anatomists and zoologists disputed whether the pygmy hippo should be assigned to a separate genus from Hippopotamus ( Flower 1887; Chapman 1894; Renshaw 1904), but majority opinion eventually supported the distinction.

Choeropsis was used until Coryndon (1977) placed the pygmy hippo in the genus Hexaprotodon , meaning “six front teeth.” Both names appear interchangeably in the scientific literature, but only Choeropsis is correct. Most recently, Boisserie (2005) concluded that the mix of primitive and derived features of the extant pygmy hippo gives it a distinct lineage, validating the genus Choeropsis for the pygmy hippo and restricting the genus Hexaprotodon to the fossil lineage found mostly in Asia (Falconer and Cautley 1836). This view was also recently embraced by Groves and Grubb (2011).

In Liberia C. liberiensis is referred to as mwe (Gola), nimwe (Bassa), or nigbwe (Grebo— Robinson 1979). In Sierra Leone the animal is known as maali (Mende—A. Conway, pers.

comm.), màylă ădămāy (Koranko— Robinson 1970), or ny omwè (Susu— Robinson 1970). In southwestern Côte d’Ivoire the local name is nu-ugbe (Oubi—K. Ouattara, pers. comm.). In Nigeria C. liberiensis is referred to as odufiowei, abein, or ebei in Ijo-group languages or as agumagu or ogomagu in Igbo-group languages ( Robinson et al. 2017).

DIAGNOSIS

Choeropsis liberiensis ( Fig. 1 ) is distinguished from the common hippopotamus Hippopotamus amphibius by its smaller size, dental formula, distribution, and habitat. The only recorded weights for wild specimens are from an adult male at 204 kg ( Heslop 1944) and two adult females at 165 and 170 kg ( Hentschel 1990). Average weight in captivity ranges from 160 to 275 kg with a shoulder height of 70–80 cm ( Lang 1975); by comparison, a wild common hippopotamus can weigh upwards of 1,500 kg with a shoulder height of 130–140 cm (Marshall and Sayer 1976). The dental formula of C. liberiensis is i 2/1, c 1/1, p 4/4, m 3/3, total 38 ( Morton 1844). Rarely there are three total mandibular incisors instead of the single pair ( Johnston 1906). In contrast, the common hippopotamus has two pairs of both maxillary and mandibular incisors, i 2/2, c 1/1, p 4/4, m 3/3, total 40 ( Laws 1968). C. liberiensis is limited in geographic distribution to the Upper Guinean forest ecosystem of West Africa and is now found only in Côte d’Ivoire, Guinea, Liberia, and Sierra Leone ( Ransom et al. 2015); the common hippopotamus is found throughout sub-Saharan Africa ( Boisserie et al. 2005). C. liberiensis inhabits primary and secondary rainforest and swamps ( Bülow 1987; Hentschel 1990), whereas the common hippopotamus frequents riverine savannahs.

GENERAL CHARACTERS

Choeropsis liberiensis is gray to black in color with occasional depigmented pink patches on the body and limbs ( Fig. 1 ). The skin is thick with a rubbery texture ( Flach et al. 1998). Subdermal glands secrete an oily clear-to-white, foamy fluid that protects the skin against sunburn, dehydration, and bacterial infection ( Eulenberger 1995; Hashimoto et al. 2007). Coarse, bristle-like hairs are found only at the ear margins, on the lips, and at the end of the tail ( Lochte 1951). The ears and nostrils function as valves that close tightly when the animal is under water ( Pocock 1923).

Cranial measurements (cm) from the initial adult skull specimen used to identify C. liberiensis as a separate species, as opposed to a smaller version of the common hippopotamus were: length 31.2 (measured from the anterior extremity to the notch between the condyles of the occipital bone); width (at the widest point between the zygomatic arches) 20.3; distance between the orbits 9.9 ( Morton 1844). Proportionally, the brain case is larger and the muzzle is shorter in C. liberiensis than in the common hippopotamus ( Fig. 2 ; Johnston 1906). Skeletal measurements (cm) for C. liberiensis were: body length (173); carpal joint to scapula (61); tarsal joint to hip joint (48); tallest point along the spine (79— Leidy 1853). Cranial measurements (cm) from one of the few known examples of subspecies C. l. heslopi were: length 35; width 21; maxillary canine length 15.9, 6.2 projecting from jaw; mandibular canine length 28.6, 10.5 projecting from jaw ( Heslop 1944). These are the only measurements reported from wild C. liberiensis .

There is a dearth of detailed information concerning Choeropsis in the wild due to the limitations for in situ research; the majority of data is derived from zoo specimens, all of which are of the nominate species C. liberiensis . Morphometric measures, including body mass, are not available for any animals taken from the wild, except for the two adult females weighed by Hentschel (1990) during his fieldwork in Côte d’Ivoire. Initial measurements from Zoo Basel, Switzerland, for captive adult females (n = 3) were 77–83 cm tall, measured at center of the back, and 142–150 cm long from nose to tail tip; one adult male was 81 cm tall and 157 cm long ( Lang 1975). Body mass for adult females (n = 5) ranged from 179 to 267 kg; for adult males (n = 2) 208 kg and 273 kg ( Lang 1975). The International Studbook reports a body mass range of 180–260 kg and a shoulder height of 80 cm ( Steck 2017); however, several textbooks and gray literature sources from zoological institutions in Europe and the United States ( Crandall 1964; Boever 1978; Laws 1984; Taylor and Greenwood 1986; Jarofke 1993; Eulenberger 1995; Nowak 1999; Thompson 2002; Miller 2003, 2007; Miller et al. 2014; Walzer and Stalder 2014) give body mass as 160–350 kg and body length of adults as 150–170 cm. The animals at the higher end of the weight range are undoubtedly overweight ( Flacke et al. 2015).

DISTRIBUTION

The extant Choeropsis liberiensis is endemic only to the Upper Guinean forest ecosystem in West Africa ( Fig. 3 ), and its distribution is limited to fragmented populations in Côte d’Ivoire, Guinea, Liberia, and Sierra Leone ( Ransom et al. 2015). The Lower Guinean species, C. heslopi , was formerly reported in southeastern Nigeria, from the Owerri and Warri provinces of the Niger River Delta eastward from the Cross River ( Ritchie 1930; Heslop 1945; Corbet 1969). If the Nigerian species persists, the only remaining habitat where it could possibly occur is the Upper Orashi Forest Reserve ( Robinson 2013).

In Guinea, C. liberiensis occurs primarily in the forested regions of the southeast ( Sidney 1965). Recent confirmed reports substantiate its presence in the Ziama and Diécké forests and the Mont Béro Reserve ( Mallon et al. 2011). In Ziama, evidence of C. liberiensis (tracks and dung) was found along the Njéré, Gniogné, and upper Ouin watercourses, and in Diécké along the Nié, Grand Gbin, Gbin-bé, and Lih streams ( Bützler 1994). Historically its most westward report is from Forécariah, Guinea, about 60 km east of Conakry ( Dekeyser 1955). The most westward extent of its range is likely the Ouatamba-Kilimi National Park in northwestern Sierra Leone ( White 1986), also the northernmost location where the species has been recorded, although there are no recent reports to substantiate its continued presence in this area.

In the first one-half of the 20th century, C. liberiensis was widely distributed in the eastern part of Sierra Leone ( Sidney 1965). Confirmed reports of its presence are from both Tiwai Island and other small islands along the Moa River ( White et al. 1986; Conway 2013), and from the Gola Forest Reserve Complex along the Mano River on the border with Liberia (Hillers and Muana 2010; Lindsell et al. 2011; Mallon et al. 2011). In both locations its presence was confirmed via camera trap photos, personal observations, and by indirect methods such as tracks and dung. One animal was also caught in a pit-trap near Tiwai Island in 2010 ( Conway 2013). It is also known from the Loma Mountains in northeastern Sierra Leone ( Mallon et al. 2011).

Liberia represents the core of its distribution, as reflected in the specific name, C. liberiensis . The earliest confirmation of this species from Liberia is from the St. Paul River and the Duquea River, a tributary of the Junk River ( Büttikofer 1890). In the early 20th century the German hunter and explorer Hans Schomburgk found C. liberiensis in many other parts of interior Liberia, including along the Duquea River as well as the Mano and Lofa watersheds in western Liberia near the Sierra Leone border ( Schomburgk 1912, 1913a, 1922). Its presence throughout both coastal and interior regions of Liberia was confirmed as recently as the 1960s ( Sidney 1965). However, by the late 20th century its distribution was limited to Sapo National Park and Grebo National Forest in the southeast, and North Lorma National Forest in the far northwest; it was deemed unlikely to persist in the central regions of Liberia ( Anstey 1991a). However, shortly thereafter a survey of the Cestos and Senkwehn rivers reported numerous signs of its presence in both watersheds (Robinson and Suter 1999).

A 2005 survey conducted in Gola, North Lorma, and Grebo National Forests in Liberia confirmed signs of C. liberiensis only in Grebo National Forest, in the far southeast near the border with Côte d’Ivoire; the other areas lacked evidence of its presence ( Barrie et al. 2007). Other recent confirmed records are from Gola National Forest in the west, from the Wonegizi National Forest in the north, on the border with Guinea, and in Sapo National Park in the southeast ( Collen et al. 2011). Hunter and bushmeat market interviews conducted in Nimba and Grand Bassa counties and near Sapo National Park revealed a declining trend in C. liberiensis numbers over the last decade ( Greengrass 2011). The species has also been confirmed in the southeast in Maryland and River Gee counties and within the proposed Grand Kru-River Gee Protected Area ( Mallon et al. 2011).

Southwestern Côte d’Ivoire represents the easternmost range of C. liberiensis . Historical reports of its existence in Ghana were rejected as erroneous accounts of juvenile common hippos ( Robinson 2013). Within Côte d’Ivoire C. liberiensis has not been recorded farther east than the Bandama River ( Sidney 1965) and Azagny National Park ( Bülow 1987), about 80 km west of Abidjan. The northernmost extent of C. liberiensis in Côte d’Ivoire is the area around Mt. Nimba in the upper Cavally River valley ( Roth et al. 2004). Field surveys conducted between 1982 and 1986 found signs of C. liberiensis within 21 protected areas in southwestern Côte d’Ivoire, with the highest abundances in Taï National Park and the surrounding N’Zo Faunal Reserve. Based on a number of field surveys and recent camera trap photos, Taï National Park probably holds many C. liberiensis (Roth and Merz 1986; Hentschel 1990; Roth et al. 2004; Eshuis 2011; van Heukelum 2011; Hoppe-Dominik et al. 2011; Bogui et al. 2016). Surveys have also confirmed its presence in Cavally and Goin Débé Classified Forests ( Mallon et al. 2011).

FOSSIL RECORD

The Cetacea and the Hippopotamidae are believed to share a common ancestry, with initial divergence estimated around 54 million years ago (Ursing and Arnason 1998). Hippopotamids are believed to have originated in Africa sometime in the mid- Miocene period, approximately 11–16 million years ago, in the areas of modern-day Kenya and Tunisia ( Pickford 1983; Boisserie 2005). The earliest remains that unequivocally resemble extant hippos appeared in the fossil record in the late Miocene, around 8 million years ago ( Weston 2000), and later within this period its range also extended into parts of Eurasia (Boisserie 2005).

The earliest Hexaprotodon fossils, dating from the Mio– Pliocene period about 5 million years ago, were recovered from the Siwalik Hills on the India – Pakistan border (Falconer and Cautley 1836). Fossil records support the much more recent presence of the now-extinct dwarf hippo species on the islands of Crete (Hippopotamus creutzburgi), Cyprus ( Phanourios minor ), and Madagascar (Hippopotamus madagascariensis). Carbon dating indicates that the Cypriot population went extinct about 10,500 years ago ( Simmons 1999), while the Malagasy population probably survived until about 2,000 years ago (MacPhee and Burney 1991).

FORM AND FUNCTION

Form. — The hippopotamids are unique among the Artiodactyla in that they have retained a semiaquatic lifestyle and all four toes are weight-bearing. Choeropsis liberiensis has retained several features more adapted to a terrestrial lifestyle compared to the common hippopotamus, including proportionally longer limbs and only moderately webbed toes that spread more widely under its own weight ( Johnston 1906; Pocock 1923; Fisher et al. 2007). Retention of intrinsic adductor muscles in the feet functions to prevent splaying of the toes, an adaptation to walking on muddy substrates ( Fisher et al. 2007). Skeletal measurements (cm) for 19 adult C. liberiensis of both sexes emphasize its diminutive size in comparison to the common hippopotamus (values [cm] given in parentheses), including maximum femoral length, 25.4–29.2 (43.4–58.0); maximal width from femoral head to greater trochanter, 7.8–9.6 (14.6–17.7); diameter femoral shaft, 2.9–3.5 (5.5–7.2— Weston 2000). The vertebral formula of C. liberiensis is 7C, 14T, 5L, 4S, and 10 or more CA, total 40 or more ( Macalister 1873). Morphobank has a scalable skeletal view of C. [ Hexaprotodon ] liberiensis (O’Leary and Kaufmann 2012) .

The myology of C. liberiensis is generally similar to that of the common hippopotamus, as detailed from an approximately 8-week-old female specimen ( Macalister 1873). However, later anatomical studies have revealed several variations in topographical muscle anatomy for C. liberiensis . For example, the rhomboid muscle group can either originate in the neck ( Campbell 1936; Fisher et al. 2007) or from the occipital bone ( Macalister 1873). Similarly, the trapezius muscle can variably originate from the last two cervical vertebrae ( Campbell 1936) or the occipital bone ( Macalister 1873; Fisher et al. 2007) in C. liberiensis , whereas the common hippopotamus lacks an occipital origin. The insertion site of the brachialis muscle can be on the radius only ( Macalister 1873; Fisher et al. 2007) or on both the radius and ulna ( Campbell 1936); the radius and ulna are functionally or actually fused in artiodactyls. Overall, the myology of the C. liberiensis resembles ruminants more closely than suids, and both extant hippo species are unique in retaining several primitive features in the forelimb, consistent with the hypothesis that hippopotamids diverged from other Artiodactyla early in the history of this group ( Fisher et al. 2007).

The highly vascularized and inelastic skin of C. liberiensis is about 1 cm thick with extensive subcutaneous fat and a rubbery texture ( Flach et al. 1998). Multiple anecdotal reports describe white foamy skin secretions produced by the subdermal glands in wild C. liberiensis , and captive animals often exhibit the same secretions during periods of physical exertion, including during mating ( Steinmetz 1937; Lang 1975; Schubert 2004; von Houwald et al. 2007). Skin secretions from a female C. liberiensis had a pH of 9.5 and lower mean sodium (Na +) and potassium (K +) concentrations than secretions from the common hippopotamus, possibly reflecting dietary or metabolic differences ( Olivier 1975). Insensible water loss from the skin of a nonsedated female C. liberiensis at a temperature range of 29–35°C was between 6.0 and 22.5 mg / 5 cm 2 /10 min, depending on the areas of the body. This range is comparable to previously published rates for the common hippopotamus at the same temperatures, indicative of similar dermal physiology for the two species ( Olivier 1975).

Hair measurements (mm) for C. liberiensis were: up to 35 for lip hairs, 6–9 for ear hairs, and 10–18 for tail hairs ( Lochte 1951). The hair structure, similar to that of stiff bristles, is comprised of multiple keratin strands that are split at the ends, especially those of lip and tail ( Lochte 1951). Similar conglomerate, bristle-like hairs are found only in seals and sea lions ( Carnivora ; Otariidae ). Natural splitting of tail hairs may function to increase the surface area and thus improve the efficiency with which dung is spread in the environment during its unusual defecation behavior ( Kranz 1982).

The eyes of C. liberiensis are less prominent and have a more lateral placement with a wider field of vision than with the dorsal eye placement of the amphibious common hippopotamus. Asymmetry of the ciliary body also gives C. liberiensis a more extensive rear field of vision ( Hegner 1967). The lateral placement of eyes and a tapering head shape also facilitate a binocular field of vision when the mouth is gaping open, allowing the animal to advance toward a threat while maintaining visual contact ( Van den Bergh 1971). The globe (eyeball) itself is rounded and slightly taller than it is wide; measurements from a subadult female (117 kg) were 24.3 mm horizontal diameter, 18.6 mm anterior to posterior diameter, and 22.5 mm vertical diameter ( Hegner 1967). The pupil is oval, 7 mm wide and 5 mm high, with a symmetrical iris width of 2 mm; the lens is 10 mm in diameter and 7 mm thick; scleral thickness varies between 1.2 and 1.5 mm ( Hegner 1967). The tapetum lucidum is light blue, reflective, and extends across the majority of the fundus; the ventral aspect is grossly obscured by the pigmented epithelial cell layer of the retina and thus the functional tapetum primarily occupies the dorsal aspect of the fundus ( Hegner 1967). The retina itself has two histologically distinct types of pigmented epithelial cells and three types of photoreceptor cells ( Hegner 1967).

The cranial and dental measurements (cm, ranges) of C. liberiensis (n = 20–30) are consistent with the morphology of browsers ( Weston 2000, 2003): height of the ramus of the mandible, 5.75–7.6; length of mandibular symphysis, 5.6–7.7; depth of the mandibular symphysis, 3.9–6.9; distance between lower canines, 11.1–16.5; toothrow length (second through fourth premolars), 5.0–6.2; toothrow length (first through third molars), 7.4–9.1 ( Weston 2000). C. liberiensis has low-crowned molars, further supporting its status as a forest-dwelling consumer of coarse vegetation ( Weston 2000). However, C. liberiensis may exhibit an intermediate feeding strategy with characteristics of both browsers and grazers, as hypothesized by Hentschel (1990).

The four stomach compartments are similar to those of a ruminant and include the nonglandular visceral and parietal blind sacs (also termed left and right diverticula), the vestibulum, the connecting chamber, and the customary glandular stomach ( Langer 1975; Macdonald and Hartman 1983). The nonglandular compartments are collectively termed the “fore stomachs” ( Langer 1975) or the “proventricular chambers” (Macdonald and Hartman 1983). The esophagus opens into the dorsal part of the visceral blind sac. The blind sacs, ventriculus, and connecting chamber are further subdivided by a series of semilunar folds, 1–3 cm in height, derived from the tunica muscularis ( Langer 1975; Macdonald and Hartman 1983). The connecting chamber is oriented in a dorsal-toventral direction in the newborn to assist with passage of milk into the glandular stomach; the orientation changes in the adult to promote retention of ingesta and its movement between compartments of the fore stomach (Macdonald and Hartman 1983). The mucosal epithelium of the fore stomachs is lined with finger-like papillae, each varying in length, thickness, and shape; these papillae are less prominent in the newborn (Macdonald and Hartman 1983; Endo et al. 2001). In the adult C. liberiensis , the papillae increase the functional surface area of the fore stomachs by 3.32 times compared to that of a smooth wall ( Langer 1975).

A distinct border along the lesser curvature marks the separation between the nonglandular connecting chamber and glandular epithelium of the stomach ( Langer 1975). As in ruminants, a fornix-like blind sac in the glandular stomach, lined with nonglandular epithelium, exposes the stomach contents to digestive enzymes ( Langer 1975). The volume of the combined fore stomachs in the newborn C. liberiensis is about 68% of the total gastric volume; in adults this proportion increases to 94% ( Langer 1975). Hippos are called “pseudo-ruminants” or ruminant-like, with microbial fermentation in the fore stomachs supporting digestion through the production of volatile fatty acids and essential vitamins ( Langer 1975; Macdonald and Hartman 1983). An elongated triangular gall bladder is present, and the bile duct empties into the duodenum 15 cm distal to the pylorus ( Macalister 1873; Langer 1975). The pylorus connects to a simple linear small intestine. The entire small intestine (duodenum, jejunum, ileum) of an adult female is about 16 m long with an average diameter of 2.5 cm; the colon is about 2.5 m long (Macdonald and Hartman 1983). There is no cecum ( Macalister 1873), but the junction between the large and small intestine is characterized by an increase in diameter and a change in the lining of the mucosal epithelium (Macdonald and Hartman 1983). Gastrointestinal anatomy and physiology of hippos is most similar to that of macropodid marsupials, endemic to Australia ( Endo et al. 2001; Clauss et al. 2004; Schwarm et al. 2006, 2008).

The cardiovascular anatomy of C. liberiensis is not different from that of other large mammals ( Macalister 1873; Macdonald 1988). Lobulated kidneys are similar in architecture to those of the Bovidae but lack infundibula and renculi ( Macalister 1873). Instead, the lobes are projections like the fingers of a hand; each kidney has 17–25 lobes ( Maluf 1978, 1994). The kidneys of an adult C. liberiensis are about 75% cortex and 23% medulla; the remaining 2% are the branched tubus maximus and the relatively small renal pelvis ( Maluf 1994). Branching of the tubus maximus and the absence of renculi within a lobed kidney are features found only in hippopotamids, making their renal anatomy unique among the Artiodactyla ( Maluf 1994) . The two kidneys equal 0.22% of body mass in the adult and 0.51–0.63% of body mass in the newborn ( Maluf 1994). The kidney of an adult is about 133 mm long by 92 mm wide with a thickness of 33 mm ( Maluf 1994). Each kidney is supplied with cranial and caudal renal arteries that branch into several perforating arteries, which in turn supply their respective lobes ( Maluf 1978; 1994). Each kidney has about three times 106 glomeruli ( Maluf 1994); however, the total length of the tubule, including the Loop of Henle, is only 6.8 mm ( Maluf 1978). The large number of short nephrons is believed to have allowed the evolution of a bigger animal, albeit with a limited ability to produce concentrated urine ( Maluf 1978, 1994).

The reproductive anatomy of C. liberiensis has not been extensively detailed for either sex, although in many aspects it is likely to be similar to that of the common hippopotamus ( Macdonald 2007). The penis and prepuce curl caudally when not erect ( Pocock 1923). Ejaculates from seven C. liberiensis had an average volume of 3.0 ± 0.9 (SEM) ml and a concentration of 2.1 ± 1.3 × 106 spermatozoa per ml ( Saragusty et al. 2010). Fluorescence in situ hybridization, used to determine the primary sex ratio in ejaculates from males (n = 10), revealed an average of 43% of spermatozoa to have a Y chromosome ( Saragusty et al. 2012).

Choeropsis liberiensis has diffuse epitheliochorial placentation and an umbilical cord with two arteries and two veins (Macdonald and Bosma 1985); however, a few specimens of C. liberiensis have only three umbilical vessels, two arteries and one vein ( Benirschke 2007). The fetal placentas of three fullterm calves weighed 900–1,125 g ( Benirschke 2007). The chorion is folded and its surface is covered with villi interspersed with rounded or irregularly shaped areolae; the villi are lined with columnar epithelium (Macdonald and Bosma 1985). There are two types of villi: areolar, which are more blunted, less branched, broader and overlie the endometrial glands, and interareolar villi. The latter have prominent interdigitation between the fetal capillaries and maternal trophoblast, a feature thought to enhance gas exchange across the placenta (Macdonald and Bosma 1985; Macdonald 2007). Accumulations of yellowbrown pigment occur in some trophoblasts, including in some of the villous tips, a histologic feature distinctly different from other placentas in the Artiodactyla ( Benirschke 2007) . The multifocal areas of squamous metaplasia on both umbilical cord and amniotic membranes are unique to the hippopotamids and appear strikingly “pustular” (Macdonald and Bosma 1985; Benirschke 2007).

Function. —Metabolite levels of progesterone, as measured in the skin secretions of two females for six consecutive months, were 104–2,662 pmol/l and 98–2,863 pmol/l (Dathe and Kuckelkorn 1989). Progesterone levels below 500 pmol/l were noted every 27–32 days, consistent with physiological estrus (Dathe and Kuckelkorn 1989). Metabolite concentrations of progesterone in the saliva during the same time period were one-third to one-half of those in skin secretions, but the trends were comparable. Metabolite concentrations of estrogen in fecal samples similarly indicated a mean (± SD) estrous cycle of 31.8 ± 7.4 days ( Flacke et al. 2017a). Metabolite levels of preganediol in fecal samples steadily increased during pregnancy, reaching concentrations markedly above luteal phase levels in the second one-half of gestation ( Flacke et al. 2017a).

The first physiological variables for C. liberiensis were reported from an anesthetized female C. liberiensis : heart rate 96–106 beats/min, rectal temperature (°C) 35.8–38.9 ( Franz et al. 1978). Flach et al. (1998) reported a heart rate (beats/ min) of 42–100 and respiratory rate (breaths/min) of 0–10 for an anesthetized female. Data from 14 immobilized adults were as follows: mean heart rate (beats/min ± SD) 34 ± 7; mean respiratory rate (breaths/min ± SD), 14 ± 6; mean peripheral hemoglobin saturation, 91–92% for animals receiving supplemental oxygen; mean end tidal carbon dioxide (mmHg ± SD), 55 ± 10 ( Bouts et al. 2012). The following values were obtained from an anesthetized male C. liberiensis : average heart rate (beats/ min) 60; average respiratory rate (breaths/min) 4; body temperature range (°C) 33.6–35.3; mean arterial hemoglobin saturation 95–99% ( Weston et al. 1996). Allometric scaling was used to estimate the heart rate (beats/min) of conscious animals at 60–65 ( Bouts et al. 2012); heart rate data for conscious animals has otherwise not been reported. Respiratory rate (breaths/min) for conscious C. liberiensis is reported at 10–16 ( Pearce et al. 1985).

Median blood gas parameters for three adult C. liberiensis were: pH, 7.43; partial pressure of carbon dioxide (mmHg), 56; partial pressure of oxygen (mmHg), 40; bicarbonate (mmol/l), 35.9 ( Bouts et al. 2012). Mean values (n = 5 anesthetic episodes) from a juvenile female were: pH 7.4; partial pressure of carbon dioxide (mmHg), 61.5; partial pressure of oxygen (mmHg), 71; bicarbonate (mmol/l), 40.8; oxygen saturation, 71.6%; end tidal carbon dioxide (mmHg), 42.5; and body temperature (°C), 36.3 ( Morris et al. 2001). Hematology reference ranges for C. liberiensis were: total white blood cells 6.1–26.6 × 103 /µl; total red blood cells 3.1–7.3 × 106 /µl; hemoglobin 10.8–17.9 g / dl; hematocrit 18–55%; mean corpuscular volume 57.9–80.3 fl; mean corpuscular hemoglobin 20.1–26.2 pg; mean corpuscular hemoglobin concentration 32.0– 35.7 g /dl; segmented neutrophils 3.5–24.5 × 103 cells/µl; lymphocytes 0.38–5.2 × 103 cells/µl; monocytes 102–1,290 cells/µl; eosinophils 74–1,560 cells/µl (Walzer and Stalder 2014). Reference ranges for serum biochemical parameters were: glucose 32–193 mg /dl; blood urea nitrogen 10–14 mg /dl; creatinine 0.5–2.7 mg /dl; uric acid 0–1.2 mg /dl; calcium 8.7–14.3 mg /dl; phosphorus 4.4–11.4 mg / dl; sodium 139–156 mEq/l; potassium 3.5–8.4 mEq/l; chloride 94–115 mEq/l; total protein 5.5–10.2 g /dl; albumin 3.5–5.9 g / dl; globulin 1.7–5.6 g /dl; alkaline phosphatase 19–309 IU/l; lactate dehydrogenase 2–867 IU/l; aspartate aminotransferase 16–493 IU/l; alanine aminotransferase 8–51 IU/l; gammaglutamyltransferase 6–189 IU/l; creatinine kinase 21–274 IU/l; total bilirubin 0.3–2.5 mg /dl; cholesterol 17–155 mg /dl; carbon dioxide 12.4–31 mEq/l (Walzer and Stalder 2014). These hematology and biochemistry values were derived from a limited number of animals (n => 10 but <50) of unknown health status; thus, they do not meet the American Society for Veterinary Clinical Pathology standards for established reference ranges (Walzer and Stalder 2014).

ONTOGENY AND REPRODUCTION

Ontogeny. —Details of ontogeny and reproduction in Choeropsis liberiensis are known only for captive animals. The most comprehensive resource detailing reproduction in captivity is the monograph by Lang (1975). Neonates only suckle 2–3 times per day, but then for comparatively long periods (Stroman and Slaughter 1972; Lang 1975). During the first 4–5 months of life, calves gain about 300 g per day ( Steinmetz 1937; Lang 1975). Weaning in captivity is as early as 3 months (Stroman and Slaughter 1972; Rahn 1978) or as late as 15 months of age ( Steinmetz 1937). Lang (1975) believed 10–12 months was ideal. Weaning age in captivity is largely dictated by logistics such as available space and behavioral interactions between mother and calf. Weaning age in the wild is not known, but likely occurs when the calf is a yearling based on footprint analysis of mother–offspring pairs in Taï National Park, Côte d’Ivoire ( Hentschel 1990; van Heukelum 2011). Calving information from wild populations are conflicting, with natives in Liberia ( Büttikofer 1890) and field researchers in Côte d’Ivoire ( Hentschel 1990; van Heukelum 2011) indicating that calving is concentrated at the end of the rainy season (August–September), but Schomburgk (1913b) reported that births occurred primarily in the dry season (November–December) in Liberia. However, all of this information is based on secondhand reports and not on observation of calves in the wild; the only observation of a wild calf occurred in November in Taï National Park, Côte d’Ivoire, during the dry season ( Galat-Luong 1981). Seasonality in births would presumably be adaptive in rainforest habitats with distinct wet and dry periods affecting availability of resources for offspring.

Reproduction. —Under managed care, male and female C. liberiensis are often kept separate except during estrus (Stroman and Slaughter 1972; Lang 1975), but breeding has also been successful when the animals are housed in pairs ( Rahn 1978). The male will mate with the female for 1 or 2 days, usually with multiple copulations per day ( Lang 1975). Sexual maturity occurs between 2.5 and 3 years of age for females ( Flacke et al. 2017a). Studbook data indicate the youngest males to successfully reproduce were between 2.8 and 3 years old, and males can continue to sire offspring until their late 30s ( Steck 2017). The youngest female to produce a viable offspring was 23 months old at the time of conception, but the average age at first conception is 5 years ( Steck 2017). Age at first conception in captivity is primarily determined by access to breeding males and availability of appropriate space for placing offspring, and thus is more heavily influenced by zoo management practices than by the natural onset of sexual maturity.

The length of the estrous cycle in C. liberiensis is 28–40 days, with a mean of 35 days based on behavioral observations ( Flacke et al. 2015) and 27–32 days based on hormonal studies (Dathe and Kuckelkorn 1989; Flacke et al. 2017a). Calves have been born during every month of the year in zoological institutions, indicating that in captivity the species is nonseasonally polyestrous ( Flacke et al. 2017a). Gestation is 188–210 days with a mean of about 200 days, after which a single calf is born ( Flacke et al. 2015). Twin births are rare and often result in mortality for one or both neonates ( Flacke et al. 2016). Pregnancy detection can be challenging because the behavioral and physical changes classically associated with pregnancy are often not obvious ( Hornaday 1920; Hediger 1946; Lang 1975). Females have one pair of mammae located in the inguinal region.

Successful parturition in C. liberiensis takes place on land; many young animals in captivity have drowned, even in very shallow water (Stroman and Slaughter 1972; Lang 1975; von Houwald et al. 2007). Birth weights range from 3 to 8 kg with a mean of 5.7 kg ( Steinmetz 1937; Roth 1962; Nowak 1999; Stroman and Slaughter 1972; Lang 1975; Leutenegger 1978; Greed 1983). A review of all known births from 1919 to 1975 (n = 219) revealed an overall calf survival rate of 62%, with the great majority of deaths occurring within the first month of life ( Leutenegger 1978). Higher birth weights were positively correlated with survival rate beyond 3 months of age.

Figures

Fig. 1.—An adult female Choeropsis liberiensis, at the Center for Conservation of Tropical Ungulates. Photo by Gabriella Flacke.

Fig. 2.—Dorsal, ventral, and lateral views of skull and lateral view of mandible of an adult female Choeropsis liberiensis (ZFMK [Zoological Research Museum Alexander Koenig] MAM 1965.0570) from Côte d’Ivoire. Greatest length of skull is 339 mm. Images used with permission of the photographer Georg Oleschinski.

Fig. 3.—Geographic distribution of Choeropsis liberiensis and C. heslopi. Map prepared by Flora Ihlow, Zoological Research Museum Alexander Koenig, based on International Union for Conservation of Nature and Natural Resources Red List map data (Ransom et al. 2015).