Citation Detail: CHIEN-HSUN LAI and ANDREW T. SMITH. 1996. Keystone status of Plateau Pikas (Ochotona curzoniae): effect of control on biodiversity of native birds. in: Conserving China's Biodiversity (II) (PETER Johan Schei, WANG Sung and XIE Yan eds.). China Environmental Science Press. Beijing. 222-230p.

Keystone status of Plateau Pikas (Ochotona curzoniae):
effect of control on biodiversity of native birds

CHIEN-HSUN LAI and ANDREW T. SMITH*
(Department of Biology, Box 871501, Arizona State University, Tempe, AZ85287-1501)

★Introduction★Methods★Results★Discussion

Keywords: Avian biodiversity, China, keystone species, Ochotona curzoniae, plateau pika
Summary
The plateau pika (Ochotona curzoniae) of the Qinghai-Xizang (Tibetan) plateau, People's Republic of China, has been considered a pest because it putatively competes with native livestock for forage and contributes to rangeland degradation. As a result the plateau pika has been poisoned across vast areas of the high alpine meadows of the plateau. The plateau pika has also been considered a keystone species for biodiversity on the plateau. We investigated the effects of poisoning plateau pikas on avian species richness and abundance. We conducted standardized censuses of birds on a number of sites across the alpine grassland of Qinghai province on which pikas either had or had not been poisoned. Avian species richness and abundance were higher on non-poisoned sites, in particular for species that nest in pika burrows such as Hume's groundpecker (Pseudopodoces humilis) and six species of snowfinch (Montifringilla spp., Pyrgilauda spp.), and species that prey on pikas (upland buzzard, Buteo hemilasius; black-eared kite, Milvus lineatus). The plateau pika thus appears to be both an allogenic engineer and a keystone species. To ensure that poisoning does not reduce biodiversity of native species, management decisions concerning plateau pikas should reflect caution and careful assessment.

Introduction

When native small mammals reach high densities they are often labeled as pests and subjected to control measures. Such is the case with the plateau pika (Ochotona curzoniae), whose distribution largely coincides with the high alpine grasslands of the Qinghai-Xizang (Tibetan) plateau of China. For decades the pikas have been blamed for reducing available forage for domestic livestock (yak, sheep, horses), and causing habitat degradation (locally termed "black sands") (Ekvall, 1968; Schaller, 1985; Wang et al., 1995; Fan et al., 1999). As a result of these preconceptions, and concomitant with the increasing degradation of these alpine grasslands over the past forty years (Lang et al., 1997), the plateau pika has become the object of an extensive control effort. Poisoning of the plateau pika and the Chinese zokor (Myospalax fontanierii), began in 1958 and escalated greatly by 1962 (Fan et al., 1999). In 1964 and 1965 more than 26,667 km2 in 20 counties in Qinghai province were treated with both zinc phosphate and floroacetate (Compound 1080; Fan et al., 1999). Fan et al. (1999) estimate that in Qinghai from 1960 to 1990 "Cumulatively, more than 208,000 km2 ... was treated with rodenticides..." A separate estimate by Drandui (1996) concludes that between 1986 and 1994 insect and "rodent" control programs were broadcast over an area of 74,628 km2 - nearly one-fifth of Qinghai's provincial grazing lands. These massive poisoning programs are ongoing in spite of the absence of a comprehensive analysis of the role of the plateau pika in the ecosystem (Smith and Foggin, 2000).

An alternative viewpoint to the perspective that common native small mammals are pests is that they may be important ecological components of the ecosystem (Dickman, 1999). In many cases, small mammals may be considered "allogenic engineers" because their activities alter the environment by transforming living or abiotic materials from one state to another by mechanical or other means (Dickman, 1999). An example would be the construction of burrows by one species that are later used by other species such as birds, reptiles, or other small mammals. In extreme cases, because of their abundance and potential ecological importance, certain small mammals can be considered "keystone species." A keystone species is one whose elimination or major decimation from an ecosystem would have a greater than average effect on other species' populations or ecosystem processes - one that has a disproportionately large influence on other species in a community (Heywood, 1995; Power et al., 1996). The role of a keystone species should be unique, compared to other ecological processes or the roles of other species (Kotliar, 2000).

The plateau pika has been considered a keystone species for biodiversity on the plateau, based on a review of the natural history and ecology of the pika and those species living in sympatry with it (Smith and Foggin, 1999, 2000). The burrowing activity of pikas may minimize soil erosion, enhance the ability of soil to absorb precipitation, contribute to nutrient cycling, and create microhabitats resulting in increased plant species richness. Finally, the pika serves as the principal prey for nearly all of the plateau's predators (Smith and Foggin, 1999, 2000).

Which is the correct view? Is the plateau pika a pest or a keystone species? Should the massive poisoning of pikas be continued, or should plateau pikas be managed with an objective of preserving the characteristic biodiversity of the Qinghai-Xizang plateau? A first step to answering these questions is to obtain a clear picture of the role played by the plateau pika in the alpine meadow ecosystem.

In this study we examined avian species richness in relation to the presence of plateau pika populations. As allogenic engineers, the burrows constructed by the plateau pika function as breeding habitat for many species. Hume's groundpecker (Pseudopodoces humilis) and several species of snow-finch (Tibetan snowfinch Montifringilla adamsi, white-winged snowfinch M. nivalis, plain-backed snowfinch Pyrgilauda blanfordi, small snowfinch P. davidiana, rufous-necked snowfinch P. ruficollis, white-rumped snowfinch P. tacazanowskii) nest primarily in pika burrows (Prejevalski, 1876; Meyer de Schauensee, 1984; Feng et al., 1986; Ma, 1995; Schaller, 1998; MacKinnon, 2000). Similarly, Small snow finches (P. davidiana) and Isabelline wheatears (Oenanthe isabellina) 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). The large predatory birds on the plateau may depend on the plateau pika as a food source. Schaller (1998) determined that 90% of pellets under the nest of a saker falcon (Falco cherrug) contained pikas, while all of the pellets beneath the nest of an upland buzzard (Buteo hemilasius) 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 (Aquila nipalensis), 62%; upland buzzard, 17%; Eurasian eagle owl (Bubo bubo), 73%; and saker falcon, 22% (Peshikov, 1957, 1967). These observations indicate that avian species richness may be higher in areas where plateau pikas occur, as opposed to areas where they have been controlled (Ma, 1995; Smith and Foggin, 1999, 2000), however there have been no direct tests of this hypothesis. This study directly examines avian species richness in areas where pika populations have been drastically reduced or eliminated by poisoning in contrast to areas harboring undisturbed populations of pikas.

Methods

Study areas
Thirteen sites on the alpine grasslands of Qinghai province were chosen for sampling during summer 2000 (Figure 1) using a variety of criteria. Evidence on each site indicated current or prior presence of Ochotona, and areas classified as a winter pasture were selected to ensure as much consistency among sites as possible. In addition, it was essential that we could obtain historical information about each site from interviews with representatives of the local Animal Husbandry Bureau (AHB) and/or local pastoralists. The sites stretched across most of Qinghai province, occurring in four of eight provincial prefectures and in eight different counties (Figure 1), and ranged in altitude from 3,200 - 4,300 m.

Figure 1. Map of Qinghai Province, People's Republic of China. Shaded portion delimits the high alpine grasslands of the Qinghai-Xizang (Tibetan) plateau. Soft shaded lines outline the boundaries of major provincial prefectures. Numbered circles indicate study sites: 1 - 7 = sites where plateau pikas (Ochotona curzoniae) had not been poisoned; 8 - 13 = sites where plateau pikas had been poisoned.

At six sites plateau pikas had been extensively poisoned within the past 3 - 5 years, reducing their respective populations to less than 5% of their former density. These sites were characterized by abandoned and collapsed pika burrows. Seven sites had not been poisoned within the past 3 - 5 years, and scan censuses gave estimates of healthy pika populations with a minimum density of 50 pikas per ha. Two of the non-poisoned sites had been targeted for poisoning by the local AHB, but the resident pastoralists had not complied with this recommendation because of their religious beliefs.

Survey methods
At each site three standardized 1 km transect lines were established (only two transects were established at Site 1) that extended for the length of each study site. The placement of these transect lines reflected the variability of terrain and vegetation at each site. The transects were marked with colored flags.

An index of presence and prior presence of pikas was obtained by counting the number of fresh (open) pika burrows and/or the number of collapsed pika burrows along and within 1.5 m to each side of each transect line (3 m wide area). Collapsed burrows were further classified as partially collapsed or fully collapsed.

An index of the presence and abundance of bird species was used to indicate biological diversity. Each transect line was censused three times each day (morning, mid-day, and afternoon) for two days, and each census took one hour. The index of the presence and abundance of each bird species at each site was calculated as the number of individuals sighted per hour averaged across the three transect lines, three times of day, and two days of each census. All birds sighted during a survey were tallied, and care was taken to avoid double counting. A total of 228 hours was dedicated to bird censuses. Common and scientific names of bird species follow MacKinnon (2000).
Figure 2. Observations per hour (see text for sampling regime) of 14 bird species on sites where plateau pikas (Ochotona curzoniae) had been poisoned (open bars) and sites that had not been poisoned (closed bars) sites, Qinghai Province, People's Republic of China. Bird species include Tibetan snowfinch (Montifringilla adamsi), white-winged snowfinch (M. nivalis), plain-backed snowfinch (Pyrgilauda blanfordi), small snowfinch (P. davidiana), rufous-necked snowfinch (P. ruficollis), white-rumped snowfinch (P. tacazanowskii), Hume's groundpecker (Pseudopodoces humilis), Eurasian tree sparrow (Passer montanus), red-billed chough (Pyrrhocorax pyrrhocorax), horned skylark (Eremophila alpestris), fork-tailed swift (Apus pacificus), crested lark (Galerida cristata), black-eared kite (Milvus lineatus) and upland buzzard (Buteo hemilasius).

Results

Pika populations
Counts of fresh and collapsed pika burrows reflect the differences in population density of plateau pikas between the poisoned and non-poisoned sites. Pikas continually maintain their burrows (Smith and Wang, 1991), and burrows collapse and degenerate quickly after pikas have been eliminated from an area. The average number of fresh burrows tallied along each three 1 km transect lines at each of the seven non-poisoned sites was 111.7 (SEM = ±9.6), which was indicative of the high populations of pikas at these sites. In contrast, no fresh burrows were recorded along the transects at any of the six poisoned sites, and the average number of partially and fully collapsed burrows was 15.2 (SEM = ±1.9) and 28.3 (SEM = ±3.9), respectively. At three of the poisoned sites, pikas were encountered infrequently (an average of less than three sightings/h), and no pikas were seen at the remaining three sites.

Avian censuses
Overall, an average of 6.4 bird species was encountered on non-poisoned sites (N = 7; SEM = ±0.3), whereas less than half this number (X = 3.7; N = 6; SEM = ±0.9) was tallied on poisoned sites. Even more striking was the difference in overall abundance, indicated by the total number of birds seen per hour at each site. Nearly five times as many birds were present on non-poisoned sites (X = 61.2/h; N = 7; SEM = ±5.5) as on poisoned sites (X = 12.5/h; N = 6; SEM = ±2.0).

The differences in abundance between poisoned and non-poisoned sites were largest for those species that rely on pika burrows for nest sites and cover (Figure 2). For the obligate burrow-dwellers (Hume's groundpecker and six snowfinch species), an average of 3.57 species (SEM = ±0.3) was found on non-poisoned sites, whereas an average of only 0.67 species (SEM = ±0.3) was found on poisoned sites.

The rate of sighting of all snowfinch species combined was significantly higher on non-poisoned than poisoned sites (Mann-Whitney U = 42; P < 0.01), and there were almost thirteen times as many snowfinches on non-poisoned sites (X = 16.6/h; N = 7; SEM = 2.4) than poisoned sites (X = 1.3/h; N = 6; SEM = 0.7). Three of the snowfinch species were never seen on poisoned sites, and five of the six species were sighted more frequently on non-poisoned versus poisoned sites. These differences were significant for two species: white-winged snowfinch (Mann-Whitney U = 42; P < 0.01) and rufous-necked snowfinch (Mann-Whitney U = 33; P < 0.01).

Hume's groundpecker was nearly 20 times more abundant on non-poisoned sites (X = 19.9/h; N = 7; SEM = ±2.3; 2,418 sightings in 120 census hours) than poisoned sites (no sightings in 108 census hours), a difference that was significant (Mann-Whitney U = 42; P < 0.01).

The other most commonly sighted species, the horned skylark, was also seen somewhat more frequently on non-poisoned sites than poisoned sites (6.8 versus 4.8 sightings/h; Figure 2), even though it is not an obligate nester in pika burrows. It commonly occurred on all sites. In contrast, the other passerine birds were seen infrequently, and three of them (the crested lark, red-billed chough, Eurasian tree sparrow) were seen more frequently on the poisoned sites (Figure 2). None of these differences, however, were significant (Mann-Whitney U tests).
Both raptor species were tallied more often on non-poisoned versus poisoned sites, but the difference was significant only for the black-eared kite (Mann-Whitney U = 34; P = 0.05). Black-eared kites and upland buzzards were seen respectively 3.6 times and 11.2 times more frequently on non-poisoned than poisoned sites. Of raptor sightings on poisoned sites, 80% occurred at the two localities (Sites 9 and 10) where remnant populations of pikas yielded censuses exceeding one sighting per hour.

Discussion

For approximately 2,200 years the high alpine grasslands of the Qingahai-Xizang (Tibetan) plateau (extending over 2.5 million km2 or approximately one-quarter the area of the People's Republic of China), have supported sustainable animal husbandry of domestic yak and other livestock by pastoralists, along with a characteristic flora and fauna adapted to the harsh conditions of intense cold and high elevation (Goldstein and Beall, 1990; Zhou, 1992; Miller, 1995; Richard, 2000). Now, this region is at a crossroads. Over the past few decades these rangelands have suffered severe degradation, and many independent research programs have concluded that the current productivity of plateau rangelands is around 30 percent less than the productivity measured only two decades ago (Foggin, 2000). One comprehensive study that sampled 111 villages in a relatively productive region of Qinghai province found that during a 30 year period above-ground biomass decreased by over 73%, while prevalence of toxic plants increased by 5.6 times (Lang et al., 1997). Simultaneously the native fauna has been greatly reduced, and nearly every medium-to-large mammal species is now classified as threatened (Category I or Category II species under the PRC Wildlife Protection Law; MacKinnon et al., 1996; Schaller 1998). It has become increasingly clear that the alpine grassland ecosystem on the Qinghai-Xizang plateau is in danger of losing much of its native biodiversity and its ability to support sustainable pastoralism, the region's primary economic activity (Foggin, 2000). Accurate data and analyses of the alpine grassland ecosystem and its ecology are necessary to reverse this trend and ensure that the next generation of environmental management decisions in the region will lead to long-term restoration of the health of the grasslands and its biodiversity.

This paper addresses one "development" activity which has been promulgated widely on the plateau for approximately 40 years: the poisoning of plateau pikas in an attempt to increase rangeland productivity for livestock and to minimize the (putative) rangeland degradation caused by pikas (Wang et al., 1995; Fan et al., 1999). As highlighted in the Introduction, such control efforts have been extensive and are ongoing. Plateau pikas have already been eradicated from vast expanses of Qinghai's alpine grasslands, and the poisoning continues. Sadly, this control campaign has been executed without an examination of the ecological role of the plateau pika. What have been the effects of this widespread poisoning? Here, we evaluate the role of plateau pikas with relationship to the biodiversity of native birds.

Bird species richness and overall avian abundance were greater on sites that had not been poisoned, compared with poisoned sites. Pika control on the high alpine grasslands of Qinghai province has been pursued actively for approximately 40 years. One of us (ATS) has conducted field studies on the grasslands of Qinghai since 1984 and qualitatively noted vast expanses of habitat virtually devoid of birds. In this study, however, we chose sites that were poisoned relatively recently (within the past five years) to assess the potential for direct impact of poisoning on avian diversity. Within this short timeframe changes in the avian community were dramatic.

An important feature of biodiversity analysis is an assessment of those species that are endemic to a region. In the present study Hume's groundpecker represents a monotypic genus that is found only on the Qinghai-Xizang plateau, and four of six snowfinches (Tibetan, white-rumped, rufous-necked and plain-backed) are plateau endemics (MacKinnon, 2000). Hume's groundpecker and two plateau endemic snowfinch species were observed only on non-poisoned sites, and overall species richness and abundance of these forms were higher on non-poisoned sites. Following the poisoning of plateau pikas, the burrow systems constructed by pikas degrade rapidly, and the avian forms that rely on the burrows for nesting and cover essentially disappear. Thus, the plateau pika clearly plays the role of an allogenic engineer in the alpine meadow ecosystem (sensu Dickman, 1999). But, does this also lead to the classification of the plateau pika as a keystone species?

Controversy concerning the keystone-species concept (Power et al., 1996; Stapp, 1998; Kotliar, 2000; Miller et al., 2000) has led to refinement in the definition of a keystone species. A species is generally considered to play the role of a keystone species if: 1) it exerts a large effect on community structure and function (i.e., high overall importance); 2) these effects are disproportionately large relative to abundance (i.e. high community importance; Power et al. 1996); and 3) it provides unique functions not performed by other species or processes (Kotlier, 2000).

Clearly the plateau pika meets the first of these criteria. In its functional role as an allogenic engineer, the pika provides habitat for many bird species (also lizards; Li, 1989; Smith and Foggin, 1999, 2000). When present at its characteristic high densities, the plateau pika serves as food for most predatory birds and mammals on the alpine grasslands (Schaller, 1998; Smith and Foggin, 1999; 2000). The two raptor species in this study were more abundant on areas where pikas were present than on poisoned sites. Our knowledge of the effects of plateau pikas on ecosystem-level processes, however, are indirect and extrapolated from ecologically similar species (the Daurian pika, O daurica, and North American prairie dogs, genus Cynomys; Smith and Foggin, 1999, 2000). There is a need for controlled experiments on the effect of plateau pikas on ecosystem-level processes such as nutrient cycling, disturbance, rates of erosion, etc.

Application of the second criteria has been more problematic (Kotlier, 2000). Of particular concern is the issue that a species must exert an effect larger than would be predicted by their abundance. Using prairie dogs as an example, Kotlier et al. (1999, 2000) found that the assumption that overall importance was a linear function of abundance was invalid, and determined that community importance was sensitive to levels of abundance and issues of scale. It would be useful to examine these issues for plateau pikas. However, in our controlled examination of the effect on avian diversity and abundance caused by poisoning pika populations, the crucial dynamic was presence or absence of pikas, not gradations in their abundance. Thus, we were unable to test this criterion directly.

The third criterion for the determination of plateau pikas as a keystone species concerns their unique role on the high alpine grassland ecosystem of the Qinghai-Xizang plateau. No other mammalian species provides analogous ecological services. Among herbivorous mammals, native large grazing ungulates have been nearly eliminated from the areas in which we worked (c.f. Foggin, 2000). Similarly, wooly hare (Lepus oiostolus) and Himalayan marmot (Marmota himalayana) populations have been greatly reduced across the plateau; these species are rarely sighted in the project area. Most other species of Ochotona in the region occupy different habitats (rocks, shrubs, or more xeric biomes; c.f. Smith et al., 1990). The Chinese zokor (Myospalax fontanieri) makes burrows, and thus may contribute to aspects of ecosystem-level processes, such as disturbance and nutrient cycling. But, its burrows are closed and cannot function as habitat for Hume's groundpecker or snowfinches. One of us (ATS) has conducted two long-term studies at different sites within the region of the present study (Smith et al., 1986; Wang and Smith, 1988, 1989; Smith and Wang, 1991; Dobson et al., 1998, 2000), and no other small herbivorous mammals have been sighted. The plateau pika is the only small burrowing mammal distributed commonly across the alpine grasslands of the Qinghai-Xizang plateau, and this lack of ecological redundancy contributes to its classification as a keystone species.

The purpose of the keystone-species concept in conservation is to identify those species that, if lost, would trigger a cascading effect of impoverishment in an ecosystem. We have identified the alpine grassland ecosystem of the Qinghai-Xizang plateau as being at a crossroads - the health of the land, its native biodiversity, and its ability to sustain the pastoralist economy are all at risk. We have demonstrated that eradicating populations of plateau pikas contributes to the loss of avian species richness and abundance, and in particular those forms which are most unique to the plateau ecosystem. In addition, ongoing efforts to control pikas may lead to further losses of predators and disruption of ecosystem-level processes (Smith and Foggin, 1999, 2000). Thus, the poisoning of pikas represents one factor contributing to the loss of functionality of the plateau grasslands. The keystone role played by plateau pikas should be considered in any future management measures that involve poisoning in the fragile alpine grassland ecosystem.

Acknowledgements
We thank Robert Hoffmann and Harriet Smith for their conscientious review and comments on the manuscript. Financial support for this project was provided by the Kadoorie Charitable Foundation (Hong Kong). We appreciate the logistical support provided by the Biodiversity Working Group of the China Council for International Cooperation in Environment and Development.

References
Dickman, C. R. (1999) Rodent-ecosystem relationships: a review. In Ecologically-based rodent management (G. Singleton, L. Hinds, H. Leirs and Z. Zhang, eds) pp. 113-133. Canberra, ACT, Australia: Australian Centre for International Agricultural Research.
Dobson, F. S., Smith, A. T. and Wang, X. G. (1998) Social and ecological influences on dispersal and philopatry in the plateau pika (Ochotona curzoniae). Behavioral Ecology 9,622-635.
Dobson, F. S., Smith, A. T. and Wang X. G. (2000) The mating system and gene dynamics of plateau pikas. Behavioural Processes 51,101-110.
Drandui, L. (1996) A comparative study on the animal husbandry of Qinghai and Tibet. China Tibetology 4,31-42.
Ekvall, R. B. (1968) Fields on the hoof. Prospect Heights, Illinois: Waveland Press.
Fan, N., Zhou, W., Wei, W., Wang, Q. and Jiang, Y. (1999) Rodent pest management in the Qinghai-Tibet alpine meadow ecosystem. In Ecologically-based rodent management (G. Singleton, L. Hinds, H. Leirs and Z. Zhang, eds) pp. 285-304. Canberra, ACT, Australia: Australian Centre for International Agricultural Research.
Feng, Z., Cau, G. and Zheng, C. (1986) The mammals of Xizang. Beijing: Science Press.
Goldstein, M. C. and Beall, C. M. (1990) Nomads of western Tibet: the survival of a way of life. Berkeley: University of California Press.
Heywood, V. H., ed. (1995) Global biodiversity assessment. Cambridge: Cambridge University Press.
Kotliar, N. B. (2000) Application of the new keystone-species concept to prairie dogs: how well does it work? Conservation Biology 14,1715-1721.
Kotliar, N. B., Baker, B. W., Whicker, A. D. and Plumb, G. (1999) A critical review of assumptions about the prairie dog as a keystone species. Environmental Management 24, 177-192.
Lang, B. N., Huang, J. S. and Wang, H. Y. (1997) Report on the pasture and livestock survey in Hainan IFAD Project Area. Xining, Qinghai, PRC: International Funds for Agriculture Development.
Li, D. H., (ed). (1989) Economically important vertebrates of Qinghai province. Xining, Qinghai, PRC: Qinghai People's Publishing House.
MacKinnon, J. (2000) A field guide to the birds of China. Oxford: Oxford University Press.
MacKinnon, J., Sha, M., Cheung, C., Carey, G., Xiang, Z, and Melville, D. (1996) A biodiversity review of China. Hong Kong: World Wide Fund for Nature.
Ma, M. (1995) Suggestions for the protection of some pikas. China Nature 2,26.
Meyer de Schauensee, R. (1984) The birds of China. Washington, DC: Smithsonian Institution Press.
Miller, D. J. (1995) Herds on the move: winds of change among pastoralists in the Himalayas and on the Tibetan plateau. Kathmandu, Nepal: International Centre for Integrated Mountain Development.
Peshikov, B. I. (1957) Data on numbers and diet of feathered predators of southeast Transbaikalia. Izv. Irkutsk. Nauchno-Issled. Protivochumnogo. Ins. Sib. Dal'nego Vost. 16,143-153.
Peshikov, B. I. (1967) On the biology of the upland buzzard. Izv. Irkutsk. Nauchno-Issled. Protivochumnogo. Ins. Sib. Dal'nego Vost. 27,167-174.
Power, M. E., Tilman, D., Estes, J. A., Menge, B. A., Bond, W. J., Mills, L. S., Daily, G., Castilla, J. C., Lubchenco, J. and Paine, R. T. (1996) Challenges in the quest for keystones. Bioscience 46,609-620.
Richard, C. (2000) Rangeland policies in the eastern Tibetan plateau. Issues in Mountain Development 2000/4,1-4.
Schaller, G. B. (1985) Wildlife in the middle kingdom. Defenders 60,10-15.
Schaller, G. B. (1998) Wildlife of the Tibetan steppe. Chicago: University of Chicago Press.
Smith, A. T. and Foggin, J. M. (1999) The plateau pika (Ochotona curzoniae) is a keystone species for biodiversity on the Tibetan plateau. Animal Conservation 2,235-240.
Smith, A. T. and Foggin, J. M. (2000) The plateau pika (Ochotona curzoniae): a keystone species for biodiversity on the Tibetan plateau. In Formation and evolution, environmental changes and sustainable development on the Tibetan plateau (Z. Du, ed) pp.387-396. Beijing: Academy Press.
Smith, A. T., Formozov, N. A., Hoffmann, R. S., Zheng, C. L. and Erbajeva, M. A. (1990) The pikas. In: Rabbits, hares and pikas: status survey and conservation action plan (J. A. Chapman and J. E. C. Flux, eds) pp. 14-60. Gland, Switzerland: International Union for the Conservation of Nature.
Smith, A. T., Smith, H. J., Wang, X. G., Yin, X. C. and Liang, J. (1986) Social behavior of the steppe-dwelling black-lipped pika. National Geographic Research 2,57-74.
Smith, A. T. and Wang, X. G. (1991) Social relationships of adult black-lipped pikas (Ochotona curzoniae). Journal of Mammalogy 72,231-247.
Stapp, P. (1998) A reevaluation of the role of prairie dogs in great plains ecosystems. Conservation Biology 12,1253-1259.
Wang, Q. J., Lang, B. N., Wu, Y. S., Jing, Z. C. and Zhang Y. (1995) Report on the study of the cause of formation of black sand regions in the Qinghai-Xizang plateau grassland and possible preventative measures. Beijing: Chinese Academy of Sciences.
Wang, X. G. and Smith, A. T. (1988) On the natural winter mortality of the plateau pika (Ochotona curzoniae). Acta Theriologica Sinica 8,152-156.
Wang, X. G. and Smith, A. T. (1989) Studies on the mating system in plateau pikas (Ochotona curzoniae). Acta Theriologica Sinica 9,210-215.
Zhao, S. (1992) The Qinghai-Xizang (Tibetan) plateau: its physical environment and economic development. Chinese Journal of Arid Land Research 5, 203-222.