Citation Detail: Andrew T. Smith and J. Marc Foggin. 1996. The Plateau Pika Is A Keystone Species For Biodiversity On The Tibetan Plateau. in: Conserving China's Biodiversity (II) (PETER Johan Schei, WANG Sung and XIE Yan eds.). China Environmental Science Press. Beijing. 211-221p.

The Plateau Pika Is A Keystone Species For Biodiversity On The Tibetan Plateau

Andrew T. Smith and J. Marc Foggin
(Department of Biology, Box 871501, Arizona State University, Tempe, AZ85287-1501)

¡ï1. Introduction¡ï2 Pikas as a Keystone Species¡ï3 Control of Pikas¡ï4 Discussion

1. Introduction

Biodiversity is the totality of genes, species, and ecosystems in a region, and is the product of hundreds of millions of years of evolutionary history (WRI, IUGN and UNEP 1992). Biodiversity provides humankind with a variety of direct and indirect values which, combined, are essential for the well-being of the human species (Costanza et al. 1997, Primack 1998). The Biodiversity Working Group of the China Council for international

Cooperation in Environment and Development (CCICED) has calculated that the total value of biodiversity for China equates to approximately $255 - 410 billion per year (CCICED1996), and these figures - in the light of recent work (c.f Costanza et al. 1997) - are certainly underestimates.

China has recognized the importance of biodiversity conservation in many fundamental ways. China was the fourth country to ratify the Convention on Biological Diversity and actively participates in CITES. China has aggressively adopted a proactive Agenda 21 (becoming the first developing country to do so), and has produced a China Action Plan for Biodiversity Conservation and a China Country Study for Biodiversity. The CCICED is a model organization for the integration of wide-ranging ideas into a policy framework for sustainable utilization of biological resources throughout China. Environmental policy in China is as advanced as in any country on earth. The primary problem faced by China, as well as all nations, is how to successfully implement environmental policy in a world that is inherently complex.

The Qinghai-Tibet (Tibetan) Plateau occupies 2. 5 million square kilometers, approximately 25% of the area of the PRC. It has been estimated that 70% of the Plateau is rangeland, and Tibetan pastoralism is the primary sustainable use of this rangeland (Ekvall 1968, Miller and Craig 1997). Although the Plateau is not as biologically rich as some other areas in China, it does contain a distinctive flora and fauna (MacKinnon et al. 1996, Schaller 1998). Most ungulate species found on the Plateau are listed as Category I or Category II Nationally Protected Fauna in China (MacKinnon et al. 1996). At the present it is impossible to know the full extent of endangered species on the Plateau, because the list of Nationally Protected Fauna in China is heavily biased toward charismatic species and does not contain less prominent fauna even when they have been internationally recognized as threatened (Baillie and Groombridge 1996).

Biodiversity conservation on an area as vast as the Tibetan Plateau is a complex endeavor. Here, we concentrate on one feature of this ecosystem and show its relevance to the overall of biodiversity and ecosystem health on the Plateau. We present the Plateau Pika (Ochotona curzoniae) as a Keystone Species in the Tibetan Plateau ecosystem. In spite of the critical ecological role played by the Plateau Pika in this ecosystem, this species has been the target of widespread poisoning campaigns designed to eliminate it. We argue that an attempt to selectively eliminate the Plateau Pika is detrimental to the preservation of native biodiversity and the normal functioning of the Tibetan Plateau ecosystem. If China is to implement successfully its forward policy of biodiversity conservation on the Tibetan Plateau, the pika should be considered as a positive element and widespread poisoning activities to kill pikas should be halted.

2 Pikas as a Keystone Species

A keystone species is one that, if lost, will lead to a cascading effect of reduced biodiversity and interfere with the proper functioning of an ecosystem (Primack 1998). The Plateau Pika is a Keystone Species because it 1 ) makes burrows that are the primary homes to a wide variety of small birds and lizards; 2) creates microhabitat disturbance that results in an increase in plant species diversity; 3) serves as the main prey for most of the small to medium to large predators on the Plateau; 4) contributes positively to ecosystem-level dynamics by recycling soil; and 5) enhances above-ground and below-ground (root) biomass .

2.1 Biodiversity Of Symbiotic Animals
The Tibetan Plateau is largely a treeless environment, and the open meadows that constitute the majority of the plateau ecosystem offer little in the way of protection for nesting animals. The burrows constructed by the Plateau Pika offer breeding habitat for many species. Hume's ground jay (Pseudopodoces humilis) and several species of snowfinch (Montilnngilla spp.) nest mainly in pika burrows (Meyer de Schaunensee 1984, Ma 1995, Schaller 1998, Smith et al. 1990). These birds are more abundant in area of pikas (Ma 1995). Similarly, Pere David's snow finches (Pyrgiauda davidiana) and Isabelline wheatears (Oenanthe isabelfina) regularly nest in the holes of Daurian pikas (O. daurica), an ecologically similar species that also occupies portions of the Plateau (Smith et al. 1990). In addition, native lizards (Phrynocephalus, Eremias) use pika burrows for cover and as breeding sites. Loss of pikas, and thus the burrows they create, contributes to a direct loss of these species and a reduction of native biodiversity on the Tibetan Plateau.

2.2 Increased Plant Species Richness
Evidence for increased plant species richness caused by the burrowing activities of the Plateau Pika is indirect, because there are insufficient studies on this phenomenon in the Tibetan Plateau ecosystem. However, a wide variety of studies on similar species on grasslands in other areas, including Asia, all indicate that the surface disturbance caused by burrowing animals commonly increases plant species richness compared to areas without burrowing animals (Ellison 1946, Grant et al. 1980, Grinnell 1923, Huntly and Reichman 1994, Tilman 1983). On the Mongolian steppe, Dmitriev (1985) and Tsendzhav (1985) found that local floral diversity is enhanced due to the burrowing of Daurian pikas (some shrub plants only grow on pika burrows). Certain plants (Euphorbia altaica, Artemesia spp., and some crucifers) grow only on the burrows of Pallas's pika (O. pallasi pricei; Kholodova 1975). Thus, it is likely that effects of burrowing of the Plateau Pika would mirror these other studies and yield a higher plant species richness than in areas where pikas have been exterminated. Under current heavily overgrazed conditions on the Plateau, however, it would be difficult to determine the direct effects of this relationship. It will take long-term controlled experiments in areas that are not overgrazed to evaluate the full association between pikas and plant diversity.

2.3 Pikas Serve as Prey for Many Native Wildlife Species
Most of the predatory animals on the Tibetan Plateau rely heavily on pikas in their diet (Schaller 1998, Smith et al. 1990). Pikas are not only the most abundant source of food for predators during the summer, but as pikas do not hibernate, they become almost the sole source of food during winter. When pikas are exterminated, this important source of food disappears and the effect is to starve these species resulting in a loss of local biodiversity. We have driven across vast stretches of the Plateau without seeing any raptors soaring in the sky - these are the areas void of pikas due to prior poisoning campaigns. Conversely, the presence of raptors indicates areas of healthy pika populations.

The ecological niche of the Plateau Pika is similar in many respects to that of prairie dogs (Cynomys) of North America, and it is possible that the polecat (Mustela eversmanni), like its counter part in North America (the black-footed ferret M. nigripes), is tied in some way to the dynamics of pika populations (Nekipelov 1954, Schaller 1985, Smith et al. 1990). In North America the control of prairie dogs has resulted in the near extirpation of the black-footed ferret, ferrets are only being rescued from extinction with very expensive intervention (Miller et al. 1996). Similarly, other small mammals such as weasels (Mustela spp.), foxes (Vulpes spp.), and Pallas's cat (Otocolobus manul) rely heavily on pikas for food (Schaller 1998, Smith et al. 1990).

Even many larger mammalian predators such as wolves (Canis lupis), snow leopards (Uncia uncia), and brown bears (Ursus arctos) can prey on animals as small as the plateau pika (Schaller 1998). Snow leopards largely specialize on big game and pikas act only as a buffer species (Schaller 1998). However, pikas comprise over 50% of the diet of wolves in some areas (Schaller1998), and brown bears appear particularly reliant on pikas for food. One study on the Chang Tang found almost 60% of the diet of brown bears to be pikas (Schaller 1998). In fact, pikas form such an important part of the diet of brown bears on the Plateau, that when Przewalski (1883) secondarily defined this form, he called it U. lagomyiarius or "bear pika-eater." Kozlov (1899)found 25 pikas in the stomach of one bear.

Most of the large predatory birds on the Tibetan Plateau depend upon pikas as a food source: Steppe eagles (Aquila nipalensis), upland buzzards (Buteo hemilasius), saker falcons (Falco cherrug), goshawks (Accipiter gentilis), black kites (Milvus migrans), little owls (Athene noctua), etc. Schaller (1998) determined that 90% of pellets under the nest of a saker falcon contained pikas, while all of the pellets beneath the nest of an upland buzzard contained pika remains. The similar Daurian pika has been shown to comprise the following percentages of the diet of avian predators in southeast Transbaikalia steppe eagle, 62%, upland buzzard, 17%, eagle owl (Bubo bubo), 73%, and saker falcon, 22% (Peshkov 1957, 1967).

2.4 Contributions To Ecosystem Function
The Plateau Pika may contribute in many ways to enhanced functioning of the Plateau ecosystem. Like fossorial animals in other ecosystems, they may act to increase local primary plant productivity, aid in the formation, aeration and mixing of soil, and enhance infiltration of water into the soil (Ellison 1946, Grant et al. 1980, Grinnell 1923, Huntly and Reichman 1994, Tilman 1983). Burrowing mammals, such as pikas, do not increase levels of erosion rather erosion in most cases is caused by overgrazing of vegetation by domestic livestock (Ellison 1946. Schaller 1985). Jiang and Xia (1985, 1987) determined that the foraging of small and moderate populations of Plateau Pikas is selective and may play an important role in the stabilization of the alpine meadow vegetational community. Studies on the similar Daurian pika have shown that their digging activity loosens and improves the soil, and the accumulation of their excrement and leftover stores in the burrow system yields high levels of organic materials. Soil temperatures and humidity are higher, and concentrations of nitrogen, calcium, and phosphorus are greater near burrow systems then in nearby areas without pika burrows (Tsendzhav 1985). The result of these interactions is a greater biomass of roots taller plants, and increased density of plant cover near the burrow system then elsewhere, the overall biomass of plants when growing over the burrow system of Daurian pikas is almost five times greater than in the surrounding steppe (Tsendzhav 1985). In addition, in spring the phenology of plants growing over burrow systems is 10 to 15 days advanced over plants growing on areas of nearby steppe without pikas. Similar studies have been recently been conducted on another fossorial animal found on the Tibetan Plateau and which is also subject to control, the zokor (Myospalax baileyi). The above ground biomass of plants on zokor mounds which emerged on an area of primary vegetation was significantly higher than on an area of secondary vegetation. The soil of fresh and old mounds was higher in available nitrogen and phosphorus content than randomly collected samples. Above ground biomass surrounding mounds was significantly higher than control areas (Wang et al. 1993).

2.5 Summary
The Plateau Pika fulfills all of the component parts of the definition of a Keystone Species: 1) its presence adds ecological niches in the environment which are used by native species; 2) it serves as an important source of food to a wide variety of predators - many of which can not survive without the pika; and 3) it contributes in a meaningful way to many aspects of ecosystem function. Of this list, the first two constitute direct benefits to biodiversity on the Tibetan Plateau. The third function concerning ecosystem services is indirect, and currently there are no long-term studies as to how the poisoning of pikas (and zokors) may change ecosystem functioning on the Plateau. There is a pressing need for controlled investigations and analyses of this problem.

3 Control of Pikas

3.1 Background
In spite of these contributions of the Plateau Pika to the Tibetan Plateau ecosystem, this species has been targeted for control (Fan et al. 1986, Liu et al. 1980, Ma 1995, Schaller 1985, Shen and Chen 1984, Smith et al. 1990, Zhong et al. 1985). Here "control" is defined as the deliberate poisoning of a native species with the intent to reduce its density or eliminate it altogether from a geographic area. The Plateau Pika is the focus of this paper, but other species of pika (O. daurica, O. pallasi) and the zokor (Myospalax spp.) are also targets of control. Within China these activities are normally termed "rodent control" - but as pikas are lagomorphs and not rodents, this is a misnomer (a better term would be small-mammal control, in this paper we simply use the term control).

Plateau Pikas have been targeted for control primarily because they putatively cause negative effects on rangeland habitat and are competitors for forage that could otherwise be utilized by livestock. In addition, the many burrows constructed by pikas are hazards to individuals who ride horses across the grasslands. These reasons are identical to those given in the United States for control of prairie dogs (Cynomys ) - a species that is ecologically equivalent to the Plateau Pika (Schaller 1985). A final reason for pika control efforts is more insidious: the practice continues because of vested interests by agencies responsible for control. In China (like in the United States and elsewhere), agencies and research institutions with a primary responsibility for control issues can take on a life of their own - and for them to admit that some of their practices are outdated or unnecessary would put some employees out of work.

3.2 Extent of Control
Placing the blame for degraded rangelands on the Plateau on pikas is an old pursuit. Ekvall (1964) who worked in the Amdo region during the 1930s commented that pastoralists attributed "black sands" areas to pikas (see below). This observation pre-dates rangeland degradation that has occurred in the past three decades. In Qinghai currently it has been estimated that "rodent" (viz. pika or zokor) infested areas total 44,720 km2 (Jing et al. 1986). Control is recommended when the populations of these small mammal reach high densities (thus the term "infested").

Control of small mammal populations on the Plateau began in 1958 with tests on the Plateau Pika. Large-scale control efforts were initiated in 1962 and reached a peak between 1963 and 1965 (13 million ha, Smith et al. 1990) and continues on a reduced scale. Between 1986 and 1994 alone, nearly 7.5 million ha of grassland were controlled (poisoned) to eradicate small mammals and, to a lesser degree, insects (QAHB 1996). In 1997 we observed television programs in Beijing extolling the virtues of rodent control measures on the Plateau. The continued emphasis on control programs indicates that extremely large tracts of pastureland on the Plateau have been poisoned to kill pikas (Ma 1995, Schaller 1985, 1998, Smith et al. 1990).

An unfortunate consequence of these campaigns to kill Plateau (and Daurian and Pallas's) pikas is that other pika species may inadvertently be targeted and killed. The IUCN/SSC Lagomorph Specialist Group recognizes several species and subspecies of pika in China (most inhabiting the high plateau country) as being threatened (Baillie and Groombridge 1996, Chapman and Flux 1991; see also Ma 1995). Most pikas look alike, thus there is opportunity for some of these threatened forms to be inadvertently poisoned - with direct negative ramifications of efforts to preserve native biodiversity on the plateau.

In the early years, when extensive control occurred, the main chemicals were Compound 1080 and Fussol, applied from hand spreaders, tractors, and often broadcast from airplanes. Application of Fussol was discontinued because of expense and environmental contamination from poisonous secondary derivatives. Use of Compound 1080 continued until 1978 when it was determined that it also killed carnivores that ate pikas (Smith et al.1990). Because of the importance given to pest control in China, research on poisons has been well-funded, and almost all rodenticides have been tried. Gophacide and Zinophos then became the primary control agents. A major emphasis is now the use of anti-coagulants to avoid secondary by-products and damage to the environment. Anti-coagulants, however, are expensive and must be distributed in large quantity over several applications to be effective (Smith et al. 1990).

3.3 Relationship of Pikas and Livestock
Plateau Pikas are controlled because they putatively compete with livestock for forage and are found at very high densities (which may reach 300/ha; Liu et al. 1980, Shen and Chen 1984. Smith et al. 1990). This condition normally occurs at the end of summer when the pikas normally high rate of reproduction has yielded its greatest population density and the meadow vegetation has begun to dry up.

There is also an apparent relationship between grazing of domestic animals and pika population density. When the combined grazing of yaks, sheep and horses lower the degree of cover and the height of vegetation, Plateau Pikas may be found at greater densities than on natural meadows (Shi 1983). Similarly, Daurian Pikas are more likely to contribute to the deterioration of rangelands that are already overgrazed (Zhong et al.1985). The zokor, another species that is often held responsible for soil and vegetation disturbances in alpine plant communities on the Plateau, reportedly has a higher survival rate in heavily grazed sites (Cincotta et al. 1992). Apparently under conditions initiated by pikas and other small mammals on the Plateau are in a position to do greater harm to the grassland environment. The question can be raised as to whether the high densities of pikas (and zokors) are causal of rangeland degradation, or merely symptomatic of overgrazing by livestock (see also Cincotta et al. 1992). Shi (1983) concluded that the most effective way to control damage by pikas would be to improve the condition of the range, which would presumably mean to reduce the intensity of grazing by domestic livestock.

Competition between Plateau Pikas and livestock is dependent on the density of pikas and the quality of the rangeland. The foraging of pikas at low and moderate densities (such as found in normal situations) is selective and overtaps little with the diet of domestic grazing animals, thus the pikas may play an important role in the stabilization of the alpine meadow vegetation community, (Jingo and Xia 1985, 1987). However, the diet of pikas may overlap extensively with livestock when pika density is extremely high (such as found in situations where rangelands are already heavily grazed, see above, Jiang and Xia 1985, 1987). The situation is compiicated further in that overgrazing incr6ases the proportion of ptants unpalatable or poisonous to livestock (Lang et al. 1997), and pikas tend to eat those plants that livestock do not eat (Stellaria, Ligularia, Schaller 1985, 1998). Schaller (1985, 1998) further determined that by eating herbs, pikas may slow the spread of plants not palatable to livestock, and thus improve conditions of grasses and sedges preferred by livestock. Thus, pikas do not appear to compete with livestock for forage on well-managed ranges, and on degraded ranges where pika densities are high and competition might exist, pikas still provide a beneficial service to the livestock.

The alpine meadows of the Tibetan Plateau contain areas of small depressions with steep edges and larger areas primarily devoid of vegetation, termed locally as "black sands". Areas of extensive black sands lack the deep sod base that is so typical of the alpine meadows, although on many of these areas a luxurious monoculture of mint (Elscholtzia sp.) grows late in the summer season. Plateau Pikas are thought to contribute to the origin of these areas, which are indicative of degraded rangeland. As pikas have been observed frequently to utilize the edges of these depressions as cover and to eat plant roots there, it has been suggested that pikas may expand the extent of black sand areas (Schaller 1998). As mentioned earlier, Ekvall (1964) commented that pikas were responsible for the creation of black sand areas as early as the 1930s.

One of us (Smith) has worked extensively with fully marked populations of the Plateau Pika, and one analysis directly addressed the issue of habitat utilization by pikas on the alpine meadow. Working on a 2 ha study area containing 26 pika families, we identified three distinct habitat types: 1) meadow dominated by sedges (Carex spp.; Kobresia spp.), 2) small depressions (1 - 5 m across) in the sedge meadow that contained a variety of grasses and sedges, and 3) black sands. Although meadow accounted for an average of 84% of the area, 47.4% of our 800 grid cells contained black sand habitat. Additionally, we measured habitat edges, slope of the habitat, burrows and "duck holes" (burrows that did not link to an underground burrow system). We examined the relationship of dispersal movements of marked individuals to determine if they favored any of these components of the habitat. Although 192 correlational tests were run with both parametric and non-parametric procedures, we found that none of the family ranges of dispersing and philopatric pikas exhibited significant differences in any of the habitat characteristics (Dobson and Smith, in press). Thus, although casual observations have suggested that pikas prefer edges offered by depressions and black sands, our work does not confirm this conclusion; rather, it shows that pikas tend to utilize the alpine meadow randomly.

There are other hypotheses as to the origin of black sand areas. One hypothesis is that trails cut on hillsides by livestock cause the compaction of soil and formation of erosion terraces - processes that may initiate the decline of a meadow and lead to a black sands situation (Ma et al. 1997, Schaller 1998). Another hypothesis is that anthropogenic activities, such as the cutting of sod by pastoralists to make walls and fireplaces, create clearings that expand into black sands. Finally, a demonstrable warming trend has been detected on the Plateau (Miehe 1988, 1996), and the resulting desiccation may be changing the meadow into a semiarid alpine steppe across much of the Tibetan Plateau. These changes in climate may be upsetting the natural balance of the vegetation community leading to formation of black sands.

4 Discussion

4.1 Potential Outcomes of a Pika Control Program
We began this paper by depicting China's progressive policies toward the promotion of biodiversity and conservation. We now find that there are other governmental policies in China that promote the control of an ecological keystone species such as the Plateau Pika, which operate in opposition to China's biodiversity policies. Contravening policies are common in all countries, but they establish a situation that requires a close look at the costs and benefits of each policy. In the case of the Plateau Pika, let us look at the potential outcomes of a pika control program and then weigh each of these against the goal of biodiversity and ecosystem conservation.

Potential Outcome #1: All pikas can be eliminated by control programs, and this can be done without
affecting other components of the Tibetan Plateau ecosystem and its biodiversity

1£© Any attempt to eliminate (exterminate) all pikas would be very costly, and may be ecologically impossible.
2£© Even if all pikas were to be eliminated, the analysis presented here clearly shows that pikas are a Keystone Species - their presence is necessary for the survival of many different classes of animals and plants, and they perform essential ecosystem services. Thus, Potential Outcome #1 is an impossible one - the pikas cannot be eliminated without seriously comprising the biodiversity and health of the Tibetan Plateau.

Potential Outcome #2: All pikas can be eliminated and the result will be a stimulation of productivity of livestock.

1£© Any attempt to eliminate (exterminate) all pikas would be very costly, and may be ecologically impossible.
2£© This outcome is based on the premise upon which most control programs are based that controlling the Plateau Pika will yield positive benefits to the economically important utilization of the Tibetan grassland ecosystem. Massive control has taken place over the past three decades, and many broad regions of the Tibetan Plateau have been cleansed of pikas. In spite of the extent of control, however, productivity and health of livestock on the Plateau has been declining. Livestock weight has declined. Animals are weak when entering the critical winter season and massive starvation of herds is becoming a too frequent event (viz. the situation in Yushu and other ounties in southern Qinghai recently). Clearly, the control policy has not improved the situation on the grasslands appreciably.

Potential Outcome #3: Pikas are eliminated only temporarily or in restricted geographical areas.

The cost of control is high, thus the cost of partial control, with the pikas and other controlled animals in a position to bounce back in population within a short time, is not cost-effective. Short-term control may be warranted in conjunction with other activities to restore the health of degraded pasture (such as removal of livestock), but these situations must be evaluated carefully to assure that they do not jeopardize the role of pikas as a Keystone Species in the ecosystem. In a general sense, if control is too pervasive, then biodiversity conservation and ecosystem function are jeopardized, but if control is too restricted (temporally or spatially), it is bound to be ineffective. Thus, it appears that short-term control is unsustainable and not cost effective.

Potential Outcome #4: Control fails to eliminate pikas.

The cost implications of failure are obvious. In this outcome, control is a lose-lose situation - cost is high and there are no potential benefits of control. That expensive control measures that accomplish little should not be undertaken is obvious. Yet, we have recent examples of just such cases. We were recently informed (1997) by individuals in an international aid agency in Beijing that they were promoting a control program directed at Plateau Pikas in spite of the fact that they knew such an activity was counter-productive to long-term sustainability in the area in which they were working. Their rationale for funding the control program was to demonstrate that they were actually spending money in China - regardless of the outcome. Ultimately this program collapsed because it failed to control pikas and was too expensive. It is unfortunate that the resources spent on this misguided venture were not expended for programs consistent with a long-term goal - a sustainable program for utilization of the rangelands.

4.2 Comprehensive Rangeland Management
In this paper we have addressed a single component of the complex of factors governing the biodiversity and sustainable development of the alpine grasslands of the Tibetan Plateau. We wish to stress that these rangeland resources must be viewed comprehensively. Just as we have documented that campaigns to poison Keystone Species such as pikas and other small mammals in this ecosystem is short-sighted, when viewed from the perspective of biodiversity and ecosystem conservation, so should we be skeptical of other apparent quick fixes to this ecosystem (such as planting grasses, fencing, winter housing, etc.). Some of these approaches may be justified in a restricted geographical area, but the entire alpine grassland ecosystem on the Tibetan Plateau is at risk (Lang et al. 1997, Miller 1995, Schaller 1998). What are needed are studies to evaluate the long-term sustainability of the ecosystem - studies that begin with the health of the alpine meadow which is the cornerstone for all biodiversity and pastoralist activity. Here we have shown that the Plateau pika is one key element in this system and have argued that managing pika populations should be viewed as part of the solution, not part of the problem, facing a comprehensive solution for the restoration of Tibet's rangelands.

Acknowledgements
We would like to thank IUCN - The World Conservation Union, the Biodiversity Working Group of the China Council for International Cooperation in Environment and Development, and the Center for Asian Studies at Arizona State University for their generous support of this endeavor. Harriet Smith kindly read and commented on the manuscript.

Bibliography
Baillie, J., and B. Groombridge (eds). 1996. 1996 IUCN Red List of Threatened Animals. IUCN. Gland, Switzerland.
CCICED. 1996. Proceedings of the fourth meeting of the China Council for International Cooperation on Environment and Development. Beijing.
Chapman, J. A., and J. E. C. Flux (eds). 1990. Rabbits, hares, and pikas: Status survey and conservation action plan. IUCN. Gland, Switzerland.
Cincotta, R., Y. Zhang, and X. Zhou. 1992. Transhumant alpine pastoralism in northeastern Qinghai Province: An evaluation of livestock population response during China's agrarian reform. Nomadic Peoples 30:3-25.
Costanza, R., R. d'Arge, R. deGroot, S. Farber, M. Grasso, B. Hannon, K. Limburg, S. Naeem, R.V. O'Neill, J. Paruelo, R.G. Raskin, P. Sutton, and M. van den Belt. 1997. The value of the world's ecosystem services and natural capital. Nature 387: 253-260 .
Dmitney, P. P. 1985. The relationship between some shrubs of the Mongolian steppes and colonies of mammals. Zh. Obshch Biol. 46: 661-669.
Dobson, F. S., and A. T. Smith. In Press. Social and ecological influences on dispersal and philopatry in the plateau pika (Ochotona curzoniae). Behavioral Ecology.
Ekvall, R. B. 1968. Fields on the Hoof: Nexus of Tibetan Nomadic Pastoralism. Waveland Press. Prospect Heights, Illinois.
Ellison, L. 1946. The pocket gopher in relation to soil erosion on mountain ranges. Ecology 27: 101 -114.
Fan, N., Z. Jing, Q. Wang, and W. Zhou. 1 986. Studies on bromadiolone against the pika and the zokor. Acta Theriologica Sinica 621 1 -21 7.
Grant, W., N. French, and L. FoISe. 1980. Effects of pocket gopher mounds on plant production in shortgrass prairie ecosystems. Southwestern Naturalist 25: 215-224.
Grinnell, J. 1923. The burrowing rodents of California as agents in soil formation. Journal of Mammalogy 4: 137-149.
Huntly, N., and O. J. Reichman. 1994. Effect of Subterranean mammalian herbivores on
vegetation. Journal of Mammalogy 75: 852-859
Jiang, Z., and W. Xia. 1985. Utilization of food resources by plateau pikas. Acta Thenologica Sinica 5: 251 -262 .
Jiang, Z., and W. Xia. 1987. The niches of yaks, Tibetan Sheep and plateau pikas in the alpine meadow ecosystem. Acta Biologica Plateau Sinica 6: 115-146.
Jing, S., et al. 1986. The Situation of Qinghai Province. Qinghai People's Publishing House. Xining.
Kholodova, M. V. 1975. Studies of the connections between the Mongolian pika (Ochotona pricei Thomas) and other vertebrates of Tuva's stony mountain steppe. Unpublished report. Dept. of Vertebrate Zoology, Moscow State University, Moscow.
Koslov, P. K. 1899. Proceedings of the Expedition of the Imperial Russian Geographical Society in Central Asia 1893-1896 under the direction of Roborovski, volume 2.
Lang, B-N., J-S. Huang, and HY. Wang. 1997. Report on the pasture and livestock survey in Hainan IFAD Project Area. IFAD.
Liu, J., Y. Zhang, and G. Xin. 1980. Relationship between numbers and degree of harmfulness of the plateau pika. Acta Zoologica Sinica 26: 378-385.
Ma, M. 1995. Suggestions for the protection of some pikas. China Nature 2:26.
Ma, Y., B. Lang, and Q . Li. 1997. Improve yak productivity through resuming "black soil type" deteriorated grassland, pp. 291-294. In: R. Yang, X. Han, and X. Luo, (eds.). Yak Production in Central Asian Highlands. Qinghai People's Publishing House, Xining.
MacKinnon, J., M. Sha, C. Cheung, G. Carey, Z. Xiang, and D. MelVille. 1996. A Biodiversity Review of China. World Wide Fund for Nature International. Hong Kong.
Miehe. G. 1988. Geoecological reconnaissance in the alpine belt of southern Tibet. GeoJournal 17: 635.-648.
Miehe, G. 1996. On the connection of vegetation dynamics with climatic changes in High Asia. Palaeogeography, Palaeoclimatology Palaeoecolgy 120: 5-24.
Miller, B., R. P. Reading, and S. Forrest. 1996. Prairie Night: Black-footed Ferret and the Recovery of Endangered Species. Smithsonian Institution Press. Washington, D. C.
Miller, D. J. 1995. Herds on the move: Winds of change among pastoralists in the Himalayas and on the Tibetan plateau. International Centre for integrated Mountain Development. Kathmandu, Nepal.
Miller, D. J., and S. R. Craig (eds). 1997. Rangelands and Pastoral Development in the Hindu Kush-Himalayas. International Centre for integrated Mountain Development. Kathmandu, Nepal.
Meyer de Schauensee, R. 1984. The Birds of China. Smithsonian Institution Press. Washington DC.
Nekipelov, N. V. 1954. Changes in numbers of the Daurian pika in southwest Transbaikalia. Izv. Irkutsk. Nauchno-issled. Protivochumnpeo. Ins. Sib. Dal'nego Vost. 12: 171 -180.
Peshikov, B. I. 1957. Data on numbers and diet of foathered predators of southeast Transbaikalia. Izv. Irkutsk. Nauchno-Issted. Protivochumnogo. Ins. Sib. Dal'nego Vost. 16: 143-153.
PeShikov, B. I. 1967. On the biology of the upland buzzard. Isv. Irkutsk. Nauchno-Issled. Protivochumnogo. Ins. SIb. Dat'nego Vost. 27: 167-174.
Prezewalski, N. M. 1883. From Zaisan Lake through the Kham region of Tibet and the head of the Yellow River, second edition (1948). Moscow.
Primack, R. B. 1998. Essentials of Conservation Biology, Second edition. Sinauer. Sunderland, Massachusetts.
QAHB. 1996. Summary of experiences, goal clarification and the promotion of animal husbandry. Provincial animal husbandry development and its future tasks. Qinghai Animal Husbandry Bureau. Xining.
Schaller, G. B. 1985. Wildlife in the middle kingdom. Defenders 60:10-15.
Schaller, G. B. 1998. Wildlife Of the Tibetan Steppe. University of Chicago Press. Chicago.
Shen, S., and Y. Chen. 1984. Preliminary research on ecology of the plateau pika at the Dawu area, Guoluo, Qinghai Province. Acta Theriologica Sinica 4: 107-115.
Shi, Y. 1983. On the inftuence of rangeland vegetation to the density of plateau pikas (Ochotona curzoniae). Acta Theriologica Sinica 3: 181-187.
Smith, A. T., A. N. Formozov, R. S. Hoffmann, Zheng Changlin and M. A. Erbajeva. 1990. The pikas. pp. 14-60. In: J. A. Chapman and J. E. C. Flux (eds.) Rabbits, hares, and pikas: Status survey and conservation action plan. IUCN. Gland, Switzerland.
Tilman, D. 1983. Plant succession and gopher disturbance along an experimental gradient. Oecologia 60: 285-292.
Tsendzhay, D. 1985. The role of the Daurian pika (Ochotona daurica Pallas, 1776) in the
biogeocenosis of Eastern Khangai. Abstract of Candidate degree thesis. UIan-Bator, Mongolia.
Wang, Q., J. Bian, and Y. Shi. 1993. Influence of plateau zokor mounds on the vegetation and soil nutrients in an alpine meadow, Acta Theriologica Sinica 13: 31-37.
WRI, IUCN, and UNEP. 1992. Global Biodiversity Strategy. World Resources Institute. Washington, DC. .
Zhong, H., Q. Zhou, and C. Sun. 1985. The basic characteristics of the rodent pests on the pasture in Inner Mongolia and the ecological strategies of controlling. Acta Theriologica Sinica 5: 241-249.