Form

Pelage of Przewalskium albirostre consists mainly of needle-shaped guard hairs (Chen and Wang 1991) that are brownish at the base and have an off-white ring before a black tip (Flerov 1930). Rigidity and thickness of the pelage of P. albirostre exceeds all other species of deer, and hairs are “hollow, very brittle, slightly wavy and abruptly narrowed towards the base” (Flerov 1930:118). Guard hairs vary in length (mean in mm ± SD): 152.0 ± 22.6 on the back, 4.52 ± 0.54 dorsally, and 3.07 ± 0.65 on the legs; depth of the pelage is 33.5 ± 7.5 in winter (Sheng et al. 1993). Winter pelage is twice as thick as summer pelage and longer than that of other Cervidae, even moose (Alces alces—Geist 1998; Sheng et al. 1993). As in other Cervidae, the central medulla of guard hairs of P. albirostre has a lattice arrangement of cells: medulla width, 635 µm (range: 341–659 µm) and cortex width, 12.0 µm (Sheng et al. 1993). Conflicting accounts exist on the extent of an undercoat in P. albirostre; fine undercoat hairs were noted by Sheng et al. (1993), but others did not note a developed undercoat (Chen and Wang 1991; Flerov 1930, 1952; Jaczewski 1986).

The skull of P. albirostre is “large, massive, shortened, very long and relatively flattened [with] a maximum length of about 400 mm” (Flerov 1952:109; Fig. 3). It is differentiated from other cervids by wide posterior nasals, convex upper profile of the braincase, large lachrymal fossa, broad, flat palate, and short incisive foramina (Allen 1940; Flerov 1952; Geist 1998). Early descriptions include various skull measurements of individual specimens (Allen 1940; Blanford 1893; Flerov 1930, 1952; Sclater 1889), but Ohtaishi et al. (1993) provide the most thorough measurements (mean in mm ± SD) of 2–6 males and 24–27 females, all ≥5 years old: total length, 397 ± 9.7 and 364 ± 11.4; condylobasal length, 374 ± 8.8 and 335 ± 9.2; nasal length, 131 ± 8.1 and 115 ± 7.0; median palatal length, 214 ± 2.5 and 200 ± 6.5; maximum width of lachrymal fossa, 52.8 ± 2.2 and 45.8 ± 2.7; length of upper cheek toothrow, 107 ± 5.5 and 102 ± 3.6; length of lower cheek toothrow, 120 ± 8.6 and 112 ± 3.6; length of diastema, 91 ± 6.9 and 83 ± 4.7; and length of mandible, 299 ± 10.1 and 280 ± 6.4.

Fig. 3.

Ventral, dorsal, and lateral views of skull and lateral view of mandible of an adult male Przewalskium albirostre (Bristish Museum [Natural History], specimen 1892.16.11.1). This is the type specimen of Cervus thoroldi collected by Dr. G. W. Thorold and described by W. T. Blanford (1893:plate 34), now considered a synonym of P. albirostre. The annotation of “Pr. albirostris” on the top of the skull no doubt postdated C. Flerov's (1930) naming of the genus Przewalskium. Greatest length of skull is 373.1 mm.

During growth, antlers are covered in modified skin, or velvet, that nourishes the growing bone (Bubenik 1993); velvet is particularly prized in Chinese medicine (Koizumi et al. 1993). Antlers of male P. albirostre are whitish (Flerov 1952), yellowish, or brownish, likely dependent on the vegetation used to rub off the velvet. Antlers are grown annually, are shed early in March, begin to grow almost immediately, and attain full growth in late August and September (Geist 1998; Schaller 1998). Characteristically, all antler tines lay on the same plane, unlike antlers of male red deer whose 4th tines branch inward (Ohtaishi et al. 1993). Length and mass of antlers and the number of tines (maximum  =  7) increase with age; maximum circumference of beam is reached at about 4 years of age (Ohtaishi et al. 1993). Total antler length along the beam (mean in mm ± SD) relative to the number of tines is: 3 tines, 48 ± 5.4; 4 tines, 66 ± 7.8; 5 tines, 92 ± 8.2; and 6 tines, 93.8 ± 18.5 (Ohtaishi et al. 1993). Total antler mass (mean in kg ± SD) is: 3 tines, 0.45 ± 0.13; 4 tines, 0.85 ± 0.21; 5 tines, 2.9 ± 1.5; and 6 tines, 3.0 ± 0.86 (Ohtaishi et al. 1993). Males can be grouped in 4 general age classes (I–IV) based on their antler characteristics: I, yearlings with spike antlers; II, juveniles with 2 or 3 antler tines; III, subadults with 3 or 4 tines; and IV, adults with 4–7 tines (Miura et al. 1993).

Dentition of adult P. albirostre is typical of cervids: i 0/3, c 1/1, p 3/3, m 3/3, total 34 (Flerov 1952; Fig. 2). Teeth are small compared with species of Cervus and Rusa, dental row is “short and flat,” cheek teeth are subhypsodont, roots of P2 are fused, lower canines are incisoriform, and upper canines in males and females are rudimentary (Brokx 1972; Flerov 1930:118; Flerov 1952). P. albirostre can be aged by patterns of replacement of deciduous teeth and eruption of permanent teeth in ≤2.5-year-old individuals; for example, m1 and M1 erupt at about 3 months old; i1, m2, and M2 erupt at about 1.5 years old; and m3 and M3 erupt at about 2.5 years old (Ohtaishi et al. 1993). Other methods for aging P. albirostre include the appearance of lingual and buccal cervical lines in permanent molars of 2.5–8.5 year olds, assessment of wear patterns as individuals age (particularly useful for >8.5 year olds), and counts of cementum annuli of incisors and molars (Ohtaishi et al. 1990).

Function

In captive herds at elevations of 2,300–4,200 m in Qinghai, China, basic hematology of Przewalskium albirostre is: hemoglobin, 126.0–145.6 g/l; packed red blood cell volume, 42.13–46.75%; and mean corpuscular hemoglobin concentration, 28.39–36.60% (Li et al. 1993). Representative serum constituents of P. albirostre in captivity in Germany are: protein, 61.9–68.1 g/l; alkaline phosphatase, 189.6–428.9 IU/l; creatinine, 181–190 mmol/l; sodium, 148–165 mmol/l; potassium, 5.1–7.5 mmol/l; calcium, 2.2–2.4 mmol/l; and phosphorus, 2.1–3.4 mmol/l (Seidel 1993). Additional measurements of blood and serum chemistry were provided (in Chinese) by Wu and Wang (1999).

Similar to the wild yak (Bos mutus—Leslie and Schaller 2009) and other Tibetan species, P. albirostre is adapted to living in high-elevation areas with associated low temperatures, intense solar radiation, and limited oxygen availability. Vascular adaptations of P. albirostre include a high number of relatively small red blood cells (mean diameter, 4.37 µm ± 0.13 SD compared to 5.6 µm in cattle), low red blood cell packed volume, and low mean corpuscular volume, resulting in greater surface area of red blood cell membranes to enhance oxygen transport; plasma viscosity is low, which also speeds oxygen transport (Li et al. 1993). Hypertrophy of the right ventricle of the heart—an expectation if hypoxia from elevated pulmonary arterial blood pressure at high elevations occurs (cf. Heath et al. 1984)—has not been observed in P. albirostre (Li et al. 1993).

The modulus of fineness (MOF—Poppi et al. 1980) of fecal particle size from digesta residue suggests that P. albirostre (MOF  =  2.12) is an intermediate feeder (Clauss et al. 2002). Percentage of fecal particles passing through various sieve sizes are: 4-mm sieve, 1.17%; 2-mm sieve, 1.72%; 1-mm sieve, 4.47%; 0.25-mm sieve, 15.06%; and <0.125-mm sieve, 45.69% (Clauss et al. 2002).

As occurs in other large Tibetan mammals (Leslie and Schaller 2008, 2009), enlarged nasals and associated cranial features of P. albirostre permit greater passage of rarified air in its high-elevations habitats (Wu and Wang 1999). Its dense pelage dominated by long, air-filled guard hairs protects it from the extreme low temperatures of the Tibetan Plateau.

Population characteristics

Densities of Przewalskium albirostre are difficult to estimate because of fragmented small populations and the expansive nature of the Tibetan Plateau. Density estimates of P. albirostre in Qinghai and Sichuan vary greatly from 0.01 individuals/km² to 4.41 individuals/km² (Koizumi et al. 1993:table 1; Schaller et al. 1988; Yu et al. 1993; Zheng et al. 1989). Disparity in density estimates reflects some combination of real densities, sampling differences relative to methodology, size of the area sampled, time of year, and human impacts (e.g., protection, livestock grazing, and poaching—Harris 2008; Schaller 1998).

In a sample of 45 P. albirostre collected in Qinghai and Sichuan in 1986–1988, maximum age of females was 12 years in the wild and 19 years in captivity; maximum age of males was 9 years in the wild and 12 years in captivity (Koizumi et al. 1993). Two females lived 21 years and 2 months and 21 years and 3 months in captivity at the San Diego Wild Animal Park (Weigl 2005). From observations during a variety of months in 1985–1988 and 1990, sex ratios of P. albirostre ranged from 26 males : 100 females in Qinghai to 42 males : 100 females in Sichuan (Kaji et al. 1993). During rut in Zhenda, Sichuan, when males and females occurred together, Miura et al. (1993) noted 59 males : 100 females, as calculated by Schaller (1998). Low ratios of males in some areas are thought to reflect selective poaching for antlers, particularly in velvet, although population statistics from areas around monasteries and in heavy hunted areas are lacking (Schaller 1998).

Space use

Przewalskium albirostre occupies the highest elevation (up to 5,100 m) and most open habitats of any deer species on the Tibetan Plateau. Among other deer species of the world, only the Peruvian huemel (Hippocamelus antisensis) occurs at such high elevations, rarely below 3,960 m and up to 5,000 m in South American Andes (Putz 2003). P. albirostre generally prefers mosaics of grassland, shrubland, and forest, often above timberline, and alpine meadows and even arid, treeless areas (Schaller 1998; Wu and Wang 1999). Unlike other deer species, its “short-legged body form, large haunches, and large, blunt hooves” allow it to use steep mountainous terrain to escape danger (Geist 1998:97). P. albirostre often is found in the same areas as other mountain specialists such as the Tibetan argali (Ovis ammon hodgsonii—Fedosenko and Blank 2005) and blue sheep (Pseudois nayaur—Wang and Hoffmann 1987), which have similar habitat preferences (Cai 1988; Zheng et al. 1989). In summer, high-elevation sedge meadows may be preferred by P. albirostre (Harris and Miller 1995).

Unlike the strongly migratory ungulates of the Tibetan Plateau such as the Tibetan antelope or chiru (Pantholops hodgsonii—Leslie and Schaller 2008; Schaller 1998) but similar to the wild yak (Leslie and Schaller 2009), P. albirostre moves seasonally from high-alpine summer ranges to low-elevation winter range and has been described as nomadic (Kaji et al. 1989). Miura et al. (1989) observed a herd of 25 P. albirostre that moved across a relatively large area of 35 km² in only 9 days in August. Harris and Loggers (2004) suggested that P. albirostre moved among the large drainages in Yeniugou, Qinghai, unlike wild yak (Leslie and Shaller 2009), and although not confirmed definitively, they believed that 1 herd (based on comparable group composition) moved 21.5 km in 9 days. Like other cervids, P. albirostre is adept at swimming. Miura et al. (1989) and Koizumi et al. (1993) noted summer use of islands in Gyaring Lake, Qinghai, which required up to a 25-min swim.

Diet

Przewalskium albirostre is an herbivorous ruminant and is foremost a grazer that is not particularly selective of species or plant parts (Takatsuki et al. 1988). As many as 95 species of plants have been noted in the diets of P. albirostre, of which 35 species are regularly eaten (Wu and Wang 1999). Foraging preferences are understood only from limited direct observations and evaluation of stomach contents and feces (Cai 1988; Harris 1993; Harris and Miller 1995; Takatsuki et al. 1988). Overall, grasses (Stipa and Poa) and sedges (Kobresia and Carex) dominate the summer diet of P. albirostre, not unlike other ungulates of the Tibetan Plateau (Leslie and Schaller 2008, 2009); 1 composite fecal sample from mesic sedge meadows in Yeniugou, Qinghai, suggested a summer diet of <90% Kobresia (Harris and Miller 1995). Woody plants (Rhododendron, Salix, and Potentilla) also are eaten depending on availability (Cai 1988). P. albirostre drinks water (Cai 1988), but specific requirements are unknown.

As occurs for other Tibetan ungulates (Leslie and Schaller 2008, 2009), diets of P. albirostre probably are deficient in protein (<6%) from October to May because of low forage quality on the Tibetan Plateau (Schaller 1998). Forage also may be deficient in minerals such as phosphorus and potassium depending on local soil fertility (Schaller 1998).

Interspecific interactions

The Tibetan Plateau has a rich ungulate fauna (Hoffmann 1991), although many species have declined numerically because of human activities (Harris 2008; Schaller 1998). P. albirostre occurs at higher elevations than other deer species of the Tibetan Plateau (MacKinnon 2008) but is most likely to be sympatric with red deer (Schaller et al. 1996) and alpine musk deer (Moschus chrysogaster), albeit the 3 species are rarely observed together (Kaji et al. 1993). P. albirostre can occur generally in the same areas as Tibetan argali, blue sheep, wild yak (Leslie and Schaller 2009; Schaller and Liu 1996), and Tibetan gazelle (Procapra picticaudata), and less so with Tibetan wild ass or kiang (Equus kiang—St.-Louis and Côté 2009) and chiru (Cai 1988; Harris and Miller 1995; Kaji et al. 1993; Schaller 1998). Because of the rarity of many of these species in the eastern Tibetan Plateau, direct interactions among them are now reduced. As in mixed ungulate assemblages elsewhere, Tibetan species likely partition food and space, relative to size and digestive capabilities, to minimize competition (Harris and Miller 1995; Schaller 1998; Schaller et al. 1991), but competition with livestock remains the most common threat.

Scant information exists on the predator-prey dynamics on the Tibetan Plateau, and current dynamics are a product of reduced populations of both due to various human activities (Harris 2008; Schaller 1998). Predators within the range of P. albirostre include the wolf (Canis lupus), snow leopard (Uncia uncia), lynx (Lynx lynx), and Tibetan brown bear (Ursus arctos pruinosus), but only wolves and snow leopards have been documented to consume P. albirostre, albeit infrequently. Schaller (1998) found P. albirostre in only 1 of 46 fecal samples of snow leopards and 1 of 29 fecal samples of wolves in Qinghai. No information on diseases and parasites of P. albirostre is available.

Grouping behavior

Przewalskium albirostre is a gregarious herding ungulate that typically occurs in 3 types of groups: male groups outside the breeding season, female groups outside the breeding season, and mixed groups throughout the year (Fig. 4), particularly during rut in September–November, depending on location (Miura et al. 1993; Schaller 1998). Group sizes in Qinghai and Sichuan varied from 1 to 169 individuals; 97% of groups contained ≥10 individuals and tended to be the largest during rut; and group cohesion appears to be high (Miura et al. 1993). Groups observed by Schaller (1998) in eastern Tibet and Qinghai ranged from 8 to 92 individuals. P. albirostre rarely occurs alone; Miura et al. (1993) observed only 2 males alone out of 48 observations. Historically, group sizes were larger; for example, groups of up to 800 individuals (200 males) were observed in Zhidoi, Qinghai, in the late 1960s, but by the mid-1980s, few P. albirostre were encountered (Kaji et al. 1989). P. albirostre is crepuscular in its daily activity patterns (Cai 1988; Miura et al. 1989).

Fig. 4.

Mature male Przewalskium albirostre guarding an estrous female by standing up against her in a typically mixed group during rut in September–November, Qilian Mountains, northeastern Qinghai; note the diagnostic reddish buff rump patch that circumscribes the short, similarly colored tail. Photograph by G. B. Schaller used with permission.

Reproductive behavior

Przewalskium albirostre displays sexual segregation outside of rut (Miura et al. 1993; Schaller 1998). The breeding system of P. albirostre is “micro-scale” harem-based, wherein multiple mature males (class IV, 4–7 tines) often occur in mixed groups and either maintain small harems within the group or tend individual females when they are receptive to breeding (Miura et al. 1993:227). In Zhenda, Sichuan, the number of mature males per mating aggregation was 1–8, and females were divided into as many as 8 harems depending on the size of the group (Miura et al. 1993). Such mating aggregations often have peripheral groups of subordinate males (Miura et al. 1993). In captivity, male P. albirostre >5 years old breed, but 7-year-old males display the most rutting behavior (Koizumi et al. 1993). During rut, adult males display to each other with behaviors typical of cervids (Geist 1998): regular vocalizations, marking by antler thrashing with open preorbital glands, urine spraying, pawing, and aggressive interactions (e.g., head-down postures, rushing, chasing, parallel-walking displays, and clashing with joined antlers, but not a head-up threat—Koizumi et al. 1993; Miura et al. 1988, 1993). Biting and kicking also occur among males (Wu and Wang 1999). Males wallow in dry and wet spots; muddy wallows are conspicuous and sought out in particular by mature males (Fig. 5; G. B. Schaller, pers. comm.).

Fig. 5.

Male Przewalskium albirostre standing near a muddy wallow in Qilian Mountains, northeastern Qinghai; mature males seek out such wet wallows, a behavior that departs from some other large deer species. Photograph by G. B. Schaller used with permission.

In Qinghai and Sichuan, courtship behavior among wild P. albirostre was always the purview of mature males with fully developed antlers (Miura et al. 1993). Courtship behaviors, typical of cervids (Geist 1998), include lip-curling, licking, low-neck stretch with rapid tongue flicking, and herding or tending with head-down postures and rushing (Miura et al. 1993). When receptive, males rest their chin on the female's back prior to mounting; copulation is accomplished with a “single violent thrust,” but only after an average of 6.7 unsuccessful mountings based on 23 breeding events observed in Qinghai and Sichuan (Miura et al. 1993:227). Males guard females (Fig. 4), even from their offspring from the year before, for ≥3 h postcopulation (Miura et al. 1993).

Yu et al. (1993) provided the following details on behavior of female P. albirostre during parturition. Females separate themselves from other females and seek secluded places to give birth and hide their neonates. Among 37 birthing sites in Sichuan, 46% occurred in Rhododendron shrub habitats, 32% occurred at the ecotone of grassland–shrub habitats, and 22% occurred in forested, but not dense, areas; east-facing slopes at elevations of 3,850–4,450 m were preferred. One wild female was observed giving birth in a standing position, and she consumed the afterbirth and licked the amniotic fluid on the ground. Females vigorously lick, muzzle, and push the anogential region of their neonates, encouraging them to stand soon after birth; 2 wild neonates stood 30–38 min postpartum (Yu et al. 1993). Neonates begin to move away from the birth site within a couple of hours; during their 1st day of life, neonates may move 40–80 m from the birth site, and during the next 4 days, they may use several hiding sites per day and move 80–200 m between them (Yu et al. 1993). Females visit their offspring to nurse in the morning and evening; otherwise, they remain apart. After about 2 weeks, neonates follow their mothers and begin associating with other conspecifics (Yu et al. 1993).

Communication

Vocalizations of Przewalskium albirostre include alarm barks by both sexes; “gurgling roars” and “short, whiny bugles” (Schaller 1998:149) and “low quavering” growls of males in rut; and grunts, bleats, and “soft, mew-like” sounds of females and their offspring (Miura et al. 1988:107, 1993:225). Roars of males consist of 3–5 syllables, last for about 5 s with a frequency range of 1.2–1.8 kHz and 4–6 harmonic bands, and can be heard >500 m away (Miura et al. 1988, 1993). Grunts of females when approaching their offspring last about 5 s with a frequency range of 0.4–2.2 kHz (Miura et al. 1988). Although nonvocal, both sexes and all ages older than calves emit “resonant snapping sounds” from their carpal joints that can be heard ≤30 m away (Miura et al. 1988:107; Schaller 1998). The function of this sound is unknown; it is absent in most deer but present in caribou (Rangifer tarandus) and Père David's deer (Elaphurus davidianus—Schaller 1998; Schaller and Hamer 1978).