 |
resources forexplore exploreresources for |
Mist enveloped us as we meandered through fallow winter fields of maize and beans tucked in the narrow Pitae He valley in southwestern China. At the edge of the fields the path changed direction and angled upslope floating in the fog that hung on the surrounding peaks. It was a classic scene in a Chinese painting. We climbed a steep pitch for another hour through shrubs and saplings and images of the surrounding mountains moved in and out of focus in the swirling fog. The mountains were vertical and spectacular, but intimidating and they looked impossible for field work. The mist turned to snow. The path suddenly leveled off as we traversed a spur and we entered the Choushuigou Valley clothed in spruce forest with bamboo in the understory. This was our destination, the forests pandas called home. We walked another ten minutes on a narrow icy trail and waited at our camp, Wuyipeng, for the half dozen Tibetan porters who were carrying the research team's equipment and supplies. Wuyipeng was a cluster of canvas tents perched at 2500 m in the heart of the Wolong Natural Reserve in Sichuan Province (map 1). I had no idea then, that I would spend almost three years during the next decade doing research in these mountains on the bamboo forests which pandas need to survive.
Map 1: Panda Distribution, Past and Present
Our research team was international and interdisciplinary and we had been invited by the Chinese government and World Wide Fund For Nature to study the endangered giant panda and its habitat. Both were virtually unknown. The aim of our project was to develop a sound scientific foundation for future panda conservation initiatives. Our group included zoologists, botanists, and geographers from China, the United States, Great Britain, Canada and Japan. My research, and that of my colleague Qin Zisheng, focused on two broad questions: 1) what controls the distribution and abundance of bamboos in forests pandas use and 2) have people changed panda habitat and if so are these changes detrimental to pandas? Other members of the research team focused their efforts on identifying the food habits, movement and activity patterns, and the population dynamics and social behavior of the pandas. Boundaries between our lines of research were fuzzy. We hoped to get an integrated picture of the panda's life.
The plight facing the giant panda surfaced in the mid 1970s. In 1975-76, in the remote Min Mountains, 138 pandas were found dead of starvation [Schaller et al. 1985] (map 1). There are about 1500 wild pandas and the death of so many caused widespread concern among government officials and conservationists [Schaller et al. 1985; O'Brien & Knight 1987]. The pandas starved because they suddenly ran out of their only food, bamboo. At least three bamboo species in the Min Mountains flowered synchronously and then died. Pandas have obviously survived such bamboo die-offs in the past. Why did so many pandas die after the bamboo die-off in the mid 1970s? High panda mortality was caused by a combination of factors: habitat destruction, the pandas specialized food habits, and bamboos peculiar lifecycle [Schaller et al. 1985; Taylor and Qin 1987].
Pandas once ranged over much of eastern China (map 1). But now they occur only in six forested tracts that cover about 30,000 km. along the eastern rim of the Tibetan Plateau in parts of Sichuan, Gansu, and Shannxi provinces in southwestern China. The contraction of the panda's range is partially the result of climatic change during the Pleistocene period but mostly due to landscape change caused by people [Schaller et al. 1985]. Forests with bamboo have been cleared for agriculture for centuries to feed China's burgeoning population [Campbell 1984a, 1986; Schaller et al. 1985]. Recently, loggers have moved into remote mountain areas to clearcut timber China is starved for [DeWulf et al. 1988]. The once large and more continuous distribution of pandas is now smaller and fragmented. Fragmentation of panda populations into small groups may be a threat to pandas because of inbreeding depression. Inbreeding reduces reproduction, fecundity, and survival of young [Schaller et al. 1985; O'Brien & Knight 1987].
Panda habitat: background, primary coniferous forests with bamboo understory; foreground, area clearcut in 1970s.
Panda's are absolutely fettered by bamboo. It comprises 99% of their diet [Schaller et al. 1985; Schaller et al. 1989 Johnson et al. 1988a,b; Reid et al. 1989; Reid and Hu 1991]. A vegetarian diet is typical of herbivores who digest both the content of plant cells and much of the tough cell wall. In an herbivores stomach, the tough cell wall is digested by microbes after plant parts ferment in their complex stomach. But the panda has the stomach of a carnivore, with a short intestine, so it can digest only cell content and a small portion of cell walls [Dierenfeld et al. 1982; Schaller et al. 1985]. The nutritive content of bamboo is low but the supply and nutrient quality of bamboo is relatively constant year round [Dierenfeld et al. 1982; Schaller et al. 1985]. So the gastrointestinally ill-equipped panda digests only 12-23% of the bamboo it eats [Dierenfeld et al. 1982] and it must spend much of the day feeding (57%) to meet its nutritional requirements [Schaller et al. 1985]. Our team discovered that pandas eat 14 kg or more of bamboo a day but despite these gargantuan diets, they barely consume enough forage to meet their daily energetic needs. Pandas are vulnerable when there are large fluctuations in their food supply due to bamboo die-offs.
Professors Qin Zisheng and Alan Taylor in matrue arrow bamboo stand.
Bamboo die-off's are natural. The twenty or so bamboo species eaten by pandas grow and spread vegetatively most years by producing finger-width shoots from extensive root systems [Campbell and Qin 1984b; Schaller et al. 1985; Taylor and Qin 1987, 1993a). But periodically, each bamboo species flowers, sets seed, and then dies back. The next bamboo generation emerges from seed that germinates on the mossy forest floor [Taylor and Qin 1988a, 1993b]. The habit of gregarious flowering followed by die-off, in bamboos, is unusual in the plant kingdom [Janzen 1976; Simmonds 1980].
Seven-year-old arrow bamboo seedlings.
We don't know why bamboos mass-flower and then die-back but several evolutionary mechanisms that would promote mass-flowering have been suggested. Mass-flowering in bamboos may have evolved to satiate bamboo seed eaters such as rodents and birds [Janzen 1976]. If bamboos produced all their seed at once, then seed™eaters would consume all the seed they want and some seed would be left over to start the next generation. The probability of a seed surviving seed-eaters would be greatest when most seeds are produced. Over time, pressure by bamboo seed-eaters would promote the development of synchronous flowering and seed production [Janzen 1976]. But bamboos don't have to die after they produce seed. Why do they die-back? We don't know, but bamboo seedlings may need their parents to die to establish and grow. Mature bamboos cast deep shade and few plants survive beneath a dense bamboo canopy. If mature bamboos did not die back, light levels would probably be too low for bamboo seedling growth [Simmonds 1980; Taylor et al. 1991a]. We still don't know why bamboos wait so long to flower and then do so all at once. But we now know something about how pandas respond to a bamboo die-off.
Radio-collared panda Zhen-Zhen eating arrow bamboo.
In 1983, arrow bamboo (Bashania fangiana) flowered throughout Wolong where our research team was working. There was immediate concern that pandas would starve as they did in the Min Mountains in the mid 1970s. A few pandas did starve [Johnson et al. 1988a], but we found that most pandas just changed their eating habits. Arrow bamboo carpets the understory of the subalpine fir (Abies faxoniana) and birch (Betula utilis) forests [Campbell and Qin 1984b; Taylor and Qin 1988b,c, 1992] that pandas spent most of the time in. These bamboo forests were their favorite [Schaller et al. 1985; Johnson et al. 1988a,b; Reid et al. 1989; Reid and Hu 1991]. Yet there were other bamboos that pandas sometimes ate, notably umbrella bamboo (Fargesia robusta) especially during the spring shoot season [Campbell and Qin 1984b; Schaller et al. 1985; Taylor and Qin 1987]. About 85% of the arrow bamboo died after it flowered in 1983 [Reid et al. 1989]. The remaining live arrow bamboo (15%) was scattered in small patches [Johnson et al. 1988a; Reid et al. 1989; Taylor et al. 1991a]. So how did pandas respond to the change in their food supply? Our research team discovered that pandas became less selective in their choice of arrow bamboo plants parts, and they began to eat more umbrella bamboo at lower elevations [Johnson et al. 1988a; Reid et al. 1989; Reid and Hu 1991].
Primary forest in Wolong Panda Reserve.
Our understanding of how pandas adjust to changes in their food supply pointed to a clear goal for panda conservation: maintain two or more bamboo species within the range of groups of pandas so pandas don't starve when a bamboo flowers and dies. The arrow bamboo dieback in 1983 did not cause widespread starvation because umbrella bamboo was available to the group of pandas we were studying. In the Min mountains in 1975 in contrast, several bamboos flowered simultaneously and others had been removed and replaced by fields [Schaller et al. 1985; Johnson et al. 1988a] so pandas starved.
Recovery of bamboo groves after flowering is a gradual process. We found that bamboo seedlings take about 15-20 years to grow back to full size but pandas can eat seedlings once they are ten years old or a meter or so tall [Taylor 1993b]. We also found that bamboo seedlings grow best beneath small opening in the forest canopy where light penetrates to the forest floor [Taylor and Qin, 1992]. But large openings in a forest are poor places for bamboo seedlings establishment and growth. Large tracts of mixed fir, hemlock (Tsuga dumosa), and birch forest with arrow bamboo were clearcut in Wolong, and elsewhere, in the panda range in the early 1970s. Arrow bamboo proliferated in these clearcuts forming dense bamboo thickets [Schaller et al. 1985; Taylor and Qin 1987, 1988a; Reid et al., 1991]. But we found few trees, especially fir or hemlock, regenerating in these bamboo thickets so forest recovery will be slow [Schaller et al. 1985; Taylor et al. 1991b]. Despite the dense bamboo in clearcuts, we discovered that pandas avoided them except for short periods in the summer when they feed mostly on bamboo leaves [Schaller et al. 1985; Johnson et al. 1988; Reid et al. 1989]. Arrow bamboo died back in the clearcuts in 1983 but I found that few bamboo seedlings established in the clearcuts because bamboo seedlings require shade from a forest canopy to establish [Taylor and Qin 1988a, 1993b]. In contrast, areas beneath forest had a carpet of arrow bamboo seedlings.
The loss of arrow bamboo in the Qionglai Mountains due to clearcutting and bamboo flowering was unexpected and alarming. Many of the Natural Reserves that protect pandas were established in the 1960s and 70s after they were logged [Seidensticker et al. 1984; Schaller et al. 1985]. In Wolong alone there are 15-20 km of clearcuts that bamboo has not regenerated and I calculated that these areas would support 8-10 pandas [Taylor and Qin 1993b]. The message from our research was clear, bamboo will not reestablish in recent clearcuts and clearcutting should be stopped in panda habitat because it reduces future bamboo supplies for pandas. Sadly, the whack of the loggers axe still resonates through much of the pandas' forest home.
But loggers and pandas can coexist in forests outside natural reserves [Taylor and Qin 1989a; Reid et al. 1991]. Forests of fir, hemlock, spruce, and birch that were selectively cut have a thick carpet of arrow bamboo that pandas used extensively both before and after the flowering [Schaller et al. 1985; Johnson et al. 1988a; Reid et al. 1991]. I found that arrow bamboo seedling established well in selectively cut forests and that selective cutting promoted tree seedling regeneration amongst the ranks of arrow bamboo stems. Moreover, loggers left large hollow fir, hemlock, and spruce because they had little timber value. These large trees are an important source of seed that speeds forest recovery after cutting and we found that hollows in their bases are used by giant and red pandas as maternity dens [Schaller et al. 1985; Taylor and Qin 1989a; Reid et al., 1991b]. So selective cutting does not harm panda habitat and it can be used to extract timber where pandas live on production forest land. The 1983 arrow bamboo dieback, overall, was not a crisis for pandas because they adjusted their behavior to the changing bamboo supply. In fact, I have promoted the view that the bamboo die-off is a golden opportunity for panda conservation. Arrow bamboo chokes the understory of now degraded hardwood forests that developed after clearcutting. But arrow bamboo can only dominate these old clearcut sites when it is mature. When it flowers and dies back, space for other plants becomes available. I have recommended that managers plant fir, spruce, and hemlock seedlings amongst the dead bamboo in the clearcuts so trees can grow rapidly in a competition free environment and speed recovery of mixed forest; preferred panda habitat [Taylor and Qin 1989b; Taylor et al. 1991b; Reid et al. 1991a]. Planting trees after bamboo dieback on forest production lands will also shorten the harvest rotations for valuable softwood timber.
Tree planting should be done on certain landscape positions. Pandas prefer to feed on bamboo where slopes are gentle [Schaller et al. 1985; Reid and Hu 1991] and, therefore, flat sites should be planted with trees first [Reid et al. 1991b; Taylor et al. 1991b].
Managers have been urged to respond quickly to the habitat restoration opportunity given by the arrow bamboo dieback. Bamboo seedlings are slowly developing into mature bamboo, except in the clearcuts where it has disappeared and is now being replaced by shrubs [Taylor and Qin 1988a; Taylor et al, 1991b]. Bamboo seedlings will soon be tall enough to strongly compete with planted tree seedlings. The opportunity to easily reintroduce fir, spruce, and hemlock to degraded forests with arrow bamboo is infrequent. It will not occur until about AD 2028 when arrow bamboo is scheduled to flower again [Taylor and Qin 1989b; Taylor et al. 1991b].
A panda's needs are straightforward: large interconnected tracts of forest with a diversity of bamboo species free of poachers and other human interference. But the forests pandas live in are coveted for timber, and poachers trails criss-cross even remote areas in Natural Reserves. Poachers fetch up to $10,000 for a panda pelt, and poaching is the most urgent conservation problem facing the panda [Schaller 1987, 1993]. Giving pandas the space they need will require change in people's behavior, and change in forest practices. These are enormous tasks. But without change the panda will slip silently away, its life reduced to mere script in a book or specimens in a museum. The panda can be saved, but only if people, who have pushed them to the brink of extinction, have the will to do so.