2.1 Shop interviews

For the shop interviews, we accessed the government database of registered TCM retailers (including pharmaceutical shops, clinics, and hospitals) in Hong Kong. As of September 2021 (Chinese Medicine Council of Hong Kong, 2021), a total of 5193 retailers (henceforth “shops”‘) were listed. We randomly sampled 520 shops from the list to conduct semistructured interviews. In instances where the target shops were not at the registered location, possibly due to time lags in updating the list, shops nearby were approached for interviews instead. Interviews were conducted over a period from September 2021 to December 2021. All interviews were carried out in Cantonese by members of the research team. All interviewees were fully informed about the study purpose and their rights to withdraw from the study at any time. Oral consent was obtained from each participant before the interview. Gel ink pens (3.5 HKD each) were given as gifts at the beginning of the interview to increase participation rate and thank participants for their contribution. To protect participants’ privacy, we did not collect or record any data that could reveal shop names participant individual identities.

Interview questions (Supporting Information: Appendix S1) covered three aspects: knowledge about medicinal uses (tokay geckos and pangolin scales), knowledge about the medicine trade, and the impact of COVID-19 on the trade. Questions regarding the tokay gecko were asked first, followed by inquiries about pangolin scales. Additional data were also gathered, including shop nature (i.e., whether it was a chain shop or private shop), number of people in the shop during the interview, observed gender and age range of the interviewee, and attitude of the interviewee (e.g., resistant, or welcoming). We bought gecko products priced below 200 HKD during the shop interview. For the purchased products, we measured width and length of the dry geckos and recorded package style (e.g., rubber band or metal wire to tie gecko pair together). The duration of each interview varied, ranging from approximately 5 to 20 min, depending on the interviewee's knowledge and willingness to engage in discussion.

2.2 Public survey

Public surveys were conducted using a snowball sampling technique. Research team members disseminated the survey link through their personal networks, coinciding with the timeframe of the shop survey. The survey was written in traditional Chinese (Supporting Information: Appendix S1). We used the Theory of Planned Behavior (TPB) as our guiding framework for the questionnaire design, with certain adaptions (Ajzen, 1991). We hypothesized that the consumption or the intention to consume pangolin/gecko products for medicinal use is influenced by several factors: people's attitude toward TCM and its trade, perceived social norms, and their understanding of behavior control, particularly in terms of legal regulation and perceived expensiveness of the products. We extended the normal TPB model by including two additional factors. First, we considered an individual's understanding of species conservation status and the product origin (whether derived from wild or from captive sources) as behavior control factors. Second, past TCM consumption was included as a predictor for habituated behavior (Chen & Chao, 2011). To ensure the reliability of the survey data, responses were filtered using the time taken to finish the survey. Any survey completion time under 1 min was removed as it suggests potentially unreliable inputs. To maintain the relevance and the context-specific nature of our study, responses from participants who had not lived in HK for the past 5 years were also excluded from the final analysis.

2.3 Statistical analysis

Quantitative data were summarized and analyzed using generalized linear models (GLMs) to investigate correlations between reported product origin, sale quantity, product price, product size, package style, and shop location (administrative district). From the public survey results, GLMs were used to analyze factors correlating with public's consumption intention. A stepwise algorithm (R function StepAIC) was used to find the final model. All statistical analyses were performed in RStudio (R version 4.1.3). We used the Z-test to find significant differences between proportions with a significance threshold of p < 0.05 and effect sizes were evaluated using Cohen's H with a significance threshold of more than 0.2 (Cohen, 2013). Qualitative data were first coded and then grouped into themes, allowing us to understand the general patterns and trends in the data (Newing, 2010).

3.1 Shop interviews

In total, we approached 474 shops for interview (Figure 1), of which 389 were privately owned (82.1%), and the rest were chain shops under different brands. We successfully obtained consent for 169 shops to participate in the interviews (35.7% of 474), the majority of which, 156 in total, were private shops (92.3%). Chain shops exhibited a higher rejection rate than private shops, which might be due to company policy restrictions cited frequently as rejection reason.

For tokay geckos (henceforth “geckos”), 165 shops provided information, with 97 of them (58.8%) reporting recent sales activity. Geckos were normally sold in pairs, though powdered forms were also observed in markets. The reported price of geckos varied from 50 HKD to 500 HKD per pair. Price increases during the past few years were mentioned specifically by 19 sellers, potentially due to COVID-19 and related difficulties with importing. We bought a total of 65 gecko products, with an average price of 120.9 ± 36.1 HKD per pair. According to our generalized linear models (GLMs), the price of purchased gecko products did not correlate with factors such as gecko size (measured by wrist-to-wrist width and head-to-tail-tip length), reported origin, shop location (HK district), or product package (e.g., whether rubber bands or plastic tape were used) (for statistical results, see Appendix II, GLM 1). The interviews revealed a notable variation in the pricing of geckos, with many shops having both cheap and expensive options. Interviewees often differentiated the price based on size or geographic origin of the geckos. Larger geckos and those originating from Wuzhou, Guangxi Province, were often priced higher. However, these two characteristics did not necessarily co-occur; sometimes geckos claimed to be from Wuzhou were smaller in size, leading to contradictions in the pricing rationale. Most respondents also claimed their products were only from Wuzhou (32 out of 77), with another 13 claiming their geckos came from both Wuzhou and other sources, including Thailand and Vietnam.

A total of 36 shops reported annual sales quantity which added up to approximately 1468 pairs of geckos in 1 year (shop average = 40.8 ± 58.0 pairs of geckos per year). GLMs showed no significant correlation between sale quantity and input factors, including gecko price and reported origin as Wuzhou only or other (Supporting Information: Appendix S2, GLM 2). The distributions of reported price and reported sale quantity share a similar pattern that most shops had low/middle price with a low/middle sales quantity (Figure 2). A few shops reported significantly higher sales than others, but no consistent patterns were found across these shops such as geographical region or pricing. Therefore, we computed a simple extrapolation assuming that the reported average sale quantity (40.7 pairs of geckos per year for shops selling geckos) and the number of TCM sellers that reported selling gecko products (58.8% of all registered sellers) from our survey results were representative of the situation in Hong Kong. The extrapolation estimates that Hong Kong could potentially sell 248,553 geckos annually, given that there is a total of 5193 registered TCM shops in Hong Kong.

Although this estimation is very high, it is important to note that 55 of the interviewed shops reported that sales quantity has been in decline in recent years or even a cessation of sales altogether. Reasons proposed for these declines in gecko sales included decreased demand (e.g., “not a popular medicine anymore” n = 27), high prices (n = 20), COVID-19 impacts (including fewer buyers, higher prices, and difficulty in importing) (n = 13), and legal concerns (protecting animal and restrictions in sale/import) (n = 10).

Different uses of geckos were reported by 158 respondents, with the top six uses being treating asthma (n = 85), good for kidney (n = 73), alleviating coughs (n = 66), good for lungs/trachea (n = 53), reducing phlegm (n = 24), and for general healthcare (n = 20). Treating joint pain and skin disease was also mentioned by a few respondents (n < 5). To put in soup or tea was also mentioned by 16 respondents. Overall, respiratory system diseases and general healthcare were the main categories of gecko medicinal applications. More than 75% of respondents (N = 136) showed a positive attitude toward substitutes and agreed that geckos could be substituted to some extent (Figure 3). Four medicines were mentioned more than 10 times by respondents as potential substitutes, including the plant Fritillaria cirrhosa (chuan bei mu in Chinese, 35 mentions), crocodile meat (35 mentions), Lignosus rhinoceros (tiger milk mushroom, 15 mentions), and Ophiocordyceps sinensis (caterpillar fungus, 11 mentions). Other substitutes, including apricot kernels, seahorse/sea dragon, sun-dried aged mandarin peel (chen pi in Chinese), the bark of Eucommia ulmoides, and the fruit of Borassus flabellifer, were mentioned by at least five respondents.

For pangolin-related questions, 145 respondents participated in the interview, and only one indicated recent selling behavior. Respondents showed high awareness of legal restrictions, and 94 out of 145 (64.8%) stated it was illegal to sell, with many specifically mentioning the ban had been in place for more than 10 years. Due to this awareness of legal restrictions, respondents were more cautious with pangolin-related questions, and 21 of them showed resistance to provide in-depth information. For pangolin scale medicinal uses, 127 respondents provided information. The top three mentioned medicinal applications were detoxification (49 mentions), promoting blood circulation/removal of blockages in channels (42 mentions), and clearing heat (23 mentions). Symptoms or illnesses that would use pangolin scales included skin diseases (especially itchiness and acne, 54 mentions), cancers (23 mentions), helping milk secretion (13 mentions), and bone/joint-related symptoms (such as rheumatism, 5 mentions). A total of 113 respondents provided information on potential substitutes. High approval toward substitutes was observed in respondents (Figure 3). However, only 68 respondents provided specific substitutes which covered more than 40 medicines. Most of the proposed substitutes were mentioned only once or twice. Six medicines were mentioned more than five times: wang bu liu (seeds of Vaccaria segetalis), zao jiao ci (dry thorn of Gleditsia sinensis), and jin yin hua (flower of Lonicera japonica) were mentioned by eight respondents, while turtle shell, dandelion, and tu fu ling (dried root and stem of Smilax glabra) each has seven mentions.

3.2 Public online survey

A total of 155 usable inputs were collected from the public survey (demographic information of respondents is provided in Supporting Information: Appendix S2). Although the sample size was small relative to the HK human population, the results revealed some interesting patterns worth discussing. For general TCM use tendency, more than 91% of respondents felt at least half of their friends/relatives support the use of TCM. This aligns with respondents’ self-reported usage, with more than 48% of them using TCM at least occasionally. Most respondents obtained their TCM prescriptions from pharmaceutical shops (57.7%), followed by clinics (46.2%), friends/family (23.1%), online (9.0%), and hospitals (7.7%). Model showed gender was a significant predictor of TCM use tendency, with men being more likely to use it than women (Supporting Information: Appendix S2, GLM 3, p < 0.001) while other factors such as the attitude of surrounding people, monthly income, age, and education level, were not significant.

Only seven respondents reported recent gecko consumption, mostly to treat coughing or for lungs/trachea health benefits. No recent pangolin scale users were found. However, the intentional user group is much larger, as shown in the hypothetical use tendency question (“If you were sick and pangolin scale/gecko appeared in your prescription, would you use it?”, i.e., “Hypothetical use” in Figure 4). No significant difference was found between the proportions of respondents who would hypothetically use pangolin and those who would use hypothetically gecko (Figure 4, z-test, p = 0.07). Demographically, only the age group correlated significantly with gecko hypothetical users (p = 0.006, Supporting Information: Appendix S2, GLM 4), which showed older people were more likely to be potential users than young people. No demographic variable correlates significantly in the pangolin model with hypothetical use (Supporting Information: Appendix S2, GLM 5). Support for the trade was the only significant TPB variable for predicting hypothetical use in both the gecko and pangolin models.

We conducted further GLMs to understand who the trade supporters were and what factors correlated with the supporting attitude. The gecko GLM model showed that men and people who perceived the trade to be legal were more likely to support the trade (Supporting Information: Appendix S2, GLM 6, p < 0.05). The pangolin GLM model showed that perceived trade legality, age, gender, and perceived expensiveness of the products correlated with the supportiveness toward the trade (Supporting Information: Appendix S2, GLM 7). Overall, older people, men, people who perceived the trade to be legal, and people who do not know whether the products are expensive or not are more likely to support the pangolin medicine trade compared to others.

A significant difference was found between the proportion of respondents who perceived pangolin trade as legal versus respondents who perceived gecko trade as legal (Figure 4, z-test, Cohen's H = 0.62, p < 0.001). In parallel, there was a significant difference between the proportion of respondents supporting the medicinal trade of geckos versus the trade of pangolin products, with 19.9% supporting pangolin and 39.7% supporting gecko (Figure 4, z-test, Cohen's H = 0.44, p = 0.0001). Significantly more respondents understood pangolins were threatened compared to gecko threat status, but this was not a significant factor in shaping their consumption intention or supportive attitudes toward the trade according to GLMs.

4.1 Trade legality and quantity

Most sellers had a clear understanding of trade legality related to geckos and pangolins, which is very different from the general public's perceptions. Most sellers perceiving pangolin products as illegal and gecko products as legal. This awareness led to differences in attitudes toward the two products. For example, more hesitation and higher instances of participation withdrawal as different forms of risk aversion were seen with pangolin-related questions compared to gecko-related questions. Knowledge-wise, gecko use was reported in more detail by sellers compared to pangolin use. Most gecko uses were related to the respiratory system and general healthcare, which aligned with the uses reported by the seven recent gecko users in the public survey. Reported uses of pangolin medicine in Hong Kong were different from those reported in a recent study in mainland China (Wang et al., 2020). In Hong Kong, most interviewees mentioned that pangolin medicine was effective for skin-related diseases, while in the mainland, TCM practitioners reported that pangolin was more commonly used to helping milk secretion or treat infertility. The potential explanations for these different applications could be a result of varying prevalent diseases in different regions, with Hong Kong's humid and hot subtropical climate making allergic and acne-related skin symptoms more common compared to the locations in mainland China where other studies have taken place (i.e., Henan and Hainan Province) (Caraballo et al., 2016; Narang et al., 2019). Another explanation could be that these differences may be because the pangolin trade has been practically banned in Hong Kong for more than 10 years. For pangolin medicine to be legally traded in Hong Kong, a product-specific permit needs to be issued by the Agriculture, Fisheries and Conservation Department (AFCD). Information provided by the AFCD upon enquiry showed no legal pangolin trade has been permitted within HK for at least 10 years (AFCD, personal communication). Since TCM practitioners in HK have had no legal access to pangolin medicine for the past decade, they may have less accurate recollections of the medicine or might have restricted its application to minor diseases requiring smaller quantities, such as skin symptoms, thus avoiding the use of large quantities.

Regarding trade quantity, we estimate that the Hong Kong medicinal market could sell over 249,000 geckos a year. This estimate of trade volume comes from a very simple extrapolation based on our survey data, but international trade records showed a similar pattern. Since tokay geckos were uplisted to CITES Appendix II in 2019, legal commercial import and export should have been recorded in the CITES trade database (CITES, 2022). The database shows that HK had a net import of 546,476 wild-origin individuals from November 2019 to December 2021, resulting in an average annual import of around 273,000 individuals. Numerically, this seems to align with our trade estimate. Moreover, trade records from other regions are also on the magnitude of millions of individuals per year (Caillabet, 2013). However, it should be noted that our estimation does not include the pet trade and other potential trade types, which are of significant size according to other studies (Dufour et al., 2022). Therefore, HK's actual gecko trade volume could be higher than our estimation. On the other hand, the CITES records for HK only report imports from Indonesia and Thailand, yet genetic data from Dufour et al. (2022) reveal other potential origins of geckos traded in HK, including mainland China. The absence of mainland China in the CITES trade records suggests these products are likely to be illegal. However, it is also possible that geckos currently for sale in the markets were imported before 2019, as dry gecko medicine can be preserved for a long period of time if stored properly. Furthermore, reported product origin from the sellers includes places like Vietnam, suggesting these are either imported illegally or misreported due to information loss during the trading chain. The mismatch between reported origins, genetic data, and CITES records also highlights that species such as geckos may be neglected in import assessments and calls for enhanced monitoring and regulation to ensure legal and sustainable trade.

Overall, the significant demand for geckos, as shown through our study, highlights the urgent need for conservation attention. For trade to be sustainable, wild gecko populations should be harvested at a rate that will not compromise their long-term survival (Cusack et al., 2020; Hughes et al., 2023), whether to directly supply the market or to stock for commercial farms. Research such as Population Viability Analysis or other models should be used to assess the sustainability of wild gecko populations under this heavy demand (Lacy, 2019). Demand reduction might therefore be a key aspect for gecko conservation. Furthermore, countries like Indonesia have progressively increased their export quotas (currently at 6 million annually) without any assessments of the impact on source populations, and this combined with stockpiling could negatively impact their long-term viability (Nijman & Shepherd, 2015). As long as demand exists, supply will continue, and thus it is essential to develop methods to ensure trade sustainability.

In contrast to the gecko trade, no active pangolin medicine trade was reported in our study which is consistent with a previous study in Hong Kong looking at pangolin consumption behavior (Zhang et al., 2022). More importantly, this absence of pangolin use is significant as it shows that TCM healthcare remains prevalent locally without relying on pangolin scales. This scenario serves as a positive example for places with on-going pangolin medicine trade such as mainland China to encourage for using substitutes and restricting the trade (Wang et al., 2023). However, it needs to be highlighted that HK continues to serve as a transit hub for the international pangolin trade (Zhang et al., 2020). This role as a transit point highlights the urgent need for conservation efforts both within Hong Kong and in the broader international context.

4.2 The importance of setting up legal protection

In Hong Kong, TCM is a healthcare system used by a sizable proportion of the population, with around 17% of residents using it on a regular basis (HKSAR Census and Statistics Department, 2017). We see that the intention to use pangolin or gecko products is mainly influenced by public attitudes toward the trade, which correlates to perceived trade legality. Notably, the awareness of a species’ endangered status is not a significant factor in decision-making process. However, this is not because people have little awareness on the endangered status of these two animals, as shown in our data. Significantly more respondents know that pangolins are very threatened while geckos are not. Therefore, raising people's awareness of species’ endangered status as a common message used in conservation campaigns might not be as effective as assumed. Rather, our results suggest that setting up legal protection and raising awareness of these regulations could be more influential. This is consistent with previous pangolin trade study in mainland China, where perceived trade legality was identified as an important factor in consumption decisions (Wang et al., 2020).

Setting up legal protection is particularly important for species like the gecko which despite an abundant wild population, might face significant threats from unsustainable trade that could cause rapid population decline. The yellow-breasted bunting is a pertinent example illustrating this risk. Once a superabundant species, the bird experienced drastic population collapse due to intense and unsustainable trapping (Kamp et al., 2015). Therefore, prompt and up-to-date sustainability assessment for species involved in high-volume wildlife trade is critical (Hughes et al., 2023). The resulting policy decisions and conservation actions can help to prevent species from becoming endangered, which is of comparable importance to saving species from already endangered status.

4.3 Discrepancies between trade supporters and hypothetical users

Our study also highlights the notable discrepancy between public attitudes toward trade and hypothetical usage behavior. Although the results show more respondents oppose the pangolin medicinal trade compared to gecko trade, this opposition does not translate into a statistically significant difference between hypothetical user proportions of the two products. In other words, despite stronger opposition to the pangolin trade, the hypothetical willingness to use pangolin medicine is not statistically lower in proportion than that for gecko medicine. This disconnect between attitude and behavior is also observed in other studies (Balmford et al., 2017; Nishizawa et al., 2021). It suggests factors such as legal and conservation status, which differentiate the two animal groups, are not the sole predictors of people's usage intentions, as shown in the model results. There might be important factors not included in our adapted TPB model influencing the decision-making process, for example, adherence to medicinal advice (Vermeire et al., 2001). This aspect warrants future research attention to better understand the complex dynamics between public attitudes, legal status, and actual usage behavior in the context of the wildlife medicinal trade.

In parallel, the demographic profiles of trade supporters and intentional users differ in both pangolin and gecko models. Pangolin trade supporters tend to be older men, while gecko trade supporters are predominantly men regardless of age. However, when examining hypothetical gecko users, age emerges as a significant factor that older people might be more likely to use, but gender is not influential anymore. The different demographic profiles between trade supporters and intentional users again highlight the gap between attitude and behavior. This divergence suggests that conservation programs and initiatives need to be tailored according to their target outcomes. Those aiming to change public attitudes might need to focus on a different stakeholder group compared to those designed to alter consumption behavior (Jones et al., 2019).

In conclusion, the sustainability of wildlife-derived TCM medicine needs more conservation attention, especially for commonly used medicinal species. While species like the tokay gecko might still be abundant in the wild, the high market demand and trade volume present some worrying signals for their future. Systematic sustainability assessments of commonly traded TCM ingredients, including both plants and non-mammal animals, are important for avoiding “invisible extinction” (Eisenhauer et al., 2019). The disparity between attitudes, behaviors, and the associated differences in user profiles highlights the complexity of human behavior and underscores the significance of evidence-based conservation. Overall, sustainability is a shared objective for both the TCM and conservation communities. Collaboration between the TCM community, conservation practitioners, and relevant government sectors is crucial to guarantee the long-term survival of both endangered species and the conservation of TCM practices and traditions.