S35.1: Sustainable management to prevent extinction: The case of the Sichuan Partridge Arborophila rufipectus

1School of Biological & Earth Sciences, Liverpool John Moores University, James Parsons Building, Byrom Street, Liverpool, L3 3AF, UK, e-mail s.d.dowell@livjm.ac.uk; ²Wildlife Division, Sichuan Forestry Department, 15, Renminbeilu, Chengdu, Sichuan Province, 610081, China, e-mail habitat@public.cd.sc.cn; ³6 Connaught Road, Norwich, Norfolk, NR2 3BP, UK, e-mail rodney.martins@virgin.net; ⁴Biosphere Consultants, 6 Waldeck Road, Norwich, Norfolk, NR4 4PG, UK, e-mail RobSRW@aol.com

The Sichuan Partridge Arborophila rufipectus (formerly known as the Sichuan Hill-partridge) is recognised as one of the most globally threatened Galliforms (Collar et al. 1994; McGowan et al. 1995) and it has been the subject of conservation attention since the late 1980s. It is known only from a limited area of southern Sichuan in China and possibly northern Yunnan where it occupies subtropical deciduous/evergreen broadleaf forest between 1000 and 2235 m in altitude (Cheng et al. 1978; Johnsgard 1988; King and Li 1988; McGowan et al. 1995; Dowell et al. 1997; Dai et al. 1998). This forest type has a diverse bird community and comprises the western part of the Chinese subtropical forests Endemic Bird Area (Stattersfield et al. 1998). It is therefore part of a globally important network of areas of high bird diversity which represent priority areas for conservation attention (Stattersfield et al. 1998). An estimated 90% of the original deciduous/evergreen broadleaf forest cover in Central China has been lost (Mackinnon et al. 1996). Today, the habitat is severely threatened by commercial logging and all primary forest below 1000 m in altitude has been destroyed (King & Li 1988). Recent concern for protection of watersheds and soil resources has led to a more sustainable forestry policy in this part of China, with clear-felling being followed by widespread replanting in some areas (Cao 1989). The effects of this on the Sichuan Partridge and other endemic bird species are the subject of the investigation reported here.

The Sichuan Partridge was first described by Boulton (1932) who collected the type specimen, a male, in Ganluo county, south central Sichuan. A further specimen was collected by Tang Rui-chang in 1945 from Qingming Shan in Mabian County, close to Dianjiping village, in an area that is now agricultural land (specimen held at Wuhan University). Little additional information about the species emerged until the 1960s when studies by Professor Li Gui-yuan of the Agricultural University in Ya'an provided more information on the habitat; dense broadleaf forest between 1300 and 1800 m in altitude (Li et al, 1974). King & Li (1988) recorded up to 11 calling males per day during an expedition to Huanglianshan, Mabian county in spring 1986. They expressed alarm at the rapidity with which primary forest was being logged at this site to make way for commercial coniferous plantations and predicted the extinction of the Sichuan Partridge within 40 years (King & Li 1988). Further surveys in Huanglian Shan were undertaken between 1988 and 1992 when three nests containing between three and seven white eggs were found in recently felled areas (Li & Zhang 1992; Xu et al, 1994). It is not clear whether these were abandoned or in use. Xu et al. (1994) recorded nine calling males from 12 km² of secondary and plantation forest and 11 in 4 km² of primary forest. Although based on small sample sizes and relying only upon records of calling males, these data do indicate the importance of primary forest habitat whilst also suggesting that Sichuan Partridges may survive in secondary forest at apparently lower densities.

Concern for the species led to visits from gamebird biologists from Beijing and the UK in the early 1990s (Dowell 1995), culminating in the study reported here which was carried out with the support of the World Pheasant Association and the Sichuan Forestry Department. The aims were to determine the range and population density of the species and to investigate its ability to adapt to forest change in the region. In addition, other broadleaf forest birds were studied, especially those species identified as restricted-range endemics by BirdLife International (Stattersfield et al. 1998) in relation to their adaptability to forestry practices.

The Sichuan Partridge occupies a highly restricted range centred around the Daliang Shan and its environs in south central Sichuan and northern Yunnan. Ornithologically this is a relatively poorly studied area but the low and moderate altitude (c.1000 to 2350 m) broadleaf forests within the region are known to contain five other restricted-range bird species (Stattersfield et al. 1998) although these are also known to occur at other sites to the north-west and/or to the east. These are Emei Shan Liocichla Liocichla omeiensis, Gold-fronted Fulvetta Alcippe variegaticeps, Red-winged Laughing-thrush Garrulax formosus, Emei Leaf-warbler Phylloscopus emeiensis and Silver Oriole Oriolus mellianus.

Most of the Daliang Shan area is mountainous, rising to over 4000 m on the highest ridges in the west, with deep valleys and steep slopes. Gradients typically vary between 25 and 50^o. Drainage is extensive with all rivers flowing eventually into the Yangtze (or Jin Sha Jiang as it is called in this part of China). Climate is seasonal and continental with warm, wet summers and cool, damp winters. Recorded mean annual precipitation in valleys ranges between 1200 and 1400 mm and the annual temperature range is -4 to +30 ^oC (Anon. 1981). Generally, higher altitudes experience a substantially colder and wetter climate than other parts of the region but substantial variation in local climatic conditions occurs, reflecting the changeable nature of altitude, aspect, relief and general topography.

Forest cover was originally extensive (Yang & Li 1992) but exploitation for timber and clearance to provide land for agriculture has radically diminished and fragmented the original vegetation cover. Extensive forest no longer exists below 1000 m. Between 1000 and 2000 to 2350 m the forest may be broadly categorised as subtropical broadleaf with an admixture of deciduous and broadleaf evergreen species and a patchy bamboo understorey. Between 2000 to 2350 and 2500 m there is a narrow zone of temperate mixed forest blending to native coniferous forest above 2500 m. The tree-line typically lies between 3000 and 3500 m above which is alpine vegetation. The Sichuan Partridge has not been recorded in mixed or coniferous forest above 2335 m, despite searches for the species at commensurate altitudes (e.g. King & Li 1988).

The Daliang Shan is located at the northern edge of the area inhabited by the Yi tribal minority. The Yi comprise between 15 and 43% of the human population in the counties of E'bian, Leibo and Mabian (source: Sichuan Forestry Department). The Yi are primarily farmers and hunters, although most traditional quarry species (including the Sichuan Partridge) are now theoretically protected by legislation. The Yi have a unique language and a distinctive and colourful culture. Traditional uses of the forest by the Yi and others include hunting, collection of medicinal plants, bamboo shoot collection, livestock herding, cutting of firewood and wood logging for house and furniture construction. Of these, only the latter continues extensively today, although the extent to which the forest is used for other purposes is poorly studied.

Although forestry is one of the main local industries, there is some history of conservation. Dafending Nature Reserve in Mabian and Meigu counties was established in 1978 to protect a remnant population of Giant Panda Ailuropoda melanoleuca (Wang & Hu 1989). Most of this reserve is above 2000 m and therefore within the coniferous forest zone. Elsewhere in the region there is much evidence of replanting after clear felling and there are proposals to establish forest parks in key areas to manage forests for visitors. These were investigated during the course of the fieldwork.

Table 1 and Fig.1 give all forest tracts in the study area that were surveyed for Sichuan Partridge during field surveys in 1996, 1997 and 1998.

Extensive fieldwork was carried out in May and June 1996 (by Dai Bo) and in April and May 1997 (all authors). In May and June 1998 a detailed survey of Dafending Nature Reserve in Mabian and Meigu counties was carried out (by Dai Bo).

The Sichuan Partridge is elusive and any census of this species must therefore rely on detecting vocalisations. Two types of far-carrying call are given by the male, mainly during the breeding season between March and June (Li et al. 1974, Li & Zhang 1992). Calls are most frequently delivered during the four hours after dawn and from about 1600 until just before dusk, although they may also be made at other times. There are no records of calls at night. Little is known about the function of these calls although they are assumed to be associated with territory establishment and maintenance and/or mate attraction (Li et al. 1974, Li & Zhang 1992).

Although Sichuan Partridge calls are fairly distinctive if heard clearly, the Spotted Laughing-thrush Garrulax ocellatus, which occurs in some of the same forest areas, sometimes gives similar calls which can cause confusion. Therefore, some caution has been applied to the results of the survey. For example, where records of calls represent a potential range extension (in Yunnan), we propose that a further visit be made to sites involved to seek confirmation through sight records, signs, local reports and tape recordings.

In each year of the study up to ten 1000 m line transects were selected in each forest tract. Dense forest combined with steep slopes meant that most of these transects were sited along established pathways, forest roads or ridge lines. Transect surveys were undertaken within primary, natural secondary and plantation forest. All transect data were collected between 0530 and 0930 or between 1600 and 2000 (Greenwich +8h) in April, May and June. None were undertaken during periods of high winds and/or heavy rain. Observers walked each transect slowly, pausing for three minutes after every 50 m (approximately 100 paces) to imitate the ascending whistle call 15 times and listen for calls. The distance from the transect line of any Sichuan Partridge call detected was estimated and recorded along with the time, frequency of call and forest type from which it was thought to emanate.

Data from line transects were used to calculate density estimates of calling male Sichuan Partridges using the linear model of the two belt transect method (see Bibby et al. 1992). The relevant equation is:

D = 10nK / L

where

K = (1 - ü (1-n₁/n)) / w

Where D is density of calling males per ha; n is the total number of birds recorded on all transects; n₁ is the number recorded within the inner belt; w is the belt width (taken as 300 m); L is the total length of transect in kms; and K is a constant calculated using the formula shown.

Abundance and density of other forest birds were assessed between 12 April and 27 May 1997 through point counts. Points were selected 500 m apart along transect lines through continuous forest. Transect lines were located in primary, secondary and both plantation forest types. Point counts were carried out by observers working in pairs. On arrival at the point, observers stood silently for a two minute habituation period, recording only birds that appeared to flee from the area as a result of the observers' presence. Bird observations and calls were then recorded for ten minutes onto an appropriate data sheet. For each bird species record the estimated distance from the point to the bird was recorded. Care was taken to avoid recording the same individual more than once. If a bird moved to a new location during the point count period, its initial location was recorded as the distance from the point. Birds flying over the point during the count were excluded.

The accuracy of point count data is dependent upon the accuracy of observer judgement and identification skill. All four observers undertook trial point counts and habitat surveys together at the beginning of the field season in order to develop consistency. Point count data sheets were monitored periodically throughout the field season.

Analysis of point count data using the two belt method of density estimation for selected species (Bibby et al. 1992) is presented. The relevant equation is:

Density = log_e(n/n₂) x n/m(p r²)

Where n is the total number of individuals counted; n₂ is the number beyond the fixed radius; m is the total number of point counts; and r is the fixed radius measured in metres.

The relevant forestry bureau was visited in each county and senior officials were interviewed. The following questions were asked:

(1) Over what time scale do you plan to clear-fell the remaining primary forest in your area?

(2) How important is forestry to the local economy in terms of local employment?

(3) To what use is previously forested land to be put after clear felling?

(4) Do you have a replanting strategy, will you harvest any of this replanted forest in future years and if so, after what time period?

(5) Are there any plans to protect any forest within your area for wildlife, landscape value or tourism?

The answers to these questions were used to collect information on forestry policies in the region.

During extensive fieldwork in 1996, 72 Sichuan Partridges were recorded. In 1997 when field time was divided between carrying out point counts for other bird species and censusing Sichuan Partridges, 37 individuals were recorded. A more intensive survey of the Dafending area in 1998 revealed only two calling male Sichuan Partridges, both inside the Dafending-Mabian Nature Reserve. The only sightings were at Heitaoping south of the Xining river in 1996 and in the Yangsiba area in 1997, all in primary forest. All records from secondary and replanted broadleaf forest were of calls that were heard only. In addition to records from within the known range, calls sounding like Sichuan Partridges were heard from four places in the Ershisigeng forest area on the Suijiang/Yongshan border in northern Yunnan province (site 11 on Fig. 1). These need to be confirmed with sightings.

The mean density of calling males estimated from all transect surveys was 0.48 + 0.06 km^-2 in 1996 and 0.24 + 0.16 per km^-2 in 1997 (Table 2). Numbers recorded in 1998 were too low for density estimates to be calculated. Density estimates from primary forest were not significantly higher than those in secondary and replanted broadleaf forest in any year (1996, c ² = 0.08, df = 1, P>0.75; 1997, c ² = 1.83, df = 1, P>0.10; c ² tests were calculated using density of calling males per 100 km²; see Table 2). The species was not recorded from coniferous plantations. It should be emphasised that all records of Sichuan Partridge on transect surveys were of calling birds only and should therefore be treated with some caution in the light of possible confusion with the calls of Spotted Laughing-thrush.

All records of Sichuan Partridge were from broadleaf forest within the elevational range 1100 to 2235 m. Most forest is located within areas with extremely rugged topography and gradients ranging from 25 to 70^o but typically falling within the range 25 to 40^o. The species is not necessarily restricted to steep slopes however. At Banlixi ridge in the Yangsiba area, numerous signs detected by local guides (e.g. areas of disturbed leaf litter, dust-bathing areas), combined with sightings at the same location indicate that broad, rounded ridge crests with extensive areas of flatter forest floor are perhaps favoured where available. Depth of leaf litter and frequency of direct sunlight penetrating to the forest floor are perhaps maximal under such circumstances.

Most individuals were recorded from undisturbed primary forest and all sightings were in this habitat. The density estimates suggest that the species may be able to survive in replanted and secondary forest at comparable densities to those in primary forest, though it must be emphasised that all records from secondary and replanted forest relate to calls only. Almost all Sichuan Partridges heard calling from secondary and replanted forest were in areas of at least 15 years growth and within 1 km of patches of primary forest. There may be a relationship between the recolonisation of secondary forest and the proximity to primary forest source areas.

Where call locations could be visited, characteristics of the forest floor were remarkably similar. Five locations from which calls were heard or birds seen were visited, one in undisturbed primary, two in disturbed primary, one in secondary and one in replanted broadleaf forest. All of these had an open forest floor with sparse bamboo growth and a damp leaf litter layer typically 5 to 10 cm in depth. The canopy was closed, providing good cover.

A total of 204 point counts were carried out during the survey period in 1997, within primary and secondary forest and broadleaf and coniferous plantations. Population density estimates have been calculated using the two-belt method for three of the five other restricted-range bird species recorded and these are given in Table 3. In all three cases, density estimates in secondary/replanted forest were two or more times higher than those in primary forest. Apart from one Red-winged Laughing-thrush, no records of restricted-range birds were obtained from coniferous plantations. Insufficient numbers of Gold-fronted Fulvetta and Emei Shan Liocichla were recorded on point counts for the calculation of density estimates, although both these species were recorded in both primary and secondary forest. Emei Shan Liocichla appears to prefer early successional stages and bamboo thickets and therefore favours regenerating clear-fell. Between 19 and 26 Gold-fronted fulvettas were recorded, all in small stream valleys within 50 m of running water within an altitudinal range of 1365 to 1595 m. The majority of records were from primary forest. Those in secondary forest were in mature patches over 25 years old.

The main threat to forest bird populations in the region is habitat destruction through clear-felling of primary forest. There are currently over 20 state-owned and collective forestry enterprises logging the area (Wang 1994). Forestry policy is formulated within provincial (Sichuan and Yunnan) administrations but is applied and realised at county level. Primary forest at all elevations (broadleaf and coniferous) is harvested by clear-felling or selective logging followed by clear-felling. There is a general policy of leaving a strip of primary forest up to 50 m either side of major ridge crests, though this is not always adhered to. Primary forest on the steepest slopes is also left due to difficulty of extraction of the timber. In some places, local laws prevent the extraction of timber from the steepest slopes in the interests of prevention of erosion. Apart from these areas, all primary forest outside protected areas will be clear-felled within the next 20 years according to local forestry plans.

Some of the clear-fell land is converted to agricultural uses and local people have also encroached into the forest, clear-felling and burning forest to release the land for grazing and growing crops. Although this latter practice is illegal, it is a serious problem, especially where the land is farmed for only one or two years then abandoned in favour of further clearances.

Despite this, much of the clear-fell is replanted, either with non-native conifers or with a small selection of quick-growing native broadleaf species obtained from the original forest and grown in nurseries at the local Forestry Farm Headquarters. One to two year old saplings are planted and kept free of competing ‘weeds’ (including bamboo) for up to three years, after which the plantations are left to nature. After 15 to 20 years the broadleaf plantations resemble natural secondary forest although the diversity of tree species is lower. Some of the replanted forest will be harvested for timber after 30 to 50 years. The remainder will protect mountain sides from soil erosion and prevent excess run-off that could lead to flooding further downstream. There are local variations in forestry policy across the study area and these are detailed in Table 4.

Bamboo shoot collection for sale to the restaurant market is widespread but seasonal. The main seasons are between April and May and in early autumn when several hundred people may be temporarily based in the forest, entering it daily in search of young bamboo shoots. This is therefore an activity which creates substantial and regular disturbance, particularly where it occurs on a large scale and is uncontrolled. Evidence of bamboo shoot collection was found within all forest tracts visited though the density of collectors varied. The highest density was at Daxuecao in Yongshan county where, in early May 1997, several hundred people were collecting, living in makeshift shelters in a temporary settlement along the road. Evidence of earlier camps, including cleared areas contaminated with litter and human waste, were found at higher altitudes. Discussions with local officials established that over 1000 permits are issued annually to local people to allow them to collect bamboo shoots here during the period between 7 April and 7 May. Permits cost 1 Yuan but perhaps as many as two thirds of the bamboo shoot collectors do not purchase one. There could be as many as 3000 people collecting bamboo shoots in this 91 km² area of forest in late April and early May. Most of these are concentrated near the road which bisects the area and along adjacent paths.

Hunting of gamebirds is traditional, especially amongst the indigenous Yi people. Although hunting without a permit is illegal it is difficult to control and police in such mountainous terrain. Most hunters seek the larger quarry such as Temminck’s Tragopan Tragopan temmincki, but Sichuan Partridges are also trapped. The effect of hunting on gamebird populations is not known.

Some local farmers graze their livestock on the forest floor and their cattle and goats will browse secondary growth and understorey shrubs. This may restrict forest regeneration and damage the leaf litter where the Sichuan Partridge feeds. In most places, however, the steep gradient prevents significant incursion into the forest by grazing animals.

Local people are known to use forest plants for medicinal purposes and timber resources for fuel and building materials, but the extent of these activities requires further study.

The Sichuan Partridge was recorded at very low densities even in extensive tracts of optimal habitat (primary broadleaf forest). Where primary forest is replaced with native broadleaf plantations it can survive at similarly low densities, although there is some doubt as to whether its survival in these secondary plantations relies on the proximity of patches of primary forest. It is impossible to give an accurate population estimate based on density data, though a crude measure can be calculated by multiplying mean calling male density by total forest cover. Using the density estimates obtained during 1996 and 1997, this gives a range of 430 to 861 calling males for the ten forest tracts surveyed in 1996 and 1997. The surveys in these two years covered more than half of the forest areas within the species’ range. If the species is monogamous and each calling male represents a breeding pair, the total effective population size is unlikely to be more than 1500 pairs. It must be emphasised that this figure is based on crude evidence and the assumptions outlined, but it suggests that claims that the Sichuan Partridge is one of the China’s and indeed the world’s most endangered Galliforms (King & Li 1988; Dowell 1995) are justified. Destruction of primary broadleaf forest on a large scale has resulted in severe habitat fragmentation and consequent isolation of individual populations and accelerated the decline giving cause for considerable concern.

Of the five other restricted-range bird species occurring in the same habitat, we have presented evidence that three, the Red-winged Laughing-thrush, Emei Leaf Warbler and the Silver Oriole may actually occur at higher densities in secondary and replanted broadleaf forest than in primary forest. A fourth, Emei Shan Liocichla was recorded mainly in early successional habitats, the extent of which have increased after clear-felling. Only Gold-fronted Fulvetta appears to be dependent mainly on primary forest although it was also recorded in relatively mature broadleaf plantations around streams.

The future of the Sichuan Partridge and the other threatened birds which share its habitat would appear to depend upon the development of sustainable forestry management policies. It is essential that some primary forest is protected in order to act as a source area for native species. The only undisturbed primary broadleaf forest currently enjoying protection in the region is a small area in the Dafending Nature Reserve (Mackinnon et al. 1996). Surveys in 1998 recorded only two calling male Sichuan Partridges and very few records of other restricted-range species, so there is a need for other protected forests to be established. Proposals for Forest Parks in Suijiang and E’bian counties (see Table 4.) may provide opportunities for this in the future, but we propose that a new Nature Reserve be established covering primary broadleaf forest centred around the Banlixi ridge in the Yangsiba area. This was the area where the highest numbers of Sichuan Partridge were recorded and it contains some of the most undisturbed habitat.

The importance of maintaining forest cover to protect soil and water resources in areas where commercial logging activities are taking place is recognised by the Forestry Authorities in China (Cao 1989; Yang & Li 1992; Wang 1994). This has led to the policies of retaining primary forest patches on ridges and steep slopes and replanting elsewhere (Cao 1989). Where native broadleafs are replanted to replace the primary subtropical broadleaf cover, this can lead to the development of valuable secondary habitat. Our research has shown this to be suitable for at least four of the six restricted-range bird species in the region, including the Sichuan Partridge, although close proximity of primary forest fragments may be necessary for this to be successful. Where non-native conifers are used in the plantations the resulting habitat is unsuitable for most of the original bird species. This practice should therefore be ended in favour of using native broadleaf saplings. In addition, public education of local farmers is needed to reduce uncontrolled encroachment of agricultural activities onto forest land. Further research is planned to investigate the Sichuan Partridge’s detailed ecological requirements using radio telemetry techniques. It is hoped that the results of this research can be used to fine-tune the sustainable forestry strategy in order to enhance its wildlife benefits.

Unlike other areas where gamebirds represent significant flagship species for biodiversity conservation, allowing harvesting of the Sichuan Partridge to provide incentives for conservation is not an appropriate strategy. The low densities recorded would suggest that the species is on the brink of extinction and uncontrolled hunting would only contribute to its decline. As forestry land is owned by the state in China there are no incentives for private landowners to encourage gamebirds for commercial hunting. One area that does need investigation, however, is the use that is made of forest resources by local and indigenous people. The use of certain plant species for medicinal and/or culinary purposes could provide financial incentives for local habitat conservation efforts. This has been successful in Pakistan where conservation of primary forest resources in the Palas valley was made possible when local Palasi tribesmen found a market for the morel mushroom Morchella spp., thus enabling the protection of the most important area for the endangered Western Tragopan Tragopan melanocaphalus (Duke 1993). In the future, conservationists aiming to conserve the Sichuan Partridge and its forest habitat will need a clear understanding of the importance of forest resources in all areas of the local economy.

Survey work on the Sichuan Partridge and other restricted-range species within its habitat has received funding from the following individuals and organisations: Mr. Abdul Jalil, British Airways Assisting Conservation, The British Ecological Society, The British Ornithologists Union, The Oriental Bird Club, Stiftung Avifauna Protecta, The World Pheasant Association, and The Zoological Society for the Conservation of Species and Populations.

The authors are grateful to Mr. Wang Wei and Mr. Yan Xun of the Chinese Ministry of Forestry and Mr. Hu Tie-qing and Mr. Deng Xiang-sui of the Sichuan Forestry Department for granting permission to carry out the work. In China the project’s counterpart in Beijing, Mr. He Fen-qi of the Institute of Zoology, Academia Sinica gave invaluable advice, helped with logistical arrangements and was of paramount significance in enabling the full realisation of project goals. Ms. Xiong Bairong, Mr. Yuan Shijun and Mr. Wang Ji-wei of the Sichuan Forestry Department helped to arrange visas and permits. The directors of local Forestry Bureaux in Mabian, Leibo, E'bian, Suijiang and Yongshan counties were generous hosts and provided invaluable support and assistance throughout fieldwork.

We are grateful to all local officials, guides and forestry workers and to our drivers within the study area for their kind and valuable assistance during our fieldwork, especially Tao Yingao, Wang Wende, Li Bo-liang , Li Guang-yan and Hu Ya-qiu.

We would like to thank Ms. Sheila Ripper for preparing the figure and Robin Little for comments and help with earlier drafts of the manuscript.

Anon. 1981. Atlas of Sichuan Province. Chengdu, Sichuan, China; Sichuan Provincial Mapping Bureau.

Bibby, C.J. , Burgess, N.D. & Hill, D.A. 1992. Bird Census Techniques. London, UK; Academic Press.

Boulton, R. 1932. A new species of tree partridge from Szechwan, China. Proceedings of the Biological Society of Washington 45: 235-236.

Cao Zheng-qi (ed.) 1989. A Corpus of Forestry Policies and Statutes (1979-1989). (in Chinese) Chengdu, Sichuan, China; Sichuan People’s Publishing House.

Cheng Tso-hsin, Yao-Kuang, T., Tai-chung, L., Chanzhu, T., Gui-Jun, B. & Fu-Lai, L. 1978. Fauna Sinica, series Vertebrata. Aves Vol.4:Galliformes. Beijing, China; Science Press.

Collar, N.J., Crosby, M.J. & Stattersfield, A.J. 1994. Birds to Watch 2: The World List of Threatened Birds. Cambridge, UK: BirdLife International.

Dai, B., Dowell, S.D., Martins, R.P. & Williams, R.S.R. 1998. Conservation status of the Sichuan Hill-partridge Arborophila rufipectus in China. Bird Conservation International 9: 349-359.

Dowell, S.D. 1995. Current status and future prospects for the Sichuan hill-partridge. WPA News 49: 6-13.

Dowell, S.D., Dai, B., Martins, R.P. & Williams, R.S.R. 1997. Sichuan Hill-partridge Conservation Project: Report on the 1997 Field Season. Unpublished report, Reading, U.K.; World Pheasant Association.

Duke, G. 1993. A participatory approach to conservation in Pakistan: Safeguarding the tragopans and forests of the Palas valley, Indus Kohistan. In: Jenkins, D. (ed.) Pheasants in Asia 1992. Reading UK; World Pheasant Association: ii.- vi.

Johnsgard, P.A. 1988. The Quails, Partridges and Francolins of the World. Oxford, UK; Oxford University Press.

King, B. & Li Gui-yuan. 1988. China's most endangered galliform. Oryx 22: 216-217.

Li Gui-yuan., Liu Liang-cai., Zhang Rui-yun. & Zhang Qing-mao. 1974. Discovery of the female of the Sichuan hill-partridge (Arborophila rufipectus). (in Chinese) Acta Zoologica Sinica 20: 421-422.

Li Gui-yuan & Zhang Qing-mao 1992. The nests, eggs and calls of the Sichuan hill-partridge (Arborophila rufipectus). (in Chinese) Acta Sinica Zoologica 38: 108-109.

Mackinnon, J., Meng, S., Cheung, C., Carey, G., Xiang, Z. & Melville, D. 1996. A Biodiversity Review of China. Hong Kong; WWF International, China Programme.

McGowan, P.J.K., Dowell, S.D., Aebischer, N.J. & Carroll, J.P. 1995. Partridges, Quails, Francolins, Guineafowl and Snowcocks: An Action Plan for their Conservation. Gland, Switzerland: IUCN, The World Conservation Union.

Stattersfield, A.J., Crosby, M.J., Long, A.J. & Wege, D.C. 1998. Endemic Bird Areas of the World: Priorities for Biodiversity Conservation. Cambridge, UK; BirdLife International.

Wang Ji-gui (ed.) 1994. History of Sichuan Forestry. (in Chinese) Chengdu, China; Sichuan Sciences and Technology Publishing House

Wang Menghu & Hu Tieqin (eds) 1989. National Conservation Management Plan for the Giant Panda and its Habitat. Hong Kong; Joint Report of the Chinese Ministry of Forestry and WWF.

Xu Zhao-hui., Mei Wen-zheng, Zhang Gang, Chen Xin-hua & Zeng Zhao-qun 1994. Observations on the wintering habits of the Sichuan hill-partridge. (in Chinese with English abstract) Chinese Journal of Zoology 29: 21-23.

Yang Yu-po & Li Cheng-biao (eds) 1992. Forests in Sichuan. (in Chinese). Beijing, China; China Forestry Publishing House.