Introduction
The Covid-19 pandemic, which probably arose from zoonotic sources, has provoked wide-ranging discussion on how best to prevent future pandemics (Chen et al., Reference Chen, Li, Li, Xiao, Ye and Yan2020; Van Kerkhove et al., Reference Van Kerkhove, Ryan and Ghebreyesus2021). It is widely acknowledged that efforts to combat zoonotic disease should include both biosecurity measures (e.g. disease surveillance) as well as policies that target wildlife consumption (Yuan et al., Reference Yuan, Lu, Cao and Cui2020). Proposed interventions include reduced frequency of direct human–wildlife contact (Karesh et al., Reference Karesh, Cook, Bennett and Newcomb2005), increased surveillance of wildlife and traders to monitor potential pathogens (Daszak et al., Reference Daszak, Olival and Li2020), risk management tools tailored towards wildlife markets (Wikramanayake et al., Reference Wikramanayake, Pfeiffer, Magouras, Conan, Ziegler and Bonebrake2021), full or partial bans on wildlife consumption (ideally in conjunction with alternative sustenance options), enhanced legal protection for more wildlife species, increased prosecution of wildlife trade violations, and regulations for captive-bred wildlife (Borzée et al., Reference Borzée, McNeely, Magellan, Miller, Porter and Dutta2020; Chang & Chang, Reference Chang and Chang2020). A common rationale for enhanced regulations is that current wildlife consumption patterns pose a high level of risk, particularly in wet markets (Aguirre et al., Reference Aguirre, Catherina, Frye and Shelley2020).
In China, the government responded to the outbreak of Covid-19 with a ban issued in February 2020 (Koh et al., Reference Koh, Li and Lee2021; Sardana & Fischer, Reference Sardana and Fischer2021). This prohibits the farming, hunting, selling and purchasing of species defined by law as terrestrial wild animals for food consumption (Huang et al., Reference Huang, Wang, Yang, Valitutto and Songer2021). However, this ban does not apply to aquatic species and explicitly allows the use of wildlife for medicinal, ornamental or scientific purposes (Rizzolo, Reference Rizzolo2021; White, Reference White2021). Furthermore, for species that are used for multiple purposes (e.g. small carnivores such as the civet Paradoxurus hermaphroditus and racoon dog Nyctereutes procyonoides, which are utilized both for food and fur), only their production and consumption for food are banned (Whitfort, Reference Whitfort2021).
The February 2020 ban was not without precedent, as the 2003 severe acute respiratory syndrome (SARS) epidemic also led to changes in wildlife regulations in China. However, as the post-SARS ban on wildlife consumption was lifted a few months after its implementation (Li et al., Reference Li, Zhang, Hu, Zhao, Zhao and Kváč2020), it remains to be seen whether the severity of Covid-19 could lead to more permanent wildlife regulations (Huang et al., Reference Huang, Wang, Yang, Valitutto and Songer2021). Although the February 2020 ban can be viewed as an improvement on an outdated and ineffective legal wildlife trade management system (Yang et al., Reference Yang, Liu, Li and Zhang2020; Koh et al., Reference Koh, Li and Lee2021; Xiao et al., Reference Xiao, Lu, Li, Zhao and Li2021a,Reference Xiao, Newman, Buesching, Macdonald and Zhoub), it is unclear whether the Chinese public will view the ban as legitimate (Cawthorn et al., Reference Cawthorn, Kennaugh and Ferreira2021).
Furthermore, China has promoted the legal and intensive farming of numerous wildlife taxa, including badgers, bears, bamboo rats, civets, crocodiles, deer, ferrets, foxes, frogs, macaques, mink, muskrats, pheasants, pangolins, raccoon dogs, snakes, the tiger Panthera tigris and turtles (Whitfort, Reference Whitfort2021). By the end of 2006, there were 19,018 wildlife breeding and farming ventures in China; Wang et al. (Reference Wang, Yang, Wronski, Chen, Hu and Huang2019) estimated the number of farmed wildlife to range from the thousands (c. 5,000 tigers and 10,000 bears) to the millions (70 million mink). China also imports a large number of farmed wildlife. During 1997–2016 on average over half (56%) of wildlife imported into China was captive-bred, and this per cent was higher for birds and fish (Jiao & Lee, Reference Jiao and Lee2022). However, farmed wildlife has yet to saturate the demand of Chinese customers, as consumers are often willing to pay more for wild-sourced products (Dutton et al., Reference Dutton, Hepburn and Macdonald2011; Coals et al., Reference Coals, Moorhouse, D'Cruze, Macdonald and Loveridge2020), have complex motivations for switching between wild and farmed products (Hinsley et al., Reference Hinsley, Wan, Garshelis, Hoffmann, Hu and Lee2022) and are influenced by the social acceptability of various forms of wildlife products (Rizzolo, Reference Rizzolo2021).
The proliferation of wildlife farms can have severe and unintended consequences for the spread of zoonotic diseases (Jiao & Lee, Reference Jiao and Lee2021), as wildlife farms are both suppliers to wet markets and also potential sources of disease in their own right (Petrovan et al., Reference Petrovan, Aldridge, Bartlett, Bladon, Booth and Broad2021). Although the February 2020 ban applies to farmed terrestrial wildlife (You, Reference You2020), addressing the wildlife farming industry in China and its potential links to zoonotic diseases remains a significant logistical challenge (Wang et al., Reference Wang, Shao, Luo, Chuai, Xu and Geng2020).
Research in China post-Covid-19 has found evidence for decreased wildlife consumption (Liu et al., Reference Liu, Li, Cui, Yang, Shan and Zhang2020a,Reference Liu, Ma, Zhang and Zhangb; Lin, Reference Lin2021), although this could be motivated by legal enforcement or social stigmatization (Lin, Reference Lin2021) as well as awareness of food safety (Li & Wang, Reference Li and Wang2021) and proximity to Covid-19 outbreaks (Zhang et al., Reference Zhang, Yang, Hu, Cirella and Xie2022). Attitudes towards wildlife consumption in China post-Covid-19 are often complex; in another China-based study, although almost 80% of respondents believed that the wildlife trade facilitated disease transmission, 31% continued to consume wild animals (Li et al., Reference Li, Daszak, Chmura, Zhang, Terry and Fielder2021). Furthermore, perceptions of zoonotic risk and the impacts of these attitudes on wildlife consumption behaviours vary across taxa, with the impacts of zoonotic pandemics yielding more significant reductions in the consumption of some taxa than of others (Duonamou et al., Reference Duonamou, Konate, Djossou, Mensah, Xu and Humle2020). The ongoing debate over the origins of Covid-19 has complicated policy responses further despite documentation of a zoonotic link (Worobey, Reference Worobey2021; Zhu et al., Reference Zhu, Chen, Rizzolo and Li2023).
Since the emergence of Covid-19 the feasibility of wildlife consumption bans (either broad or targeted) continues to be debated. Although some argue that broad bans are not realistic given cultural and livelihood attachments to wildlife products (Zhu & Zhu, Reference Zhu and Zhu2020; Biggs et al., Reference Biggs, Caceres-Escobar, Kock, Thomson and Compton2021), others have noted that wildlife consumption bans are an essential first step towards effective protection of wildlife and human health (Xiao et al., Reference Xiao, Lu, Li, Zhao and Li2021a). Multiple factors such as cultural beliefs, economic conditions, demographic variables and the supposed medicinal properties of wildlife products affect whether and what type of bans will be accepted by the public and thus will be feasible to implement (Rizzolo, Reference Rizzolo2021). Yet there is evidence (albeit in a UK sample) that a narrative of Covid-19 that emphasizes its origins in human–animal interaction elicits support for bans on the commercial trade of wildlife (Shreedhar & Mourato, Reference Shreedhar and Mourato2020). This suggests that support for policies to protect wildlife and human health can still be harnessed even if research on the origins of Covid-19 is politicized or still in progress (Zhu et al., Reference Zhu, Chen, Rizzolo and Li2023). However, the current literature on China post-Covid-19 focuses on wildlife consumption patterns (Lin, Reference Lin2021; Zhang et al., Reference Zhang, Yang, Hu, Cirella and Xie2022) rather than support for particular regulations such as bans. More work needs to be done to investigate support for regulatory frameworks in China post-Covid-19 and specifically to model how perceptions of the relationship between wildlife consumption and zoonotic risk combine with other variables to influence support for wildlife policies.
Here we expand upon previous literature through an analysis of how support for wildlife policies in China differs according to taxa and whether traded individuals are wild-caught or farmed. Our objectives are two-fold: (1) to measure respondents' attitudes towards the current wildlife consumption ban in China as well as attitudes towards regulations of wildlife more broadly, and (2) to examine which variables correlate with support for bans on wild-caught and farmed wildlife for nine taxa. We discuss the implications of these results for conservation policies in China in the wake of Covid-19.
Methods
We recruited 974 adult respondents through the research firm Qualtrics (Provo, USA), which has experience in China-based sampling (Rizzolo, Reference Rizzolo2021) and offers a more representative sampling process compared to other online survey tools (Boas et al., Reference Boas, Christenson and Glick2020). We first wrote the survey in English and then a bilingual member of the research team translated it into Chinese. Before the full survey began, we conducted two pre-tests to ascertain translation clarity, data quality and survey validity. We recruited respondents from the Qualtrics panel and completed the survey online in June 2021 through the Qualtrics platform.
The sampling drew from a China-based panel of individuals recruited from diverse sources, such as website intercept recruitment, member referrals, targeted e-mail lists, gaming sites, customer loyalty web portals, permission-based networks and social media. Qualtrics validates the names, addresses and dates of birth of panel members via third-party verification measures prior to their inclusion in a panel. Panellists are compensated through various mechanisms: they could be airline customers who choose to join to receive credits that can be applied to flights, retail customers who opt in to receive points at a retail outlet, or general consumers who participate to obtain cash or gift cards. Panel members are given an e-mail invitation or are prompted on the respective survey platform to participate in a given survey. Qualtrics utilizes quality control measures, such as tracking the amount of time a respondent spends on a survey, to ensure data quality. In addition, we employed quality checks that allowed for nuance in attitudes to be expressed but excluded respondents who answered the questions seemingly without reading or comprehending them. For example, these checks removed respondents who stated ‘No, I have never consumed wild animals’ but then indicated consumption in a subsequent question or respondents who selected an answer in addition to ‘None of the above’ within the same question.
The full survey (Supplementary Material 1) consisted of sections on demographic information, beliefs and behaviours related to the origins of Covid-19 and wildlife consumption, beliefs and behaviours that pertain to health and traditional Chinese medicine, and attitudes about the regulation of wildlife consumption and the prevention of potential zoonotic transfer through bans and other means. Here we present data from the last section of the survey; the other sections are discussed elsewhere (Rizzolo et al., unpubl. data; Zhu et al., Reference Zhu, Chen, Rizzolo and Li2023a,Reference Zhu, Chen, Rizzolo and Qianb).
We selected a total of nine taxa for assessment from the diverse taxa that are subject to wildlife farming and/or have been investigated as potential transmitters of Covid-19 or other zoonotic diseases or parasites: civets (Tu et al., Reference Tu, Crameri, Kong, Chen, Sun and Yu2004), frogs (Yang et al., Reference Yang, Dai, Deng, Lu and Jiang2007), the tiger (Zhang et al., Reference Zhang, Hua and Sun2008), bears (Li, Reference Li2004), bamboo rats (Ma et al., Reference Ma, Wang, Zhang, Wu and Zhao2018; Li et al., Reference Li, Zhang, Hu, Zhao, Zhao and Kváč2020), mink (Liu et al., Reference Liu, Li, Cui, Yang, Shan and Zhang2020a; Fenollar et al., Reference Fenollar, Mediannikov, Maurin, Devaux, Colson and Levasseur2021), snakes (Wang et al., Reference Wang, Zhou, Gong, Deng, Zou and Wu2011), bats (Giri et al., Reference Giri, Bhardwaj, Shegane, Gehi, Kumar and Gadhave2021) and pangolins (Gatti, Reference Gatti2020). These taxa differ with respect to the production system used (farmed/wild-caught) and perceptions of zoonotic risk. For most taxa we worded the questions to refer to the entire animal. However, for tigers and bears (as these animals are not consumed whole) we asked specifically about tiger bones and bear bile. We measured respondent attitudes towards the current wildlife ban in China with the following question (Q35 in the Supplementary Material): ‘Do you think the ban on wild animal consumption is (a) adequate, (b) too strict, (c) not strict enough, (d) I don't know this policy or (e) other?’. Attitudes towards taxon-specific bans were measured using two questions (Q31 and Q32 in the Supplementary Material), which asked: ‘Do you support a ban on the following animals if caught from the wild (a) yes or (b) no?’ and ‘Do you support a ban on farming the following animals (a) yes or (b) no?’. As this survey employed sampling quotas that were based upon membership in various demographic categories, we recorded gender, age, education and income.
We analysed the data using SPSS 26 (IBM, 2019). Firstly, we generated descriptive statistics about attitudes towards the current wildlife ban in China. We did not detect collinearity; correlations between the consumption of various taxa were all < 0.7. Secondly, we modelled support for bans of consumption of wild-caught and farmed individuals of the nine taxa. Given that we modelled support for a ban as a dichotomous variable, we used binomial logistic regression. Independent variables were belief that a particular taxon was a probable transmitter of Covid-19 to humans (measured as yes/no), consumption patterns of reptiles, amphibians and mammals, and the demographic control variables of gender, age, education and income. The reference categories for the categorical variables were female (for gender), 70–79 years old (for age), middle school or below (for education), and < CNY 1,000 (for income).
Results
Sample
The sample included gender and age quotas representative of the adult population of China (National Bureau of Statistics of China, 2019) and was diverse in terms of gender, age, income, education and occupation (Table 1). The sample was predominantly urban (91%) but included respondents from all regions of China (Fig. 1).
Fig. 1 The geographical distribution of the 974 respondents.
Table 1 Demographic characteristics of the survey sample of 974 respondents.
1 The eight primary occupational categories in China.
Correlations indicated a moderate relationship between high earners and higher education levels. The correlation between the highest level of education (graduate degree) and the highest income bracket (> CNY 20,000 per month) was 0.236 and the correlation between the lowest level of education (middle school or less) and the lowest income bracket (< CNY 1,000 per month) was 0.146.
As measured using the Pearson χ 2 test (χ 2 = 2.864, df = 1, P = 0.091), there was no difference between urban and rural respondents in the belief that the current wildlife ban in China is adequate. There was also no difference between rural and urban respondents in support of a ban on the consumption of farmed or wild-caught wildlife (Table 2).
Table 2 Results of Pearson χ 2 tests comparing the support of rural and urban respondents of a ban on the consumption of farmed and wild-caught wildlife (df = 1 in all cases).
Attitudes towards regulation
The results indicated broad support for many pathways of regulation. The majority of respondents supported regulations on imported products (78%), wild animal farms (76%), wet markets (75%), wild animal trade (69%) and virology research (62%). Furthermore, the majority of respondents supported bans on both wild-caught and farmed wildlife (Table 3). Although there was some variation between taxa, respondents supported bans on valuable taxa (pangolins, tigers) and widely farmed taxa (bamboo rats, snakes).
Table 3 Per cent of the sample of 974 respondents who supported bans on the consumption of wild-caught wildlife and farmed wildlife.
When asked in particular about the ban on wild animal consumption that China issued after the outbreak of Covid-19, 75% of respondents thought that people were adhering to this ban (14% believed people were not in compliance and 11% were unfamiliar with the policy). Respondents had differential attitudes towards the adequacy of the current ban in China (Fig. 2) and towards overall policies governing wild animals (Fig. 3).
Fig. 2 Responses (of a total of 974 respondents) to the question ‘China's ban on wildlife consumption is _____’.
Fig. 3 Responses (of a total of 974 respondents) to the question ‘Current policies governing wild animals in China are _____’.
When asked which rationales would lead them to support a ban on wildlife consumption, respondents endorsed numerous rationales, with protection against future pandemics and protection of the environment being motivators for 80% of respondents (and protection of animal welfare and consumer health being motivators for 68% of respondents).
Support for taxon-specific bans
In the models of bans pertaining to wild-caught animals (Table 4), perceived zoonotic risk was a strong correlate for increased support for bans of all taxa except pangolins and bats. Although zoonotic risk was not significant in the bat and pangolin models, one variable that was significant in those models (as well as the models for other taxa) was belief that the current ban in China is adequate. Consumption of mammals increased the odds of supporting a ban on wild-caught tigers and bears. Furthermore, in these models being aged 60–69 years decreased the probability of support for such bans.
Table 4 Odds ratios (Exp(B) values) from models of support for bans on consumption of nine wild-caught taxa amongst the 974 respondents.
*P < 0.001.
1 For categorical variables, the reference categories are: female (for gender), middle school or below (for education), 70–79 (for age), and < CNY 1,000 (for income).
†Nagelkerke R 2 is a measure of model fit (< 0.2 indicates a weak relationship between the predictors and the outcome, 0.2–0.4 indicates a moderate relationship, and > 0.4 indicates a strong relationship).
The models of bans pertaining to farmed wildlife (Table 5) showed different results from those of bans on wild-caught taxa. Zoonotic risk was not a significant indicator for most taxa (the two exceptions being civets and frogs). In contrast to the models on bans concerning wild-caught animals, respondents aged 60-69 years were more likely to support bans regarding farmed wildlife of numerous taxa. Income had no effect in the models on bans concerning wild-caught animals, but was a significant factor in those regarding farmed wildlife: belonging to a higher-income group increased the probability of supporting a ban on farmed frogs, tigers, mink and pangolins.
Table 5 Odds ratios (Exp(B) values) from models of support for bans on consumption of nine farmed taxa amongst the 974 respondents.
*P < 0.001.
1For categorical variables, the reference categories are: female (for gender), middle school or below (for education), 70–79 (for age), and < CNY 1,000 (for income).
†Nagelkerke R 2 is a measure of model fit (< 0.2 indicates a weak relationship between the predictors and the outcome, 0.2–0.4 indicates a moderate relationship, and > 0.4 indicates a strong relationship).
Discussion
Our findings have several important implications for wildlife conservation interventions in China following the Covid-19 pandemic. The feasibility of the current wildlife ban in China and the potential for additional regulatory frameworks around wildlife consumption and zoonotic diseases have both been debated extensively in the conservation literature and policy arenas (Yang et al., Reference Yang, Liu, Li and Zhang2020). Our results demonstrate both how the current wildlife ban in China is perceived and how support for wildlife consumption bans is contextualized and segmented by demographic variables, taxa, perceived zoonotic risk and whether animals are wild-caught or farmed. Such information is essential for the design and implementation of conservation initiatives such as demand-reduction efforts and/or enhanced regulation (Naidoo et al., Reference Naidoo, Bergin and Vertefeuille2021).
A crucial implication of our research is that, at least in an urban sample, there is widespread support for wildlife consumption bans in China post-Covid-19. Overall, the majority (70–90%) of respondents supported bans on wild-caught and farmed wildlife of several taxa. This support for bans extended to species of various taxa (amphibians, mammals and reptiles) and included threatened species used for lucrative products (tiger bone), widely farmed and consumed taxa (bamboo rats) and products used for medicine (bear bile). This suggests that now is an opportune moment to support efforts in China to regulate and limit wildlife consumption. Although it is important to replicate our study with a predominantly rural population, our analysis indicated no differences in support for bans (of either farmed or wild animals) between our urban and rural respondents. Furthermore, urban residents in China are significant consumers of wildlife products (Zhang et al., Reference Zhang, Hua and Sun2008; Zhang & Yin, Reference Zhang and Yin2014), so a high level of support for bans amongst this population is an important measure of the social acceptability of wildlife bans in China. Protection against future pandemics was not the only rationale given for the endorsement of wildlife consumption bans; c. 80% of respondents viewed protection of the environment as a valid reason for wildlife consumption bans. This could be because a large number of our respondents (62%) had protective attitudes towards wildlife (Zhang et al., Reference Zhang, Hua and Sun2008; Rizzolo, Reference Rizzolo2021), which could predispose them towards environmental rationales for policy decisions.
Comparisons with studies conducted prior to the outbreak of Covid-19 demonstrate that this pandemic could have contributed to a shift in such support. For example, a survey in 2011/2012, also with an urban Chinese sample, demonstrated that c. 63% of respondents supported tiger farms (Liu et al., Reference Liu, Jiang, Li, Fang, Ping and Luo2015), whereas in our study 80% of respondents supported a ban on tiger farms. Although these samples are not directly comparable, they provide a snapshot of how views on wildlife consumption in China have shifted.
Our results also have significant implications for the differences in consumption of wild-caught and farmed wildlife in China. For almost all of the nine taxa, support for bans concerning wild-caught wildlife was greater than support for bans on the consumption of farmed wildlife (Table 3). In addition, the demographic correlates of support differed for bans regarding wild-caught and farmed wildlife. Income was a significant variable in the models concerned with farmed wildlife but not in those relating to wild-caught animals, with higher income increasing the probability that a person would support a ban on the consumption of farmed frogs, tigers, mink and pangolins. Farmed wildlife products are often perceived as less prestigious, effective or potent, whereas wild-caught wildlife is often viewed as nutritious, delicious and socially valued (Zhang et al., Reference Zhang, Hua and Sun2008; Coals et al., Reference Coals, Moorhouse, D'Cruze, Macdonald and Loveridge2020). One possible explanation for this is that higher earners can afford the often more expensive wild-caught wildlife products and thus bans on farmed wildlife do not threaten their consumption patterns. Higher earners could also have achieved higher levels of education, which could make them more aware of the connections between wildlife consumption and health outcomes. However, it is also possible that members of higher income groups would have objections to wildlife farms that are not shared by lower income groups, such as concerns about animal welfare (You et al., Reference You, Li, Zhang, Yan and Zhao2014).
For those aged 60-69 years, age had the opposite effect on the support of bans concerning wild-caught vs farmed wildlife: these respondents were less likely to support bans on the consumption of wild-caught wildlife and more likely to support bans on wildlife farming. Although older respondents probably still use or consume wildlife, especially products such as tiger bones that are used for health ailments common in the elderly (Gratwicke et al., Reference Gratwicke, Mills, Dutton, Gabriel, Long and Seidensticker2008), this segment of the population could view wild-caught species as more beneficial to their health and thus not object to farming bans.
Future research should aim to determine the reasons for these demographic variations between support for bans concerning wild-caught and farmed wildlife in terms of ethical concerns, perceived quality/price of the wildlife product, attitudes towards the expansion of wildlife farms or other possible explanations. However, the immediate indication is that interventions aimed at building support for bans should be tailored to different segments of the population depending upon whether the wildlife is farmed or wild-caught. Combining wild-caught and farmed wildlife (even of the same species) into a single category increases the risk of incorrect inferences regarding both the level of the support for bans and the segments of the population that are most receptive or resistant to a ban.
Our findings also indicate how differences between support for bans regarding wild-caught and farmed wildlife correlate with perceptions of zoonotic risk. In the models on farmed wildlife, zoonotic risk was only significant for two taxa: civets and frogs. For civets, this could be because of their much-publicized role in the zoonotic outbreak of SARS as well as the detection of the virus in civets in wet markets in Wuhan in the early days of the Covid-19 outbreak (Wang & Eaton, Reference Wang, Eaton, Childs, Mackenzie and Richt2007; Hassanin et al., Reference Hassanin, Grandcolas and Veron2021). For the models on wild-caught animals, however, zoonotic risk was significant (and by larger factors) for almost all taxa assessed. For both farmed and wild-caught wildlife, zoonotic risk (when significant) increased the support of bans. Therefore, these differences might be because of the increased tendency during Covid-19 in both the popular media and the scientific literature to discuss the wildlife trade in general terms rather than focusing on specific species or contexts (Wang et al., Reference Wang, Tilley, Phalke, Andersson, Dingle and Hatten2022). The term ‘wildlife’ could elicit conceptualizations of wild-caught wildlife, even though most wildlife markets include a mixture of wild-caught and farmed species (Xiao et al., Reference Xiao, Lu, Li, Zhao and Li2021a,Reference Xiao, Newman, Buesching, Macdonald and Zhoub). It is possible that this generalized framing of zoonotic risk leads consumers to underestimate the risks posed by farmed wildlife (and the associated incentives for bans on wildlife farming).
Another significant difference between the wild-caught and farmed wildlife models was that respondent attitudes towards the current ban in China were statistically significant only in the models concerning wild-caught animals. In these models, belief that the current ban was adequate increased the probability that a person would support a ban on products from wild-caught animals (but not a ban on farming of the same taxa). This might be because respondents evaluate the hypothetical feasibility of bans based on their beliefs about the effects of the current ban in China. In other words, if they think that the ban has been effective at protecting wildlife and human health, it makes sense that they would develop a generalized pro-ban attitude. However, in the models on farmed wildlife, respondent attitudes towards the current wildlife consumption ban in China had no effect. It could be that respondents view the current ban as applicable to wild-caught animals only and/or do not take the current ban into account when evaluating the feasibility of bans on farmed wildlife. This has important conservation implications because this ban applies to both wild-caught and farmed wildlife (You, Reference You2020). However, respondents do not seem to consider this law when responding to hypothetical farmed wildlife bans. Others have noted the challenges of implementing the current wildlife consumption ban in China within the wildlife farming industry (Wang et al., Reference Wang, Shao, Luo, Chuai, Xu and Geng2020) because of public sympathy for the economic plight of wildlife farmers (Li, Reference Li2020). Prior research has identified the economic and logistical challenges faced by wildlife farmers and government authorities (You, Reference You2020), and we found that consumer attitudes are an additional hurdle to compliance with a ban on wildlife farms. Although the current ban in China has substantially altered the legal context of wildlife farms (You, Reference You2020), more education and outreach are needed concerning the zoonotic risks of wildlife farms to ensure that consumers are aware that farmed wildlife products are included in this ban.
Finally, our findings have several important implications for species-specific conservation and regulatory initiatives in the wake of Covid-19. Mammal consumption increased the probability that a person would support a ban on wild-caught tigers and bears. This requires follow-up research, as it could be a product of substitutability: it is possible that these respondents consume other mammals (perhaps more common or less threatened species) and thus have no need to consume tigers and bears. Although less than a quarter of our respondents perceived tigers (11%) and bears (11%) as probable zoonotic transmitters of Covid-19 (Rizzolo et al., unpubl. data), if a respondent believed this, then it greatly increased the odds (by 12- and 8-fold, respectively) of supporting a ban on wild-caught individuals of these species. This could be because of underlying attitudes that predispose certain people to view wildlife consumption as risky; in our sample, respondents with a protective attitude towards wildlife were more likely to view bears and tigers as probable sources of Covid-19 (Rizzolo et al., unpubl. data).
Both in this sample (Rizzolo et al., unpubl. data) and in the wider literature (Gatti, Reference Gatti2020; Giri et al., Reference Giri, Bhardwaj, Shegane, Gehi, Kumar and Gadhave2021) pangolins and bats are amongst the taxa that are most likely to be viewed as transmitters of Covid-19. Even though bats and pangolins are perceived as probable zoonotic transmitters overall, their status as presumed zoonotic vectors was not a factor in predicting support for wildlife consumption bans in this study. However, belief that the current ban was adequate increased support for a ban on wild-caught pangolins and bats. As they have received significant policy attention, it follows that support for bans on consumption of these taxa would be connected to attitudes towards the adequacy of the governmental response. Two taxa for which beliefs about the adequacy of the current ban did not have an effect were bamboo rats and frogs. Although additional research is needed to understand this, it is possible that these taxa (neither of which are generally perceived as particularly charismatic) have been a lower priority in the enforcement of the current ban and that respondents therefore do not see the ban as applicable to them.
One limitation of our research is that it is based on self-reporting, which in the case of wildlife consumption can be subject to a social desirability bias (i.e. the tendency to present oneself positively to the researcher; Nancarrow & Brace, Reference Nancarrow and Brace2000; Rizzolo, Reference Rizzolo2021). Although we attempted to address this by surveying respondents anonymously and not asking about illegal consumption, it remains a potential limitation. An additional limitation is that because Covid-19 was unexpected, we had to rely upon a cross-sectional rather than a longitudinal approach (although we complement our cross-sectional data with longitudinal analysis of additional data sources; Zhu et al., Reference Zhu, Chen, Rizzolo and Qian2023b).
Conclusion
The global outbreak of Covid-19 has brought increased attention to the zoonotic risks posed by wildlife trade and consumption. However, there is continuing debate about the role of regulations and wildlife bans in the reduction of high-risk wildlife consumption behaviours. We have noted the strong support in our sample for wildlife consumption bans, with the majority of respondents supporting bans concerning both wild-caught and farmed wildlife. Furthermore, although zoonotic risk is an important variable in determining support for wildlife consumption bans, the relationship between perceived zoonotic risk and support for bans differs depending on the taxa and whether the animal is wild-caught or farmed. This highlights both the feasibility of the current wildlife ban in China and opportunities to further mitigate the environmental and zoonotic risks of wildlife consumption.
Acknowledgements
Funding for this research was provided jointly by Shanghai Jiao Tong University, China, and Wageningen University, The Netherlands.
Author contributions
Study design: all authors; data analysis, writing: JBR; revision: ALZ, RC.
Conflicts of interest
None.
Ethical standards
This research abided by the Oryx guidelines on ethical standards. The ethics committee of the Michigan State University institutional review board approved this research (study number 00005583) and all participants gave their informed consent to participate. Michigan State University was the primary affiliation of JBR at the time of research.
Open access
