TianZi Asia
TianZi Biodiversity Research & Development Centre


The Feeding Ecology of Giant Pandas and Asiatic

Black Bears in the Tangjiahe Reserve, China





The Asiatic black bear (Ursus thibetanus) has a wide though patchy distribution from Iran, Afghanistan, and Pakistan eastward along the Himalayas to Indochina and across China to northeastern Russia. By contrast, the giant panda (Ailuropoda melanoleuca) survives only along the mountainous eastern edge of the Tibetan plateau, confined to an area totaling about 29,500km2 mainly in Chian’s Sichuan province, but also southern Gansu and Shanxiprovinces (Figure 8.1). In these mountain forests pandas and black bears of the subspecies U.t.mupinensis (Ma 1983) are sympatric.

The giant panda and the Asiatic black bear are much alike, being solitary carnivores of similar body build and size. Both are mainly herbivores (Bromlei 1973;Schaller et al.1985),in spite of the fact that they have the short, relatively unspecialized digestive tract of carnivores. Lacking the microbial digestion in rumen or caecum typical of most herbivores, they are unable to break down cellulose and other structural carbohydrates composing the cell walls of plants. Most plants consist primarily of cell walls and water; therefore, the animals derive their nutrition principally from cell solubles (sugars, starches, lipids, protein). Because solubles represent only a small fraction of a plant , much bulk must be consumed to fulfill daily nutritional requirements. This chapter examines the strategies of panda and black bear for living not only as herbivores but also for doing so sympatrically.

In the Wolong Natural Reserve, where we first conducted panda studies, we noted that both bears and pandas seasonally forage on shoots of one bamboo(Fargesia spathacea). Yi[1985]renamed this bamboo Fargesia robusta , but we retain the old name to avoid confusion with our previous publications.) But because bears were only sporadic visitors to our study area, we obtained no data on the amount of ecological overlap or possible competition between these two similar species. Such overlap might consist of spatial usr of an area, of a similar daily activity schedule within that area, and most important, of the same food habits. The panda subsists primarily on bamboo and ,in fact, has evolved two specializations for processing this plant efficiently: the forepaws are adapted for grasping bamboo stems through the addition of a sixth digit or "thumb"-an enlarged wrist bone-and the posterior premolars and molars are broad and flat, modified for crushing bamboo(Davis 1964).The black bear lacks such morphological specializations. Its diet is more varied than the panda’s, if data from Russia (Bromlei 1973), India (Schaller 1977),and China (Wu 1983)are indicative, consisting of herbs, fruits, and nuts. Feeding adaptations and strategies determine how an animal meets its nutritional requirements for maintenance, growth, and reproduction, and these ,in turn, affect movements, activity cycles and other aspects of existence. To what extent do the panda and the black bear overlap ecologically within their area of sympatry?

We studied both species in the Tangjiahe Natural Reserve of northern Sichuan (Figure 8.1).The research continued through 1987; this chapter, based on work conducted between March 1984 and March 1985,is limited to a preliminary discussion of the variety, abundance, dispersion, seasonality, and nutritional quality of panda and black bear foods, and to the effects of such variation in food supply on the behavior of these sympatric species. The Tangjiahe study is an extension of the cooperative China-World Wildlife Fund panda project, initiated in 1980 in the Wolong Narural Reserve, as described by Schaller et al.(1985).All data from Wolong in this report were taken from that publication.


The Study Area

The Tangjiahe Narural Reserve, established in 1978,lies in the Min Mountains of northern Sichuan bordering Gansu province. It extends over about 300 km2 of rugged ridges and narrow valleys at elevations of from 1200 to 3800 m. Of the two main drainages in the reserve, we selected about 75 km2 of the upper Beilu valley as our general study area. The eastern end of this area was at Maoxiangba, the reserve headquarters at 1420 m. From there a road winds westward for 14 km up the Beilu valley, where near the mouth of the Hongshi valley it becomes impassable to vehicles. We concentrated our activities in about 17.5 km2 of the upper Beilu and lower Hongshi valleys between elevations of 1520 and 2300 m; our research base was located there at 1760 m(Figure 8.2).



The vegetation shows a vertical zonation similar to that of the Wolong Reserve, but being over 200 km farther north ,each of the three zones extends about 300 m lower on the slopes.(1) An evergreen and deciduous broadleafed forest occurs below 1700 m. The evergreen trees Lindera communis and Cyclobalanopsis oxyodon are prominent, as are the deciduous beech (Fargus longipetiolata) and oak (Quercus glandulifera). (2) Between 1700 and 2100 m a mixed coniferous and deciduous broadleaved forest predominates, although various species from the previous zone persist, especially o south facing slopes. Several species of maple (Acer), Litsea , Hydrangea, and Viburnum are common, as is birch (Betula utilis, B.albasinensis) and cherry (Prunus serices, P.brachypoda); there are evergreen rhododendrons, ranging in size from low shrubby species to trees; and, among the conifers, pine (Pinus armandii) favors dry, southern exposures, and hemlock (Tsuga chinensis) and spruce (Picea brachytyla) moist, northern ones. In valley flats and on lower slopes, where humus is deep, lush herb meadows thrive, providing an important source of bear food. (3) A subalpine coniferous zone, with hemlock and spruce at lower elevations and fir (Abies faxoniana) higher up, begins at 2100-2300 m, depending on exposure, and extends to timberline; rhododendron and birch are the main broadleafed trees. On some slopes forest gives way to tussock grassland at only 2500 m, but generally the upper limit of tree growth is approximately 3200-3300 m and appears to be edaphically determined.

Bamboo is a critical resource for pandas. Since the taxonomy of bamboo in the Min Mountains remains unsettled , we follow the terminology of Yi(1985), who examined material from Tangjiahe. There is little bamboo below 1600-1700 m in our study area, except for patches of Bashania (Indocalamus) fargeii (G.G.Camus) Kengf. et Yiaround the mouth of the Shiqiao vally at 1580m. The dominant bamboo from 1600-1700 m to 2200-2300 m is Fargesia scabrida Yi which may cover whole hillsides in the Hongshi and some other valleys. This bamboo has flowered and died in patches between 1972 and the present, with a peak dieoff in 1975. The dieoff affected the bamboo primarily on the lower slopes of some valleys, leaving ridge tops andother valleys, shch as the Hongshi, almost unaffected. This bamboo, therefore, exists as a mosaic, ranging in height from tiny seedlings to decadeold growth averaging less than 1 m tall to adult stands. From about 2100-2200 m and upward to timberline is a third bamboo, a smallleafed species closely resembling F.nitida, which has been named F.denudata Yi. It flowered so extensively in the mid-1970s that only some stands on ridges above 2600 m remained unaffected; pandas have been largely deprived of this resource since the dieoff. At least two other species-F.rufa Yi and Phyllostacbys species-occur locally in Tangjiahe, neither of importance to our study.

Human activity has greatly modified the vegetation in many areas below 2200 m, especially in the main valleys. The impact has been more severe on the relatively gentle southern and western exposures than on the more precipitous northern and eastern slopes. Cultivation was once extensive (and persisted downstream of Maoxiangba until 1986). Long-abandoned fields are now overgrown with trees (Populus spp.) and shrubs (species of Salix, Spiraea, Deutzia, Rosa, Rubus); recent fields are covered densely with herbs, conspicuously the tall Artemisia subdigitata. Buddleia davidii, willow, and other shrubs have replaced forest in the main valleys and some side ones. Roads were built into several valleys between 1965 and 1978 and timber extracted ftom all accessible parts, slopes being either clearfelled or selectively logged for conifers. Some slopes have been replanted with pines. This habitat destruction has affected both pandas and black bears. For pandas the impact has been entirely negative. Formerly cultivated slopes remain virtually devoid of bamboo, and clearfelling of timber has resulted in dense bamboo thickets without tree canopy which are little favoredby pandas. Logging also removes conifers 1 m or more in diameter that, if hollow at the base, could serve as maternity dens; we saw no potential den trees in our study area. The impace on black bears has been mixed. Stands of oak ,an important food source, have been decimated, often leaving only fringes of trees along ridge crests; however, the secondary growth on disturbed sites provides bears with Rubus species and other fruit.



Winter lasts from November to March ,months during which temperatures dropped below freezing at our camp. The coldest month was January, with an average daily maximum of 1.8 and minimum of –5.3; the absolute maximum was 8 and minimum -11.The first flower of spring � a primula - was seen on March 18, and the last snow of winter fell on March 23. During April and early May herbs grew rapidly , trees leafed out ,and rhododendrons blossomed. It rained on at least 15 days every month between May and October; 93.5% of the total 12-month precipitation of 1130 mm fell during this rainy season. June to August were the warmest months ,with an average daily maximum of 22.6-24.7_ and minimum of 13.1-14.2_;the absolute maximum was 30_ and minimum 10_. The first yellow leaves were evident in mid-september, and a month later autumn coloring was at its peak. Most deciduous trees had shed their leaves by mid-November. December 14 brought the first heavy snowfall to the valley.



Since we observed black bears only eight times and pandas 26 times, most of our data are based on examinations of feeding sites, droppings, and other spoor. Monthly samples of panda droppings were analyzed to determine food selection. Fargesia scabrida bamboo samples were collected monthly at two sites for nutritional analysis. Both sites were at 2000 m , one of mature bamboo and the other of seedlings about nine years old .(Positive species identification of seedlings has not yet been made.) Collecting methods and analyses of feeds and droppings follow those described in Schaller et al.(1985). Bear droppings were given an ocular examination, and the percentage of each major component estimated. To determine bamboo shoot and stem mortality rates,48 plots (2 m2 and 4 m2),totaling 112m2, were established in unflowered F.scabrida in and near the Hongshi valley. Sites differed in degree of slope, altitude, exposure, and percentage of canopy. The plots were established in March-April 1984;at that time all new stems (shoots of 1983)were marked and all old stems(two years old and older)were counted. From July to September, during the shoot-growing season, plots were visited at least once every month and shoot mortality noted. Dead stems were tallied in October and again in March 1985,completing one annual cycle.

In 1984 we captured two pandas and two black bears in traps baited with goat meat, sedated the animals either with CI-744 or ketamine hydrochloride, radio-collared them , and subsequently monitored their movement and activity using equipment and techniques as reported in Schaller et. al.(1985).One panda, Xue, a middle-aged female without infant weighing 67.3 kg, was caught on 14 December in one of the seven box traps of logs we had built in the Hongshi valley. The other panda, Tang, an adult male, was collared on 8 June, having been captured in a cave into which he retreated when we surrounded his bamboo thicket; on 1 June 1985,when we replaced his collar with a new one, he weighed 67.7 kg. Both pandas were considerably lighter than adults in Wolong , where two females weighed 86 and 89 kg and two males,97 and 107 kg. Both bears were caught in Aldrich foot snares. Kui, an adult male of undetermined weight-Bromlei (1973)found that adults generally range from 100 to 150 kg-was captured on 23 July. Chong, a subadult male estimated to be almost three years old ,was snared on 5 November; he weighed 70.5 kg. In contrast to the placid pandas, both bears were aggressive, lunging at us and roaring.

Radio-collared pandas were usually contacted daily, except that we seldom found Tang after he moved out of his usual haunts during the summer. Chongís signal was often received during the month between his capture and hibernation, whereas Kuiís was sometimes lost for weeks when he traveled out of the reserve. We monitored activity on a 24-h basis every 15 min, 96 signal readings per day. Tang was monitored for 19 days, Xue for 9, Chong for 7, and Kui for 3.


Giant panda

We estimated that about ten pandas frequented our 7 by 2.5km main study area some animals only part time. (Quoting other sources Schaller et al. 1985 gave a population estimate of 100-140 pandas for Tangjiahe; in 1985 a census coordinated by Hu Jinchu revealed 50-60 pandas, a figure that agrees with our impressions formed in 1984.)Of these, Tang spent much of the year at 1500-1600 m in Bashania farhesii bamboo, whereas Xue and others were in F.scabrida, usually above 2000 m, and some seasonally high up in F. Denudata. The three bamboo species differ in their annual cycle of shoot production; this in turn affects the movements, food habits, and nutrient intake of pandas.

B.fargesii is alarge-leaved bamboo, 2-3 m tall, with stems up to 1-1.5cm thick at the base. F.scabrida averages about 1.8 m high, with some stems 3-4 m high although stands on dry , logged slopes may average only 1 m. Stems rarely reach a basal thickness of 1 cm ,most being about half that .There was a mean of 27.5 stems/m2 on our plots, and the mean above-ground biomass (fresh weight)-based on sampling seven plots of 1 m2-was 1479 g/m2. New shoots appear in mid-July and reach full height in late September. Pandas also eat F.scabrida seedlings after they reach a height of 40 cm or more. A 1 m2 plot of seedlings, estimated to be nine years old, yielded 172 stems with a biomass of 609 g. Mean stem height was 58 cm, with the tallest stem 157 cm; mean stem diameter was 0.2cm. F.denudata resembles F.scabrida in stem height and thickness. Although it grows at higher elevations than F.scabrida. its shoots appear from mid-June through July.


Food Habits

Pandas may consume plants other than bamboo, and they also eat meat when available (Hu 1981;Schaller et al.1985).At Tangjiahe an animal once are the juicy stems of wild parsnip (Heracleum moellendorffii) - a favored bear food - and once chewed on old skin and leg bones of a tufted deer (Elaphodus cephalophus). F.scabrida bamboo was the principal food of most pandas. However, animals showed marked seasonal preferences for certain parts of the plant (Figure 8.3). They preferred leaves over stems from October to March in a ratio of 2:1,By mid-April animals had decreased their leaf consumption markedly ,and this trend continued until June, when they avoided leaves . They again ate leaves during July. Several of our study animals also ascended into F.denudata that month. On 30 July 45 droppings of F.denudata at 2800 m were composed of 21% new shoot and 79% leaf. During August and September animals consumed primarily new F.scabrida shoots.

Tang, living mainly in B.fargesii bamboo, displayed a different pattern of dietary preference. From September to April he ate leaves almost exclusively. On 26 October we observed him for 1 h(Figure 8.4).He either bent stems with a forepaw toward his muzzle and ate the leaves off, or he detached the stem with a bite and ,holding it upright with one forepaw, pushed the leafy branches into his mouth with the other. He did not select stems of particular age: after 2.5 days in the observed bamboo patch, he had eaten at least some leaves from 68% of the stems. During May and early June, Tang foraged on B.fargesii shoots . After shoots had grown tall and hard, he abandoned the valley for ridges, first to the east and then south. We found him on 4 July at 1990 min F.scabrida. His droppings revealed a diet of 23% leaf and 77% stem. In early September he returned to his usual B.fargesii haunts. (D.Reid told us that Tang also spent June-September 1985 at high elevations.)

Wolong pandas usually peel the enveloping sheaths off F.spathacea shoots before eating them, perhaps because sheaths of this species are unusually hairy. By contrast. Tangjiahe animals often consume the shoots of all species without peeling them. In Wolong pandas also select thick shoots,_1 cm in diameter. Shoots are seldom that thick in Tangjiahe. Taking the F.scabrida shoots in our plots as a sample ,we found that most shoots eaten were 0.7-0.8 cm thick, a size Wolong pandas seldom consumed. There was, however, selection against thinner shoots. As in Wolong , insect predators took a significantly greater proportion of thin shoots(diameter_0.7 cm)than did pandas(Figure 8.5).Rodents selected much like pandas. Tang displayed a similar preference for thick B.fargesii shoots, selecting for shoots_0.9 cm and against those_0.8 cm. It appears that pandas merely chose the thickest shoots in an area, with a lower limit of about 0.6 cm.

There was much shoot predation. One or more shoots were destroyed in 96% of our plots by insects, in 42% by pandas, and in 35% by rodents. Of the 996 shoots produced, 43.5% were destroyed between July and September, a mortality figure similar to that in Wolong (Table 8.1). Nevertheless, the number of bamboo stems increased during the year. There were 3028 stems in the plots in April 1984, of which 256(8.5%) died during the following 12 months. Of the shoots produced in 1984,535 survived, an increment of 279 new stems (9.2%).This increase appeared due to a large number of 1984 shoots: only 3 shoots per m2 of the 1983 crop survived to the age of one year, whereas 4.8 shoots per m2 of the 1984 crop did so.


Nutritional Content of Bamboo

Approximately 90% of protein, carbohydrates, and other cell solubles are nutritionally available to pandas (Dierenfeld et al.1982). Of the cell wall components, cellulose and lignin are indigestible; however, pandas can break down a fraction of the hemicelluloses digesting 18-26%, depending on season(Schaller et al.1985).

F.scabrida leaves from both adult bamboo and seedlings have more protein, ash(minerals and salts), and hemicelluloses and less cellulose and lignin than do stems(Figure 8.6). F.denudata and B.fargesii leaves are chemically similar to F.scabrida (Table 8.2). New F.scabrida shoots are of lower average nutritional quality between July and September than are leaves; the percentage of total cell solubles in shoots is similar to that in stems(Figure 8.6).A tall B.fargesii shoot ,collected on June 1, had 8.7% crude protein and 5.5% other cell content, similar to F.scabrida shoots.

All bamboo species in Tangjiahe retain green leaves throughout the year. Monthly samples show that the chemical composition of bamboo leaves and stems remains quite constant at all seasons as illustrated for crude protein in F.scabrida leaves (Figure 8.7). The protein level in new F.scabrida shoots decreases, however, as shoots grow and harden until it is similar to that in stems; there is a concomitant increase in the percentage of cellulose.

Wolong pandas seldom ate the leaves of Sinarundinaria fangiana bamboo between April and June. Silica (Sio2) levels in leaves reached their highest levels (4-5%) during those months, then dropped to low levels from July to October, a period when pandas selected for leaves. In an attempt to explain this change in food selection, Schaller et al.(1985) suggested that silica-which can inhibit digestion (Van Soest 1982) - may be implicated .Our data from Tangjiahe do not support this idea. Although Tangjiahe pands selected against F.scabrida leaves also from April to June, silica levels fluctuated little during the year and were always higher than at Wolong , even during the months when animals favored leaves(Figure 8.3).The average annual silica level was 7.4% in leaves and 0.1% in stems. The reason why pandas avoid leaves in spring remains unexplained.

The water content of F.scabrida leaves and stems is 40-60%, and in shoots it is at least 75%; water content of droppings is about 70% when animals forage on leaves and stems, and 75-85% or more when they are on a shoot diet (Figure 8.8).Pandas thus eliminate more water in their droppings than they obtain from their food, except for a month or two each year when they eat shoots. Moist droppings may be essential for smooth and rapid passage of the coarse forage through the digestive tract. Oxidation of feeds forms metabolic water, but, in addition, pandas may need to drink at least once a day (Ruan and Yong 1983).

In sum, bamboo contains high levels of indigestible cellulose and lignin(35-65%) and partially digestible hemicelluloses (20-35%),and low levels of readily available nutrients as part of the cell contents(12-24%),making it a food of poor quality. But since nutritive content remains constant all year, bamboo represents a predictable food source. Analyses of two Wolong bamboo species gave results similar to those from Tangjiahe, as did analysis of two introduced species from Washington, D.C.(Dierenfeld 1981).



Most of a panda’s day is devoted either to resting or to collecting, preparing, and eating bamboo; other activities, such as traveling and grooming, consume only about 4% of the day. In Wolong 300 days of 24-h activity monitoring of several individuals showed that animals may be active or inactive at any time of day or night. Pandas were, on the average, inactive for 9.8h(41.6%) of the day. Part of this time was devoted to one or two long rest periods lasting 2-4h or more. Pandas were active for 14.2 h(58.4%) of the day, a figure that remained relatively constant throughout the year. Daily activity reached its lowest level between 0800 and 0900h and after 1900h, and its highest level between 0400 and 0600h and between 1600 and 1900h. These activity peaks near dawn and at dusk were similar all year, regardless of amount of daylight. Although individuals showed no significant differences in activity levels, they often displayed idiosyncratic patterns seemingly unrelated to age, sex, or other obvious factors.

In Tangjiahe winter data for Xue showed that her average probability of activity in January was 0.53,in February 0.68 and in March 0.36 for a mean of 0.52, similar to Wolong animals (0.58).She had two activity peaks, one between 0200 and 0300h, and the other between 1900 and 2100h; her low activity between 0700 and 0900h was similar to that of Wolong animals (Figure 8.9).

Tang was less active than any panda we have monitored in Wolong and Tangjiahe: his average probability of activity was only 0.43. There was little variation between September and March(December excluded because of lack of data),average probability of activity ranging from 0.37 to 0.46 ;in June ,the only other month during which we monitored Tang, the figure was an exceptionally active 0.79 from only one day’s data. Tang’s sedentary habits for most of the year and his preference for leaves-leaves are less time-consuming to eat than stems-probably account in part for his restful existence. His 24-h pattern was also unusual in that he tended to be inactive between 2100 and 0800h and showed only one prolonged daily peak between 1200 and 1800h(Figure 8.9). He sometimes rested for long periods. On 14 October, for instance, he began a rest at 1830h that lasted until 1215h on 15 October(nearly 18 h),so long that we became concerned about his health; after feeding, he also slept during the following night from 2100 to 0730 h(10.5h), yet appeared healthy.


Land Tenure

Home range sizes of six Wolong pandas varied from 3.9 to 6.4 km2. Ranges of males were only as large as, or slightly larger than, those of females. Even though its range was small, an animal visited some parts only rarely; the amount of total range used each month seldom exceeded 25%. Ranges were stable and shared all or in part with other pandas. Land tenure appeared to be different in males and females: males occupied greatly overlapping ranges, whereas each female spent most of her time in a discrete core area of only 30-40 ha.

We lack detailed data on land tenure in Tangjiahe pandas. Xue confined her activities to 1.3 km2 from mid-December to March (Figure 8.2),and within this small area she moved little(Table 8.3). However, that summer she shifted at least 3 km southeast to some high ridges(D.Reid, pers.comm.).

Tang had a range of at least 23.1 km2, an unusually large area. However, within that range he had a center of activity, 1.1 km2 large, in which he remained for nine months from September to June(Figure 8.2),using generally less than 1% of his total range each month(Table 8.3).And within his center of activity he was usually found in one of four bamboo patches whose total area comprised only about 5 ha; between September and March he was radio-located on 134 of 182 days(74%) in these patches. His sedentary habits during the mating season from mid-March to early May surprised us, for no other pandas were in his center of activity, to our knowledge. Tang traveled widely during summer. In June, for example, he moved over 3 km down-valley and over a ridge to the east, where we lost contact with him; he had returned by July and remained south of his usual haunts, for a while outside of the reserve. Although some Wolong pandas shifted seasonally to lower elevations to feed on bamboo shoots , they remained within their small ranges ,showing a pattern quite different from that of Tang and apparently also Xue. Tang’s shift away from B.fargesii occurred at a time when that bamboo seemed to become less palatable. The unusual extent of Tang’s movements cannot be explained solely on the basis of a food search, although much travel was necessary to reach the patchy bamboo remnants on the ridges.


Asiatic Black Bear

Bears were not abundant in Tangjiahe, though the many branches they broke while feeding in trees were conspicuous reminders of their presence. Only an estimated 10-12 bears frequented the slopes of the Beilu valley up-river from Maoxiangba , and none seemed to remain there permanently.


Food Habits

The feeding of Tangjiahe bears falls into three periods: from April to mid-July the bears eat mainly herbs and leaves from shrubs, from mid-July to mid-September they add fruits to their diet as soon as these ripen, and from mid-September to November they harvest primarily acorns and other nuts.

Spring provides bears with a variety of succulent herb species. Of 15 spring foods listed in Table8.4, bears particularly favored the thick stalks of wild parsnip (Anthriscus sylvestris, Heracleum sp., Angelica sp.), Petasites tricholobus leaves, new Hydrangea growth, and Rubus coreanus shoots. Bears seldom lingered in a valley at this season, ignoring all but a few plant species in their travels. On 22 May one bear stopped to feed 18 times while moving 1 km along the base of a slope .It plucked some leaves and branch tops from one spiny Acanthopanax henryi and two Hydrangea shrubs; and it consumed jack-in –the-pulpit (Arisaema lobatum) at one site, Heracleum, species at a second, and Angelica species at a third, each time removing the leafy tops before eating the stalks. However, this bear seemed to prefer R.coreanus shoots about 1 cm thick, which sprout much like bamboo shoots. The bear broke or bit off shoots, removed the leafy tassel at the top, and ate the juicy stalk ,sometimes after peeling off the densely haired skin. Another bear angled down to the base of a long slope and foraged in a patch of Heracleum plants before continuing across the valley and up the other side, with only a brief halt to eat one Arisaema lobatum stalk , a total of two feeding stops in 1.5 km.

Herbs remained an important food all summer, sometimes augmented with bamboo(F.scabrida)shoots in August(D.Reid, pers, comm.).From mid-July to late August R.coreanus berries were frequently eaten(Table 8.5). The wild cherry (Prunus sp.) crop failed in 1984 ; judging by broken branches, bears had harvested fruits in previous years.

Bears had left the herb and berry patches of the upper valleys by mid-September and concentrated below 1800 m in oak stands on ridges and slopes. Acorns (Quercus sp.) supplemented by hazelnuts (Corylus sp.), butternuts (Juglans cathayensis), and fruits (Celtis biondii, Actinidia chinensis) composed the diet until hibernation. On 24 October, at 1615 h, we watched a bear feeding in an oak. Squatting or standing in a fork, the bear pulled small branches toward itself with a forepaw, sometimes breaking one with a bite, and plucked the acorns directly with its mouth. Large branches required more effort. The bear pulled them with both forepaws, occasionally using mouth as well, until they broke or snapped off. The animal usually pushed discarded branches into the fork beneath its feet and stepped on them , creating a crude platform(Figure 8.10). Bears in India harvest the cherry-like Celtis fruits in a similar manner (Schaller 1969).

Almost every oak in Tangjiahe has broken branches, some so damaged that little beyond the tree trunk and branch stumps remain. Bears have seriously reduced the acorn supply of future years. By late October most acorns have fallen, and bears search for them beneath trees.

Bears no doubt ate more than the 28 species listed in Table 8.4. Grapes (Vitis sp.) and rose tips (Rosa sp.) are eaten in India (Schaller 1977), beech buds in Japan (Hazumi 1985), and Pinus seeds , Ribes berries, Carex and Lilium leaves and stalks in Russia (Bromlei 1973). These genera are also present in Tangjiahe. Wu(1993) listed nine genera of fruits and nuts as bear food in the Qinling Mountains of Shaanxi province, 300 km northeast of our study area. Among these were Fragaria, Schisandra, Coriaria, Flaeagnus, Rbus and Castanea, most , if not all , of which occur in Tangjiahe too.

Domesticated plants were also consumed in Tangjiahe. In late August and September bears took apples from orchards in the reserve and maize and walnuts from farms.

Meat, either killed or scavenged, forms a small percentage of a bear’s total diet, and may include mammals, birds, fish, mollusks, and insects (Schaller 1969; Bromlei 1973; Wu1983). One dropping in Tangjiahe contained takin (Budorcas taxicolor) hair , and one in nearby Wanglang Reserve an infant bamboo rat(Rhizomys sinense).

Bears in Tangjiahe and northeastern Russia have similar food habits and show similar seasonal changes in selecting herbs, fruits, and nuts. Bromlei (1973) listed 27 genera of wild food plants for bears in Russia, and ten of these were also eaten in Tangjiahe.


Nutritional Content of Plants

The indigestible cellulose and lignin values were moderately high (mean 29%) in the various herbs, whereas the partially digestible hemicelluloses were low(6%).Ash content was high(16%).Protein varied from 2.6% in Angelica species to 35.2% in Hydrangea species, with an average of 17% for all herbs tested. The remaining cell contents(sugars, lipids)comprised a mean of 31%(Table 8.2). Rubus and Actinidia fruits were chemically similar to the other plants except that they were low in ash (Figure 8.11). Bears shell acorns with their mouths and eat only the kernels; our nutritional analyses were conducted on kernels only. The two acorn species differed in nutrient content from herbs and berries: they were much lower in cellulose and lignin(mean 6%)and moderately high in hemicelluloses (38%). Ash was low(3%). Protein was low as well (5%), whereas the remaining cell contents were higher(49%). Cell contents of acorns, as that of many seeds, contain much fat, 9% in Quercus ilex from northern Pakistan (Schaller 1977).Acorns thus represent a high-calorie, digestible plant food for bears.

Silica content of foods was low, averaging 0.2% for leaves and fruits and none in acorns.

The essential amino acid content of 11 spring food plants was analyzed. Only methionine showed a conspicuously low level(Table 8.6). This amino acid is usually deficient in plants, including bamboo (Schaller et al. 1985).

Herbs had a mean water content of 90% and Rubus berries of 86% ; droppings from April to mid-September were soft, with a water content of 87%. During these months bears probably had no need to drink. The water balance changed in autumn. Acorns contained 52% water and droppings 73%-figures similar to those of pandas foraging on bamboo leaves and stems.



Kui’s activity was monitored for 3 days in July, and Chong’s for 6 days in November, the month prior to his hibernation. The two bears showed similar daily cycles(Figure 8.12):they became inactive in the evening , Chong at 1900h and Kui at 2100h , and spent much of the night at rest. Kui’s probability of activity was similar to Chong’s(0.48 vs. 0.46)even though their diet differed, the former foraging for herbs and fruit and the latter for acorns.

Bears usually rested on the ground , in the manner of pandas, but occasionally an animal built a bed. Three such beds, constructed of bamboo, were observed in Wolong . In Tangjiahe a bear climbed a spur in late afternoon to a small, level spot. There it bent in a total of 51 bamboo stems and 13 beech saplings up to 3 cm thick and tucked them around and under its body; it either broke long stems to fit the rim, or , in the case of nine bamboos and 12 saplings, bit off the tops. The result of the effort was a springy, circular bed 130 cm in diameter(Figure 8.13).A total of 5.5 kg of droppings, consisting solely of herbs, were just outside the rim.

Bears in Tangjiahe , as in other temperate climates (Bromlei 1973; Maita 1985), hibernate during winter. Chong traveled on 28 November, and he was active for part of the day, apparently in and around a den, from 29 November to 6 December. He became almost inactive on December 8;24-h monitoring on that date revealed a probability of activity of 0.13. On 13 December we monitored his signal from 0915 to 1500h without recording any activity. On 26 March his signal indicated movement within the den-a rock cleft-and he emerged during the first week of April. In Wolong we found a hibernating yearling in the hollow base of a fir at 2600 m as 12 April. Late March to mid-April is the general emergence period in the region, according to local people, indicating a hibernation period of about four months.


Land Tenure

During the month before hibernation Chong meandered in search of acorns and then entered a den for four months, all within an area of 6 km2. Kui roamed widely and often rapidly, covering at least 29 km2 between July and September, excluding one or more trips outside the reserve to the south, where we did not track him(Figure 8.2);in October and November we could not find him. In Japan the annual range of a subadult female was 6km2 and the autume range of an adult male was 26 km2 (Maita 1985).



Giant pandas and Asiatic black bears showed moderate ecological overlap in activity schedules and use of space. Both were active in daytime; at night the bears tended to rest and pandas to feed intermittently. Ranges of the two species overlapped. However, animals were spatially separated in autumn, bears usually foraging below 1800 m and pandas above that elevation. And in winter the bears hibernated. There was almost no overlap in food habits. A pandas occasionally sampled Heracleum species, a major bear food, and bears ate bamboo shoots. No competition existed for these resources, since pandas seldom ate herbs and bamboo shoots were seasonally abundant, providing ample forage for both species.

The panda has an extraordinarily limited diet, usually only one or two kinds of bamboo. But it has specialized on a plant resource that is available in unlimited amounts at all seasons. Furthermore, the nutritional quality of bamboo remains fairly constant throughout the year, an unusual situation for a plant growing in an environment with marked seasonal changes. Forage in temperate climates usually declines in nutritive value during winter, and in arid climates during the dry season. This lack of fluctuation in both amount and quality has enabled the panda to subsist on bamboo in spite of a low nutritive content.

Unable to digest cellulose and lignin in the cell wall, a panda must obtain nutrients from cell contents and from the fraction (av. 22%) of the hemicelluloses it can break down. Total dry matter digestibility is only 17%, calculated from Wolong data , as compared with at least 60% from an ungulate eating green grass(Van Soest 1982).Such low digestibility implies that the panda must quickly move much bulk through its digestive tract to obtain needed nutrients. Indeed, passage time of bamboo is less than 14 h (Dierenfeld 1981) and as little as 5 h on a diet of bamboo shoots. The panda also needs to harvest bamboo selectively for the most nutritious parts. Stepwise discriminant analysis of bamboo parts eaten and not eaten in Wolong showed that cellulose was the best single variable for discriminating favored from other parts. Bamboo leaves are higher in protein, minerals, and hemicelluloses and lower in cellulose and lignin than are stems. They can also be collected and eaten quickly .One would, therefore, expect pandas to favor leaves. Animals feeding of F.scabrida selected leaves over stems(57% to 43%) on an annual basis. Some stems were eaten at all seasons, possibly because the amount of soluble carbohydrates in the cell contents is similar in leaves and stems; for unknown reasons , leaves seemed to be avoided from April to June. Shoots were an almost exclusive feed in season ,even though they rank nutritionally lower than leaves. hemicellulose digestibility in young, growing tissue is higher than in mature tissue, and pandas perhaps select for the additional energy provided by these hemicelluloses.

Wolong pandas subsisting on Sinarundinaria fangiana bamboo obtained an estimated 4354 kcal/day of digestible energy in spring, 5488 kcal/day in summer and autumn, and 5542 kcal/day in winter. On one occasion in October, Tang foraged in a small B.fargesii patch for 2.5 days, permitting us to calculate his caloric intake. He defecated an average of 12.9 kg(wet weight)or 3.3 kg(dry weight) of droppings per day, the droppings composed wholly of leaves. His daily dry matter intake was, therefore, 3.3(100)/100-23.25(where 23.25% is the autumn dry matter digestibility in Wolong), or 4.3 ke/day. Converted to fresh weight of leaves eaten, the figure is 4.3/100-45.5(where 45.5% is water content of leaves), or 7.9 kg/day. Dierenfeld et al. (1982) found that gross energy of Phyllostachys species leaf was 4800 kcal/kg, and that of B.fargesii leaf is probably similar. On a dry matter intake of 4.3 kg, Tang ate 20,640 kcal, but with a 23.25% digestibility he obtained only 4799 kcal/day of digestible energy. This figure is somewhat lower than the autumn one calculated for Wolong, if our small weight sample is characteristic. At 68 kg Tang has , for example , an energy expenditure for basal metabolism of 1658 kcal/day, as compared to 2214 kcal/day for a 100-kg Wolong male.

The total average energy expenditure of a 100-kg Wolong panda was calculated at approximately 3132 kcal/day, a figure that does not include energy invested in growth and reproduction. With an intake of only approximately 4300-5500 kcal/day, the nutritional margin of safety is fairly small. Although pandas have reduced energy expenditures to a minimum by resting much and , when active, concentrating on feeding, they are still constrained by low food quality and limited in food intake by the capacity of the digestive tract. Consequently, a panda can obtain only enough digestible energy to store a small amount of fat. But with a stable food source of constant quality and no hibernation period, a panda has no need for large fat deposits.

Conditions for the animals change, however, when at long intervals (40-80 or more years, depending on species)bamboo in an area flower and dies synchronously. Pandas then are forced to forage in patchy remnants or switch to other bamboo species at different elevations for 5-10 years until new seed-lings have grown tall enough to provide food. If alternative bamboo species are not available, the animals may starve, This happened in parts of the Min Mountains during the mid-1970s when at least three species flowered, and in certain areas of the Qionglai Mountains in 1983 when S. fangiana flowered.

Asiatic black bears differ markedly from pandas in their feeding strategy. Though selective in the plants they eat, they choose a moderately large variety-over 28 species in Tangjiahe. Emphasis on type of food changes with seasonal availability: herbs in spring, fruit in summer as soon as available , and acorns in autumn. The chemical compositions of these feeds are different from those of bamboo. Herbs average three times more cell contents than bamboo, and they are low in hemicelluloses(Table 8.2). Rubus berries are similar to herbs in nutritive levels(Figure 8.11) but presumably have higher concentrations of soluble carbohydrates, thus making it nutritionally beneficial for bears to pluck these small items. Bunnell and Hamilton(1985) found a 88.4% dry matter digestibility of blueberries in brown bear (Ursus arctos), and, judging by constituents (Figure 8.11), Rubus berries probably have a digestibility of 50-60%. Herbs and berries present bears with nutritional problems also faced by pandas when eating bamboo shoots. Succulent foods have a low retention time, and the animal must extract the maximum amount of energy within a few hours. Furthermore, it must eat much bulk to obtain few nutrients because of the high water content.

Acorns contain less cellulose and lignin than any other food analyzed; most of the kernel provides digestible energy(Figure 8.11).One method of measuring dry matter digestibility is by using an internal marker to compare concentrations of an indigestible constituent in both food and feces. Lignin is such a marker in mature plants. Lignin in two acorn samples averaged 2.55% and in three droppings 8.5%, indicating a digestibility of 70.2%. Acorns require time to collect. And about 30% of an acorn consists of shell , which bears discard, leaving kernels averaging 0.4 to 1.3 g each, depending on species. To eat 5 kg, a bear would require about 3850 large acorns or 12,500 small ones. During its 11-h daily active period in November, a bear would have to gather 5.8 to 18.9 acorns per minute to reach these numbers, a feasible effort. Time constraint is probably the reason why bears rarely eat the tiny Cyclobalanopsis oxyodon acorns.

Gross energy of bear foods is similar to that of bamboo (4400-4800 kcal/kg), with Rubus coreanus shoots 4200 kcal/kg, R.coreanus berries 5600, and acorns 4400. Unfortunately, we lack data on daily food intake of bears. Nelson (1980) calculated that the American black bear (Ursus americanus) may assimilate as many as 20,000 kcal in a day .Most aspects of bear digestion-passage rates, efficiency, and so forth-appear to be similar to the panda’s .However, A bear foods generally contain at least three times more digestible energy than does bamboo. Thus although the gross energies of feeds are much alike, bears obtain at least three times more digestible energy per kilogram eaten than do pandas. A bear’s resources are patchy and its food items small ,requiring extra energy for searching and feeding, but the animals still obtain such a daily surplus of calories that they can store enough fat for hibernation.

With digestive capabilities of the two species so similar, the question still remains why bears do not eat bamboo leaves and especially why pandas only sample some nutritious bear foods on occasion. The theory of optimum foraging predicts that an animal should attempt to obtain the best balance and amount of nutrients for the least amount of time and effort (Pianka 1978).Yet we have noted pandas walking through stands of parsnip without halting in this favored bear food, only to forage on low-quality bamboo nearby. For unknown reasons ,pandas do not always select forage of the highest available quality even though they seem to have retained the ability to assimilate it ,as well as to store fat. Meat , with a gross energy digestibility of about 90% in bears (Bunnell and Hamilton 1985), is, however, eaten readily by both species. Both have conserved their digestive adaptations for carnivory in spite of some different morphological and possibly physiological adaptation for herbivory.

The panda's emphasis on bamboo, even when other forage is available, is surprising, for a low-quality diet has affected many facets of the animal's life. With a diet that provides little more than subsistence, a panda must keep its digestive tract filled by foraging at frequent intervals both day and night. By contrast, bears rest much of the night. In spite of abundant and concentrated resources, a panda's time budget resembles that of a non-ruminant herbivore, with 50% or more of the day usually spent foraging. The bears Chong and Kui required only 46-48% of the day to fulfill energy needs, store fat, and then hibernate four months. Pandas can subsist for months within 1 km2 , whereas bears require large ranges and much travel - Kui used over 29 km2 to forage on their seasonal, patchy, and small food resources. During November, for example, Tang remained in 1 ha while Chong roamed over 600 ha. Yet their total activity levels were alike.

A panda's need to conserve energy extends to reproduction. Females do not conceive until 5.5 or 6.5 years of age, implantation is delayed one and a half to four months, and, after a true gestation period of about one and a half to two months, one or two young are born in such altricial state that they are proportionately the smallest of eutherian mammals, about 1/900 the weight of the mother. If there are twins, one young soon dies, the female raising just one off spring to an early independence at one and a half years. The American black bear is the ecological counterpart of the Asiatic one, judging by similar body size, habitat, mating activity, range size and other aspects of behavior (Jonkel and Cowan 1971; Garshelis and Pelton 1980, 1981; Johnson and Pelton 1980; Pelton 1982).For example, it consumes primarily grasses and herbs in spring, fruits in summer, and a mixture of fruits and acorns and other hard mast in autumn (Pelton 1982).With high-quality food available in summer and autumn, the mean monthly probability of activity of bears in the Great Smoky Mountains of Tennessee - a habitat similar to that of Tangjiahe - reaches a high level (0.5-0.6) only from June to October, and animals hibernate from late December to early April (Quigley 1982).The bears are primarily diurnal, with daily activity peaks around dawn and dusk (Garshelis and Pelton 1980; Pelton 1982). American black bears may reproduce as early as 2.5 or 3.5 years of age in areas with a good nutritional base (Alt et al. 1980), newborns are about 1/200-1/300 the weight of their mother, and litters average 1.7-2.9 young, depending on area, a female often raising all her offspring. A high-quality diet gives the American black bear, and no doubt also the Asiatic black bear, greater reproductive flexibility and potential than the panda can achieve, at less expense in energy.

Although a difference in feeding ecology influences various types of behavior in bear and panda, it apparently has not had an impact on the basic land tenure system, if data from studies of American black bears are valid for comparison. Adults of both species share ranges, overlap between females’ ranges is less than between males’, and a female has a more or less discrete core area in which other females are not tolerated(Rogers 1977).The ranges of male bears are generally two to three times larger than those of females. For example, in Tennessee males’ ranges averaged 42 km2 and females’ 15 km2 (Garshelis and Pelton 1981); in Pennsylvania males’ ranges averaged 173 km2 and females’ 41 km2 (Alt et al.1980); and in Idaho males’ ranges averaged 112 km2 and females’ 50 km2 (Amstrup and Beecham 1976). Both sexes of pandas have ranges of approximately similar size, at least in Wolong.

In conclusion, differences in feeding strategy have had a marked impact on the lifestyles of the Asiatic black bear and the panda. The bear has opted for a nutritional boom-or-bust economy: it stuffs itself on a variety of high-energy seasonal foods, storing excess calories as fat ,and then hibernates during times of scarcity. Opportunistic and adaptable, Asiatic black bears are widely distributed.: "The panda, by contrast, has become a specialist; dependent on a low-quality but constant and abundant food source, it has chosen security over uncertainty. Its mode of life gives the impression of being a durable triumph of evolution. But by losing the sense of struggle, its curiosity, its need to explore and be observant and try something new, by tying itself to a fate without horizon, it has become defenseless, it has lost the adaptability that it now must have to survive’(Schaller et al. 1985:224).



This project was financed through an agreement between China’s Ministry of Forestry and the World Wildlife Fund. Wildlife Conservation International, a division of the New York Zoological Society, also provided financial support .We received guidance and encouragement from many officials, including Wang Menghu(Ministry of Forest, Beijing), Hu Tieqin, Fu Chengjun, Gung Tongyang, and Bi Fengzhou (Forestry Bureau, Chengdu), and Yue Zhishun and Jiang Mingdao (Tangjiahe Natural Reserve) .A.Taylor (University of Colorado) assisted with the bamboo work and proviede us with data,Qin Zisheng (Nanchong Normal College) identified the plants, D.Reid (University of Calgary) supplied some recent information, and K.Schaller, Qiu Mingjiang, and Wang Fulin assisted with field work. Nutritional analyses were done by J.Robertson, Department of Animal Sciences, Cornell University, and the amino acid determinations by Pan Wenshi, Department of Zoology, Beijing University. The Department of Forestry, Wildlife and Fisheries of the University of Tennessee provided computer facilities through the courtesy of M.Pelton. A.Taylor and E.Dierenfeld read the manuscript critically. We are deeply grateful to these individuals and institutions for this assistance.



Alt, G., Matula, G., Alt, F., and Lindzey, J.1980. Dynamics of home range and movements of adult black bears in northeastern Pennsylvania. In: C.Martinka&K. McArthur, eds. Bears-Their Biology and Management , pp.131-136.Washington, D.C.:Government Printing Office.

Amstrup, S., and Beecham, J.1976.Activity patterns of radio-collared black bears in Idaho.J.Wildl.Mgmt. 33:340-348.

Bromlei, G.1973.Bear of the South Far-eastern USSR.New Delhi:Indian National Scientific Documentation Centre.

Bunnell, F., and Hamilton, T.1985. Forage digestibility and fitness in grizzly bears. Internat.Conf.Bear Res.and Mgmt.5:179-185.

Chu Ching and Long Zhi.1983.[The vicissitudes of the giant panda.] Acta Zool. Sinica.29:93-104.(In Chinese.)

Davis, D.1964.The giant panda:A morphological study of evolutionary mechanisms. Fieldiana.Zoology Mem.3:1-339.

Dierenfeld, E.1981.The nutritional composition of bamboo and its utilization by the giant panda.M.Sc.thesis, Cornell Univ., Ithaca,N.Y.

Dierenfeld, E., Hintz, H., Robertson, J., Van Soest, P., and Oftedal, O. 1982. Utilization of bamboo by the giant panda. J.Nutr. 112:636-641.

Garshelis, D., and Pelton, M.1980. Activity of black bears in the Great Smoky Mountains National Park.J. Mamm.61:8-19.

Garshelis, D., and Pelton, M.1981.Movements of black bears in the Great Smoky Mountains National Park.J. Wildl.Mgmt.45:912-925.

Hazumi, Toshihiro.1985.Habitat selection of Japanese black bears in Nikko. Abstracts distributed at First Asiatic Bear Conference, Utsunomiya and Nikko, Tochigi Prefecture, Japan.

Hu Jinchu. 1981. [Ecology and biology of the giant panda, golden monkey, and takin.]Chengdu:Sichuan People’s Publishing House.(In Chines.)

Johnson , K., and Pelton,M.1980. Environmental relationships and the denning period of black bears in Tennessee.J.Mamm.61:653-660.

Jonkel, C., and Cowan, I.1971. The black bear in the spruce-fir forest. Wildl.Monogr.27:1-57.

Ma Yiching.1983.The status of bears in China . Acta Zool.Fennica. 174:165-166.Maita,Kazuhiko.1985. Movements of Japanese black bears (Selenarctos thibetanus japonicus) in Taiheizan Mountain region, Akita Prefecture. Abstracts distributed at First Asiatic Bear Conference, Usunomiya and Nikko, Tochigi Prefecture, Japan.

Nelson, R.1980.Protein and fat metabolism in hibernating bears. Amer.Physiol.soc.Fed.Proc.39:2955-2958.

Pelton, M.1982.Black bear, Ursus americanus.In:J. Chapman & G.Feldhamer, eds. Wild Mammals of North America, pp.504-514.Baltimore: Johns Hopkins Univ.Press.

Pianka, E.1978. Evolutionary Ecology. New York:Harper & Row.

Quigley , H.1982.Activity patterns, movement ecology, and habitat utilization of black bears in the Great Smoky Mountains National Park, Tennessee.M.Sc.thesis,Univ.Tennessee, Knoxville.140 pp.

Rogers, L.1977.Social relationships, movements, and population dynamics of black bears in northeastern Minnesota. Ph.D.dissert., Univ. Minnesota, St.Paul.194 pp.

Ruan Shiju and Yong Yange. 1983.[Observations on feeding and search for food of giant panda in the wild. ]Wildlife 1:5-8.(In Chinese.)

Schaller, G.1969.Food habits of the Himalay black bear (Selenarctos thibetanus)in the Dachigam Sanctuary, Kashmir.J.Bombay Nat.Hist. Soc. 66:156-159.

Schaller, G.1977.Mountain Monarchs: Wild sheep and Goats of the Himalaya.Chicago: Univ.Chicago Press.

Schaller, G., Hu Jinchu, Pan Wenshi and Zhu Jing.1985. The Giant Pandas of Wolong .Chicago:Univ.Chicago Press.

Van Soest, P.1982.Nutritional Ecology of the Ruminant.Corvallis, Ore.:O $ B Books.

Wu Jiayian.1983.[Preliminary observations on the food specialization of black bear from Qinling. ]Chinese J.Zool.4:47-51.(In Chinese.)

Yi Tongpei.1985.Classification and distribution of the food bamboos of the giant panda.J.Bamboo Res.4:11-27;(2):20-45.(In Chinese.)


Table 8.1 Destruction of new bamboo shoots by predators in Wolong and Tangjiahe


Wolong:Fargesia spathacea




n=996 shoots


n=724 shoots


n=209 shoots


Giant pandas

Rodents and Ochotona sp.



















Table 8.2. Chemical composition of bamboo leaves eaten by pandas and herbs eaten by bears








and lignin





(% of dry matter)

Bamboo leafes







Mean of 12

































Table 8.3. Monthly variations in known home range used by two pandas




September 1984




January 1985

















Note. The total range for Xue, a female, was 1.3 km2; for Tang, a male, 23.1 km2.


Table 8.4. Wild food plants eaten by black bears in the Tangjiahe Reserve


Part eaten

Acanthopanax henryi

Actinidia chinensis

Angelica sp.

Anthriscus sylvestris

Arisaema lobatum

Aster ageratoides

Cacalia tangutica

Caraya sp.

Celtis biondii

Cnidium sp.

Cornus chinensis

Corylus sp.

Cyclobalanopsis oxyodon

Fargesia scabrida

Heracleum moellendorffii


Hydrangea sp.

Juglans cathayensis

Lunathyrium giraldii

Petasites tricholobus

Phlomis sp.

Prunus brachypoda


Quercus aliena



Rubus coreanus

Salvia umbratica

New stem growth,leaf


Succulent stalk

Succulent stalk

Succulent stalk





Succulent stalk





Succulent stalk

Succulent stalk

New stem growth, leaf


Young frond








Fruit, new shoot



Table 8.5. Seasonal variation in the diet of black bears, as estimated from contents of droppings




Type of food(%)


And stalks





Acorns and

other nuts

April to mid-July

Mid-July to mid-September

Mid-September so November

















Table 8.6. Essential amino acid content (in mg/100 g) of

11 spring plants eaten by black bears.

Amino acid

















Note. Tryptophan not tested. The plants tested were species of Acanthopanax, Anthriscus, Arisaema, Aster, Cacalia, Cnidium, Hydrangea, Lunathyrium, Petasites, Phlomis, and Heracleum scabridum.

spirituality : design : water : fashion
amphibia : reptilia : butterflies : panda : takin
ecotourism : Xishuangbanna : Shangrila : Yunnan : Mekong
books : Buddhist flowers : consulting : science of ecology : medicine
bananas : ginger : herbs : orchids : trees : succulents : lianas : sacred trees
Gabrielle's traditional motive Chinese jewelry : Wu Jialin's black and white photography
rain forests : river banks of the Mekong : pools : sacred forests : sinter terraces : chili growing
AIDS/HIV : asthma : acne : herpes : hepatitis : cancer : diabetes : health tea : malaria : Puer tea : orchid tea

is the Thai internet portal for biodiversity products designed to protect endangered species in cooperation with local ethnic communities
a project of the
Xishuangbanna, Yunnan, China

© 2000 - today