Conflict between cultural development and wildlife conservation: A potential threat to Reeves's pheasant (Syrmaticus reevesii)

Reeves's pheasant feathers are used to make headgears for the Chinese opera—Xiqu; however, this posed a considerable threat to the bird's population before it was banned from hunting/trade. It is unclear whether Xiqu‐troupes currently use feathers from wild or captive breeding; therefore, we investigated their source and scale of feather utilization. Through interviews, we clarified the current status of feather‐use by state‐run troupes. We conducted a cluster analysis and established a “random forest” model using stable isotopes as variables to distinguish sources of feathers from Xiqu‐troupes. Annual demand for feathers in China was estimated as 1159.4 pairs. Isotope analysis revealed that feathers used by Xiqu‐troupes are likely derived from wild sources. We suggest strengthening the supervision of acquisition of Reeves's pheasant feathers and harvesting feathers through captive breeding. Stable isotope analysis can be an effective method for identifying the source of tail feathers, which will support law enforcement efforts.


INTRODUCTION
For thousands of years, people have used and traded wild animals and their products; however, unsustainable trade can negatively impact source animal populations (Hughes, 2003;Rosen & Smith, 2010).Although the trade of wildlife occurs globally, for use in traditional medicine, decoration, rituals, magical spells, and cultural practices, details about these practices and their impacts on many species are lacking (Gnanaolivu et al., 2022).Addressing these knowledge gaps is essential to improve the initiatives aimed at reducing unsustainable wildlife trade, which are often poorly targeted and, therefore, have limited potential success (Phelps et al., 2016).Some ethnic communities incorporate feathers into traditional attire, which results in unsustainable capture and trade of wild bird populations (Adeola, 1992;Ezenwa et al., 2019).In contrast to the bird trade, the threat of the feather trade to wild populations is easily overlooked (Allen, 1886;Assou et al., 2021).There is a long history of using the tail LI et al. feathers of Reeves's pheasant in ornaments for traditional Chinese opera-Xiqu (Song, 2002).The ornaments, called Lingzi, are made of a pair of tail feathers and are used as accessory for helmets, caps, and hats; Lingzi have been used in pairs as a tool for dancing to express the feelings of a character (Liu, 1993).Lingzi has been worn by many heroic characters, making the feathers of Reeves's pheasant a symbol of courage and glory, as well as a cultural symbol in China.Before Reeves's pheasant were placed under special state protection to prohibit hunting and trade, hunting for the utilization of feathers in Xiqu was considered an important threat to Reeves's pheasant populations, as the incomplete interview survey showed that troupes had a general demand for these feathers (Xu et al., 1996).Although Reeves's pheasant can be legally bred, it is unknown whether the feathers used to make Lingzi come from wild individuals or captive-bred populations.Unlike the live animals or large-scale animal products that are prioritized in the protected list, the feather trade in China has not been managed in a traceable manner, making it difficult to effectively determine the legitimacy of feathers in the actual enforcement process.Evaluating the demand for these tail feathers, through consumer surveys, will greatly aid in the development of management and conservation strategies for Reeves's pheasant (Jepson & Ladle, 2005;Nijman et al., 2017).Therefore, our study aimed to further understand the source and market demand of Reeves's pheasant tail feathers used in Xiqu costumes and provide recommendations for the conservation and sustainable use of this species in China.
Naturally occurring stable isotopes of several elements are important tracers in wildlife ecology for the identification of origins in migration research, investigation of trophic relationships, and in the forensic determination of illegally harvested or captive fauna (Jiguet et al., 2019).For instance, stable isotopes are used for examining the origins of ivory and rhinoceros' horns (Amin et al., 2003;Baker et al., 2020).The isotopic characteristics of the food web are reflected in the tissues of organisms, which might vary spatially owing to various biochemical processes.In birds, the isotope values of feather tissues remain fixed after growth, thereby reflecting the environment in which that feather, and subsequently that bird, was grown (Hobson, 2005;Hobson et al., 2012).Isotope analysis provides a constant and long-term signal on the species' diet, making it rather difficult for captive-breeding farms to rapidly acquire isotopic signatures of target species via alteration of food (Rosenblatt & Heithaus, 2013;Warne et al., 2010).Therefore, in this study, we tested the hypothesis that isotopic signatures of carbon and nitrogen substantially differ between feathers of wild and captive-bred Reeves's pheasant, to enable the use of isotope analysis as a method to distinguish between feather sources.

Sample collection and preparation
To determine which Chinese opera use Lingzi, we referred to the National Xiqu Census Report (Rao & Wang, 2020) and watched the play videos (particular stories performed by the troupes) of all operas on Youku, Bilibili, and Douyin video platforms.The survey report only covered the Chinese mainland; Hong Kong, Macao, and Taiwan were not included in the scope of the survey.We collected information on the distribution of Reeves's pheasant from peer-reviewed publications (Lu, 1991;Wu et al., 1991;Zheng, 1978;Zheng, 2015;Zhou et al., 2015).The birthplace of operas using Lingzi and the distribution of Reeves's pheasant were compared.
We conducted semi-directive (n = 12) and structured interviews (n = 107) from June 2021 to June 2022 to collect data on the utilization of Lingzi in China.As the first author had studied in a Xiqu Academy and developed a friendly relationship with the Xiqu practitioners, she conducted semi-directive interviews with the costume managers of the academy and troupes to explore the practitioners' perspectives on the utilization and trade of Lingzi.Based on this, we devised structured questions on inventory, annual loss, and length of Lingzi for interviewing other troupes.We had difficulty in obtaining the definite length of the feathers because not all respondents could tell the exact length measurements; therefore, for the purpose of this study, the lengths were represented in 10-cm intervals.We selected representative provinces in the eastern, central, and western parts of China and contacted state-run troupes, listed in the National Xiqu Census Report (Rao & Wang, 2020), performing plays using Lingzi (n = 324) in these provinces.Some of the troupes refused to provide information or give complete information about the subject researched.Before every interview, we introduced ourselves and explained the nature and objective of the research; interviews were conducted only after obtaining informed consent.Ethical approval for this study was obtained from the Ethics Committee of Beijing Forestry University.
To determine the origin of the feathers, we conducted isotope analysis on the tail feathers of Reeves's pheasant, collected from three sources.In total, 44 damaged feathers on Lingzi were collected from 13 troupes; 44 molted feathers from captive-bred Reeves's pheasant from six farms and zoos; and an additional 30 feathers were obtained, which included feathers that were confiscated by the forest police and molted feathers from the wild during 2020−2022.
Samples were prepared by cutting 10 cm strips from the tip of each feather.The feathers were cleaned using a chloroform: methanol (2:1) solvent to remove any contaminating debris or oils and heated at 60 • C in an oven until dry (Paritte & Kelly, 2009).Stable carbon ( 13 C) and nitrogen ( 15 N) isotopes were analyzed at the Huake Jingxin Stable Isotope Laboratory (http://www.isotopelab.com),Shenzhen, China.A Thermo Fisher EA-HT Elemental Analyzer coupled to a DeltaV mass spectrometer was used for the analysis.

Statistical analyses
All statistical analyses were performed in R (v.3.5.3;R Core Team, 2019).To compare the purchasing power and demand of different levels of troupes toward Lingzi, inventory and annual supplement disparities between provincial, city, and county troupes were tested using Kruskal-Wallis H test.To distinguish the source of feathers used on Lingzi, δ 13 C and δ 15 N isotopes were used as clustering indicators for hierarchical cluster analysis of feather all samples, including samples from field collection, captivity, and troupes.To determine the most ideal clustering form, the number of optimal clustering was voted using the "NbClust" package in R. We set isotope and feather sources (from field collection or captivity) as predictive and response variables, respectively, in a "random forest" model in R to predict the source of feathers on Lingzi.Binary response variables were replaced by "1" and "0," where "1" represents feathers originating from captivity, and "0" represents feathers originating from field collection.Cross-validation function of the "random forest" package of R was used to automatically calculate the accuracy of the model.

Status of use of Lingzi
Currently, there are 348 operas on the Chinese mainland, of which 164 (47.1%) have staged plays using Lingzi.Four operas have been placed on the United Nations Educational, Scientific and Cultural Organization (UNESCO) Intangible Cultural Heritage List, and three of them have staged plays using Lingzi.Provinces with most Lingziusing operas are clustered in central and eastern China (Figure 1), overlapping or being relatively close to the historical distribution areas of Reeves's pheasants.
The results of our statistical analyses revealed a total of 1054 state-run troupes-81 provincial, 284 city, and 689 county troupes-in the Chinese mainland which perform operas using Lingzi.Of these, only 107 troupes (10.2% of the state-run troupes performing operas using Lingzi) pro-vided information on the use of Lingzi.Eleven troupes (10.3%) stated that they have stopped playing using Lingzi a long time ago and have not kept Lingzi ever since, while 96 troupes (89.7%) still continue the use of Lingzi.
Among the three administrative categories of state-run troupes, the inventory of Lingzi in provincial troupes (24.52 ± 15.07 pairs, n = 23) was significantly larger than that in city troupes (11.00 ± 13.12 pairs, n = 34) and county troupes (5.63 ± 4.64 pairs, n = 49) (H = 34.368,p < 0.001).No significant difference was observed in the annual supplement between the province, city, and county troupes (H = 0.193, p = 0.908).Therefore, to calculate the annual demand of Lingzi, we used the mean annual supplement of Lingzi from all types of troupes (1.10 ± 1.44 pairs, range: 0−6, n = 88) and multiplied it with the total number of troupes performing operas using Lingzi (n = 1054).Thus, we estimated that the annual supplement of Lingzi in China is 1159.4pairs.Majority of the feathers studied were 150−160 cm long, mostly obtained from Lingzi of city and county troupes; feathers obtained from provincial troupes were mostly 160−170 cm in length (Figure 2).

Source of feathers on Lingzi
The results of hierarchical clustering analysis showed that a majority of votes were in favor of splitting the three types of samples into two categories: samples from troupes and wild were clustered together, while samples from captive-bred birds were clustered separately (Figure S1).Data obtained from the scatter plot of isotopes from the three sources also show the same characteristics (Figure 3).The results of the random forest analysis predicted that all feather samples obtained from the troupes were from wild sources; the accuracy of the model was 100%.

Distribution and demand
Before the 1960s, Reeves's pheasant were widely distributed in the central and northern mountains of China (Zheng, 1978).Nowadays, the bird's populations have extirpated in the Hebei and Shanxi provinces, and the species' range has become divided into eastern and western regions in central China (Zheng & Wang, 1998).There are many operas using Lingzi in and around the historical distribution area of Reeves's pheasant (Figure 1).Hebei and Shanxi provinces were the cultural centers of the Yuan Dynasty (1271−1368).According to the earliest verifiable literature, it was at this time that  Lingzi use began in Xiqu (Wang, 1958); numerous operas still use Lingzi in this region.Though there are no Reeves's pheasant distributions in the Zhejiang and Fujian provinces, the economy here was developed during the Ming Dynasty (1368−1644), and regional trade and cultural exchanges with areas performing using Lingzi may have resulted in the proliferation of operas using Lingzi here (Zhang, 1993).The Sichuan and Guizhou provinces and Chongqing municipality of southwest China are also distribution areas of Reeves's pheasant; however, only few F I G U R E 3 Scatter plot of carbon (δ 13 C) and nitrogen (δ 15 N) isotopes from tail feathers of Reeves's pheasant obtained from three sources.operas use Lingzi here; this can be attributed to the small number of operas originating here owing to the presence of a closed natural environment and independently developed cultural system.However, in this region, the few operas of local origin mostly use Lingzi and continue to thrive to this day.
We observed that there is a great demand for Reeves's pheasant feathers in state-run troupes.Furthermore, some private enterprises and folk troupes also perform operas, for which scale and feather use information is unavailable.Therefore, the actual demand for feathers must be higher.We learned through the interviews that the annual demand of Lingzi is directly related to the loss of feathers by the troupes.The number of plays performed by a troupe varies each year; when troupes perform more plays using Lingzi, especially those that need Lingzi stunts (performances that require bending or shaking feathers), the loss of feathers is higher in that year.Moreover, the use of Lingzi by novice actors results in substantial damage of the feather owing to a lack of skills; this situation is more common in the Xiqu academies.

Source of feathers on Lingzi
In the Xiqu field, there is an old adage passed down by word of mouth that living Lingzi-that is, tail feathers plucked from a living body-are of the highest quality (Gong, 1987).This may reflect the custom of hunting wild Reeves's pheasant for tail feathers in ancient China, when the population was abundant; this has also been recognized by some Xiqu practitioners.Hunting Reeves's pheasant was common before the 1980s (Hu & Wang, 1981;Zheng, 1978); however, it was not until 1988 that the Reeves's pheasant was listed as a wildlife under special state protection, and hunting was prohibited.Determining whether the Xiqu industry still retains the practice of using wild-source tail feathers is crucial to the formulation and successful implementation of conservation policies.High cultural barriers in the Xiqu industry make it difficult for laymen to enter this community that has been passed down for hundreds of years (van Uhm & Wong, 2019).The lack of social relations in the Xiqu community also increases the vigilance of practitioners, making it more difficult to investigate the trade routes of Reeves's pheasant tail feathers head-on, compared to other investigations into illegal wildlife trade.Illegal harvesting of male pheasant tail feathers was reported in Henan, Hubei, and Anhui regions; however, no information about the trade chain was available from the hunters (Zhou et al., 2015).Till now, only 24 cases of killing Reeves's pheasant have been convicted in China, all without the mention of feather trade (https://wenshu.court.gov.cn/).Therefore, we decided to use animal products, particularly Lingzi, to reveal the source.The results of isotope analysis and group-ing of Lingzi feathers with wild feathers are supported by the information obtained from farms, zoos, and breeding institutions that claimed not to supply feathers to Xiqu.The breeding of this species in farms is not to meet the demand of the Xiqu market but to sell live animals and specimens to maximize profits.Our findings indicate that the Xiqu industry continues to rely on wild-source feathers to make Lingzi, and that there may be additional illegal hunting and trade that has not yet been discovered.

Cultural development and wildlife protection
Strategies to change wildlife consumption behavior should not only focus on the conservation status or threats of species of concern, but also on the social associations that these species provide to consumers (Wang et al., 2021).The tail feathers of the Reeves's pheasant have been worn by heroic characters in Xiqu since ancient times and have become a symbol of courage and glory in China; therefore, the protection of this species has both cultural and ecological significance.Balancing the demand and supply to realize the rational utilization of animal resources is one of the ideal ways to protect wildlife.For instance, to support the development of traditional medicine, China has ensured the protection and rational utilization of musk deer resources through the industrialized breeding and quota system (Zheng et al., 2022).Our survey has shown that Lingzi is expensive, and its cost is directly proportional to the length and appearance of the tail feathers used.A pair of Lingzi with feathers of approximately 1 m will cost more than 100 yuan ($13.84), and feathers over 1.8 m are priced at more than 3000 yuan ($415.32).Increasing the market supply through captive breeding might reduce the high profits generated by poaching, thereby reducing the occurrence of poaching and protecting the species (Damania & Bulte, 2007).In addition, research on captive-breeding of Reeves's pheasant has been carried out in China since 1986; the breeding station in Dongzhai National Nature Reserve in the Dabie Mountains of central China, which considers the protection of Reeves's pheasant as the primary objective, has established 25 generations and more than 400 pairs of captive-bred populations (Peng et al., 2020;Wei et al., 2018;Xu et al., 2007).Thus, it has the potential for large-scale reproduction of Reeves's pheasant for feather supply.For balancing the inheritance of Xiqu and wildlife protection, we believe that it is appropriate to establish proprietary breeding institutions to harvest high-quality tail feathers, which would ensure the traceability of these feathers by adding labeled isotopes.We also recommended the use of stable isotope analysis as a basis for combating illegal trade in law enforcement.

F
Distribution of Reeves's pheasant and operas using Lingzi (not including Hong Kong, Macao, and Taiwan).

F
Length of Reeves's pheasant feathers obtained from Lingzi of provincial, city, and county troupes.