2.1 Field Surveys

The study site was selected within the known distribution range of G. badenii, viz. around the type locality in Ba Den Mountain, Tay Ninh Province, southern Vietnam based on previous observations, literature, and interviews with local people. This granite-rocky mountain is up to 986 meters above sea level (asl), covered by evergreen forests (Supporting Information S1: Figure S1A) and characterized by a monsoon sub-equatorial climate with constant annual temperatures, but with a distinct climatic difference between a dry and a wet season with heavy rains (General Statistics Office of Vietnam 2016; Ngo et al. 2018; Sterling et al. 2006). The species is frequently observed on granite rocks, cliffs and caves, as well as often found on trunks or small branches at night.

Field surveys were carried out by two survey teams in Ba Den Mountain during the early wet season in August 2024. Each team consisted of two researchers marking, collecting and noting data and one local guide catching animals. Following the suggestion of local rangers, we mainly conducted day excursions (11 days) between 14:00 h and 18:00 h, when G. badenii is resting and hiding inside rocky caves, but actively react to prey (e.g., insects). To lure animals out of their hiding spots, we used a long stick with a prey organism (e.g., beetle) attached. Furthermore, four-night excursions were undertaken after sunset between 19:30 and 24:00 h. During the night excursions, we captured animals by hand or by using the same noose-stick without a prey organism because well-fed animals tend not to react to prey at night. A total of 11 transects were surveyed, whereof nine transects (T1–T9) were selected for day excursions and two different transects (T14–T15) were selected for night excursions (Supporting Information S1: Figure S1C). Each day-transect covered a forest area of about 200 m², characterized by a suitable microhabitat of granite rocks, cliffs, and caves and the mandatory presence of Ficus lacor trees, as geckos feed on ripe fruits too (Supporting Information S1: Figure S1). Night transects were set at different elevations (e.g., 200, 400, 500, and 700 m asl.) along the fire-rescue routes in Ba Den Mountain, containing similar microhabitat structures. Coordinates and elevations for each transect were recorded with a GPS Garmin 64. Individuals were released at the capture site after taking measurements, photos, determining the sex of each gecko, and marking with a permanent pen to identify re-captured animals.

2.2 Population Analysis

The total number of individuals per transect was counted based on direct observation. We further calculated rates of captured geckos per hour (inds/hour) and adults per hour (adults/hour) per person as the proxy for catch per unit effort. To assess the demographic structure, geckos were categorized into three age classes based on their snout–vent lengths (SVL ≤ 60 mm = juvenile, 60 mm < SVL < 75 mm = sub-adult, and SVL ≥ 75 mm = adult). Each gecko was measured with a digital slide-caliper to the nearest 0.1 mm. The sex in subadults and adults was determined by the presence of precloacal pores in males, and their absence in females (Figure 1). For testing significant differences in the age and sex structures between day and night that may give insights into sex or age-related activity patterns, we used a Chi-square test.

2.3 Sexual Dimorphism

The sexual dimorphism index (SDI) was calculated to identify the pattern of sexual size dimorphism (SSD) using the revised formula that was first suggested by Lovich and Gibbons (1992), in which SDI = (mean size of males/mean size of females) – 1. Accordingly, SDI is a positive value when males are the larger sex (SDI > 0) as a male-biased pattern, a negative value when females are the larger sex (SDI < 0) as a female-biased pattern and zero when the sexes are equal in size (SDI = 0) as unbiased pattern. Student t-tests were performed to determine the difference in the snout-vent length (SVL) and weight between males and females. Inter-sexual color variation of G. badenii was further examined based on photos of each individual taken in the field. For all tests, we applied p-value = 0.05. Statistical analyses were carried out in R v 3.1.2 (RStudio Team 2018).

3.1 Population Status

The presence of Gekko badenii was documented in all surveyed transects from the Ba Den Mountain, at various elevations from 60 to 700 m a.s.l. A total of 883 geckos were captured by the research team and local hunters during 2660 survey minutes, with an average capture rate of about 20 individuals per hour by a person (Table 1). During the day, our research team caught 565 geckos in 1190 min, with a rate of 28.49 individuals per hour and person, and local hunters caught a total of 238 individuals in 660 min with a lower rate of 21.64 inds/h and person (Table 1). Contrarily, the average rate of 20 adults/h and person captured by senior hunters was significantly higher than the rate of 16.64 adults/h and person, accomplished by our research team (Table 1). Among day-surveyed transects, a senior hunter collected the highest number of 90 individuals (including 85 adults) with rates of 25.71 individuals and 24.28 adults/h and person at transect T12, while our research team captured a lower maximum number of 55 individuals at transect T1. The highest capture rates were up to 72 individuals at transect T2, and 42 adults per hour and person at transect T4 (Figure 2; Table 1). We further noted that rates of captured geckos at elevations of above 600 m a.s.l in transects of T8 and T9 (10–13 inds/h and person), were significantly lower than those in transects at lower elevations (Figure 2; Table 1). Over a period of 810 min at night, 80 individuals of G. badenii were captured in total. The average night-capture rate was estimated notably low, at 5.93 individuals per hour and person (Figure 2; Table 1).

Considering the age structure, during the day surveys, the percentage of adults accounted for 76.39%, followed by subadults (20.37%) and juveniles (3.24%). During the night surveys, the age structure significantly differed from the day surveys with adults accounting for 53.85%, followed by juveniles (25.64%) and subadults (20.51%) (Chi-square = 54.332, df = 2, p-value = 1.592e-12 < 0.01; Supporting Information S1: Figure S2). Regarding the sex structure, we captured more female individuals, accounting for 59.09% (n = 195) during the day surveys and 77.59% (n = 33) during the night surveys (Supporting Information S1: Figure S2). The dominance of female individuals was consistently recorded across all survey transects (Table 1).

3.2 Sexual Comparisons

Regarding morphology, the mean SVL of adult males (93.0 ± 0.9 mm) was significantly higher than that of females (82.2 ± 0.3 mm) (Student-test, p-value < 0.001, Supporting Information S2: Table S1). In addition, our findings revealed that males (17.8 ± 0.54 g) are significantly heavier than females (12.2 ± 0.2 g) on average (Student-test, p-value < 0.001). The maximum recorded weight for males was 37 g, whereas the heaviest female weighted 22.34 g (Supporting Information S2: Table S1).

3.3 Interviews With Local Community Members

According to our interviewees from the local community at Ba Den Mountain, five local persons were known to rely on hunting the target species of G. badenii as their primary occupation. All five persons had more than 5 years of experience, and those with the longest tenure in the profession have been doing it for up to 20 years. Besides Golden gecko, the snail species, Cyclophorus saturnus, is also harvested by these hunters as supply for local food consumption. Those who hunt geckos and snails full-time, earn a maximum monthly income of approximately 4–7 million VND (160-290 USD) during the rainy season and up to 9 million VND (370 USD) during the dry season. Recently, the number of leisure and new hunters increased to 12 local persons, driven by rising unemployment after the covid epidemic period. However, most of them have been only hunting geckos and snails in smaller numbers for personal food use.

In the past, hunters employed a team-based approach to catch golden geckos, with one person using a noose-stick while the others used leaves to mimic insect-like sounds, effectively luring the geckos. Currently, beetles of the Scarabaeidae family are usually used as prey attached to the noose-stick to attract and catch the geckos (Figure 3A). Hunting generally takes place between 7:00 am and 17:00 pm, though some hunters prefer to hunt only in the afternoon from 13:00 pm to 17:00 pm. Another common method involves setting up a bucket trap at night, typically around 10 liters in size, without its lid, and with local ripe fruits of Ficus lacor as bait inside the bucket (Supporting Information S1: Figure S1B). Local hunters apply lubricant oil on the sides of the bucket trap to prevent geckos from escaping. Geckos are described to be more active and occur in higher densities on granite rocks around the trees during the ripe fruit season of F. lacor.

According to the interviews, the previous average stick-hunting yield was about 4 kg of geckos (about 250 adults) in half a day by two senior hunters, depending on the weather conditions, with a significantly lower yield during heavy rain or extreme heat. Currently, the yield of G. badenii from Ba Den Mountain has markedly declined with usually 1.5–2 kg of geckos (about 100–150 adults) per half a day. The yield of catching geckos with buckets can rise up to 0.5 kg of geckos per night during the dry season, while only 5–6 individuals can be usually caught in bucket traps per night during the rainy season. Thus, hunters only applied the bucket trap in the dry season.

Primarily live adults and rarely sub-adults of G. badenii are selectively targeted by hunters from the wild. Subsequently, these can be sold directly to local restaurants at higher prices, but they are often purchased by local dealers. After processing, gutted geckos are typically sold as prepared food and stored in refrigerators, either to local restaurants or transported to restaurants in Binh Duong Province and Ho Chi Minh City (Figures 3 and 5). Additionally, we have noted that some hunters and traders advertise live, processed, and cooked geckos online for food consumption. In contrast, we did not record that the species is locally used as pet. We visited and interviewed eight local restaurants at Ba Den Mountain, four of which offered dishes containing G. badenii. Previously, the price paid by middlemen to local hunters ranged between 90,000 and 110,000 VND (3.5–5 USD) per kilogram of geckos. Meanwhile, the current price has increased to 170,000 VND (7 USD) per kilogram. Likewise, prices paid by local restaurants to middlemen increased from 200,000 to 250,000 VND (8–10 USD) per kilogram. The current prices for dishes containing prepared geckos at restaurants range from 350,000 to 700,000 VND (14–28.5 USD) per kilogram (Figure 5).

3.4 International Pet Trade

According to the LEMIS database of imports into the United States (US), a total of 11,308 live individuals of golden geckos labeled as G. auratus, in 117 shipments were imported between 2018 and 2023 (Table 2). All shipments were reported for commercial purposes with a total profit of 11,363 USD (Table 2). The import of G. badenii into the US fluctuated annually with a mean number of 1,885 live individuals from an average of 20 shipments, amounting to 1894 USD per year (Figure 4; Table 2). All exported individuals of G. badenii were identified as originating from Vietnam, some of which were related to transit countries before shipping to the final destination of the US, including one shipment to Indonesia, three to Korea, three to Canada and one to France. The number of 10,253 individuals (90.7%) from 109 shipments were recorded as wild caught, whereas only 1055 geckos (9.3%) from 8 shipments were recorded as bred in human hands (Figure 4). The price for G. badenii varied from 0.7 to 8 USD per animal, with an average of 1.16 USD/ind (Table 2).

Our investigation on different online platforms of reptile trade recorded the presence of G. badenii on 28 websites of pet shops advertising for the commercial sale located in the US, United Kingdom of Great Britain and Northern Ireland, Romania, and Germany. Prices for sale of G. badenii ranged from 3.5 to 79.99 USD per individual, referring to an average of 28.5 USD/ind (Figure 5; Supporting Information S2: Table S2). The majority of offers related to animals that were wild-caught or whose origin was not specified. During two visits at the large reptile fair Terraristika in Hamm, Germany in March and September 2024, no Gekko badenii specimens were spotted for sale on the desks (pers. obs.).

3.5 Habitat Degradation

Ba Den Mountain is one of the most famous spiritual tourist destinations in southern Vietnam, situated approximately 100 km from Ho Chi Minh City, attracting millions of visitors each year. Through direct observations and interviews with local people, the development of cable car tourism in previous years, infrastructure for tourist locations on the mountain, and trekking visitors leaving garbage were identified as anthropogenic activities which degrade the forest and may impact the microhabitat quality of G. badenii at the Ba Den Mountain. Additionally, interviews with forest rangers revealed that small-scale forest fires have been sometimes occurring in the mountainous area during the dry season due to some local hunters using fire to obtain honey.

4.1 Local Exploitation

Our surveys recorded a maximum of 90 individuals of G. badenii within a 200 m² area, indicating a potential density of up to 450 individuals per square kilometer. Local hunters reported that up to 200 adult geckos, weighing approximately 3 kg in total, can be captured within an area of less than 400 m², referring to a density of 500 inds/km². In comparison, the average population density of a closely related species of G. canhi was calculated to be significantly lower with approximately 16.6 inds/km² in northern Vietnam (Ngo et al. 2023). Similarly, other endangered geckos in Vietnam were found to occur at much lower densities, such as Cnemaspis psychedelica with an estimated density exceeding 120 individuals/km² in suitable habitats (Ngo, Nguyen, et al. 2016), Goniurosaurus catbaensis in Ha Long Bay reaching up to 60 individuals/km² (Ngo, Ziegler, et al. 2016; Ngo, Le, et al. 2019), and G. huuliensis with an estimated average density of 3.54 individuals/km². Therefore, based on the results of this study G. badenii can be described as a gecko species which occurs in high population densities in Ba Den Mountain. However, the present study indicates that the wild population in Ba Den Mountain has drastically declined during recent years. According to information provided by local hunters, some years ago, hunters were still able to capture an average of about 250 adult geckos in half a day. Nowadays, the yield has been reduced to about 100 to 150 adults in half a day. If this declining trend further continues, it may threaten G. badenii in Ba Den Mountain in the future. With the monthly income per hunter of approximately 4–9 million VND (160-370 USD) for hunting animals of G. badenii, we estimate that a hunter can capture approximately 1500 to 3000 individuals per month. The main driver for local declines seems to be the targeted harvest for food consumption.

On average, the research team caught more individuals per person and hour than an experienced full-time hunter. Nevertheless, the hunters captured more adults on average than the research team. These differences in capture rates may be due to the fact that, according to the interviews with local hunters, smaller-sized juveniles and sub-adults are not intentionally caught, and they are often released or completely disregarded during hunting. Our observations revealed that on survey days lasting less than 90 min, the catch rate per hour was quite high, reaching up to 72 individuals, except for routes at elevations above 700 m (Table 1). In contrast, the catch rate on days with survey durations of 120 min or more was lower (Table 1). Within an area limited to less than 200 m², the encounter rate is likely to decrease after 2 h, once a large number of individuals has already been captured.

Online investigation and interviews indicated that the harvest of G. badenii is mainly intended for supplying food products in Vietnam, while no evidence was recorded that the species is used in the domestic pet trade. However, it may be assumed that the domestic pet trade with G. badenii still occurs in access-limited trade groups or by private transactions, and the quantity for these sales may be relatively small. Among gecko species in general, the herein identified main use of G. badenii as food for domestic consumption is relatively uncommon, as most geckos are generally considered too small to provide proper meat compared to larger reptile species in Vietnam. Regarding the morphology, G. badenii is characterized by a maximum snout-vent length (SVL) of 117.4 mm, which is significantly larger than that of C. psychedelica (maximum SVL: 75.4 mm) but lower than G. huuliensis reaching a maximum SVL of 132.2 mm (Grismer et al. 2010; Ngo et al. 2021). Overall, the size and weight of G. badenii (with a maximum of 37 g) are much lower than those of lizards commonly harvested for food consumption in Vietnam, such as the agamid species Physignathus cocincinus (SVL: 280 mm, and weight: 600 g) (Ngo et al. 2024). In contrast, other Vietnamese geckos (e.g., Cnemaspis psychedelica and five Goniurosaurus species) have been harvested for the main purpose of supplying the pet trade due to high prices that can be achieved there (Auliya et al. 2016; Ngo, Nguyen, et al. 2019). For example, 4–5 USD were paid to local hunters to collect an individual of Goniurosaurus species, and which was sold in local pet shops with approximately 35 USD (Ngo, Nguyen, et al. 2019). Individuals of C. psychedelica even fetched extremely high prices up to 2500–3000 Euro/pair on internet platforms (Auliya et al. 2016). Meanwhile, local hunters were only paid 6–10 USD/kg of G. badenii (meaning 0.1–0.17 USD/ind). The substantial commercial profits derived from the purpose of pet trade for small-sized Goniurosaurus species and C. psychedelica due to their beautiful color patterns and rarity in the wild likely explain why they are not exploited for food, in contrast to G. badenii, which is specifically targeted as a local food source and can be caught in larger quantities. The fact that G. badenii has been observed eating ripe fruits could also contribute to its commonness, and restaurants use this characteristic to advertise this as a sign of healthy and fresh food to consumers.

We were not able to detect a specific large demand for G. badenii in the European market, so that we consider the species as some kind of bycatch species, which is included in international shipments for the pet trade if available. The LEMIS data in the United States however, recorded imports of more than 11,000 live individuals of G. badenii between 2018 and 2023, with the majority of imports (90.7%) recorded as wild caught animals from Vietnam. Despite live animals of G. badenii may be only sold domestically with the quantity unit of kilogram for food advertising, local traders may purchase them for overseas export shipments. Our surveys on international online platforms recorded the lowest price at 3.5–5 USD and the highest at 79.99 USD per individual of G. badenii, with an average price of approx. 28.5 USD/ind. Meanwhile, with an average selling price of 0.17 USD/ind in Vietnam, and 1 USD/ind imported to the US, the profit of international retail trade is very high, with about 41-fold of shipment prices, and approximately 400-fold of the prices gained locally. The low prices for wild-caught animals in the local market in Vietnam could explain why there appears to be no significant trade in captive bred individuals, as captive breeding may not be economically viable. The highest price of 79.99 USD was recorded for bred offspring from an online retailer, while animals declared as wild-caught were already available at much lower prices. The relatively low prices for the sale of G. badenii in the international pet trade may be due to its somewhat less striking appearance, the availability in large quantities from the wild, and/or successful breeding in human hands.

4.2 Population Structure Shifted by Sexual Dimorphism

Regarding the sex ratio of G. badenii in Ba Den Mountain, the dominance of female individuals was recorded across all survey transects. This sex ratio imbalance naturally occurs in many reptile species and is linked to the territorial competition between males leading to a higher male mortality (Olsson and Shine 1996; Shine 1989; Stamps and Gon 1983). In this study, we recorded that males are significantly larger and heavier than females (Supporting Information S2: Table S1). An inter-sexual color variation was further observed. In particular, males show a bright yellow color at the dorsal body and tail (Figure 1A), while females present a brown dorsal body and light gray tail (Figure 1B). These observations approve the distinct sexual dimorphism in G. badenii following the male-biased pattern. According to interviews with local hunters, they only selectively catch adults of G. badenii and prefer light yellow individuals which can be identified as adult males because of their bigger size for food consumption yielding higher prices in sale. Observations and interviews with local hunters in this study in line with previous observations by Hofmann (2006) from the pet trade, noting an extremely biased ratio of 1:12 or smaller for females: males, respectively, suggest that the apparent sexual dimorphism of G. badenii leads to the targeted harvest of males for both the pet and consumptive trade. As in the case of the agamid species Physignathus cocincinus, we also consider the male-biased sexual dimorphism in sizes and colors of G. badenii as an important factor which might directly cause a decrease in the abundance of males (Gewiss et al. 2020; Ngo et al. 2024; T. Q. Nguyen et al. 2018). A continuous selective harvest could drastically shift the sex ratio of the wild population towards a significant reduction of adult males and thereby an unnatural imbalance.

4.3 Conservation

Currently, G. badenii can still be estimated as relatively abundant on Ba Den Mountain. However, under severe impacts of long-term harvesting (e.g., for food consumption and pet trade) and habitat fragmentation and degradation (e.g., due to road construction, tourism, and forest fires), the wild population of G. badenii has declined sharply. Demographic assessments of other endangered Vietnamese lizards revealed a drastic reduction in their wild populations, particularly Shinisaurus crocodilurus in Vietnam and China with up to 90% declines, even local extirpation in Goniurosaurus luii and G. araneus, and an increased risk of extinction in Cnemaspis psychedelica, mainly driven by over-harvesting and habitat loss (Grismer et al. 1999; Huang 2008; Ngo, Nguyen, et al. 2016; Ngo, Le, et al. 2019; Stuart et al. 2006; van Schingen et al. 2014, 2016). To prevent unsustainable exploitation and habitat degradation, concrete measures of conservation are highly essential to protect the wild population of G. badenii at its type locality in Ba Den Mountain. Globally threatened species, G. badenii (Endangered) together with other geckos such as Cyrtodactylus nigriocularis (Critically Endangered) and C. badenensis (Vulnerable) should be considered as high-profile flagship species to highlight the value of biodiversity in Ba Den Mountain. And, we have evidence that there occurs more so far overlooked, seemingly micro-endemic biodiversity in there, making the site even more important for upgraded conservation measures. This should be prioritized with regard to conservation policies to protect forest habitats of Ba Den Mountain rather than favoring only economic development. To address the overharvest of the species, banning the entire harvest for food use and pet trade of G. badenii may not be the most effective solution, as it has a significant role in supporting the livelihoods of some local people. Instead, encouraging the participation in forest patrol teams or providing additional jobs in relation to local tourism activities could provide sustainable livelihoods for hunters. A proper adaptive management that allows certain regulated and limited harvest and sustainable use of the species based on quotas, designated harvest areas or size restrictions should be implemented by considering the livelihoods of local people and incentivizing the long-term protection of the species. Alternatively, certain quotas for the sale of wild specimens in restaurants could be discussed to thereby increase prices for dishes from wildlife. To achieve that the species needs to be included in the national governmental decree as a protected species. Furthermore, a CITES Appendix III listing could be considered to collect reliable data on the extent and impact of international trade.

We further intend to partner with the Ba Den Mountain management board to implement a program raising community awareness and engaging tourists on the importance of local biodiversity. Recently, this involved designing and placing signboards and leaflets at tourist sites, to introduce the conservation status of G. badenii and other threatened geckos from Ba Den Mountain. As part of concrete measures against habitat loss, regular patrols should be conducted by rangers to detect forest fires timely before spreading in Ba Den Mountain. In an effort to reduce plastic waste from tourist activities, waste bins were recently placed along trekking routes and tourist sites. Some parts of the forest encompassing suitable habitats for the species should be entirely protected from human access. Establishing in-country breeding programs could aid building up reserve populations, which can facilitate future releases if needed to support the recovery of disturbed wild populations. Furthermore, it could aid to supply demand for the species and contribute to local livelihoods.