Welfare and Management of Civets in Civet Coffee Tourism Plantations

ABSTRACT Civet coffee is produced with coffee beans that have passed through the digestive tract of civets which are then processed for drinking. Outlets, known as civet coffee tourism plantations are popular tourist attractions in Bali, Indonesia. These outlets keep caged civets for public viewing and sell civet coffee. We assessed civet welfare in 29 civet coffee tourism plantations to assess husbandry factors, the five freedoms and body condition scores. We also assessed civet welfare in eight wildlife markets where civets are sourced for civet coffee tourism and calculated turnover of civets. We found 99 civets in civet coffee tourism plantations and 136 in markets, where. Annually between 400 and 800 civets are sold. Welfare standards do not meet the Indonesian laws and regulations. Diet, hygiene, presence of water and mobility affected body condition scores and no civet met all of the five freedoms. Enforcement of welfare and trade laws is needed to improve civet welfare, and to curb the illegal civet trade. Tourists frequenting civet coffee tourism plantations should be made aware of the low welfare standards to inform their choice about supporting this industry.


Introduction
Millions of animals are used in tourism every year globally (Fennell, 2013).Tourism has grown over the past two decades, including wildlife tourism, which, depending on the definitions used, totals 30-40% of all tourism worldwide (Moorhouse et al., 2015;Tapper, 2006).Wildlife tourism in its broadest sense deals with all tourism aspects centered around observing and engaging with animal and plant life in a local setting, but in a narrower sense focuses on interacting with wild animals in captive or semi-captive settings.When wildlife tourism includes a captive component, the welfare of the animals can be compromised due to the failure of the captive environment to meet basic biological needs (Tapper, 2006).Moorhouse et al. (2015) estimate that global wildlife tourism attractions negatively impact the welfare of 120,000-340,000 wild animals worldwide.Animal welfare has long-received promotion and scientific research, yet animal welfare in a tourism setting is less studied (Hughes, 2001).
The contradictions between welfare, captivity and tourism play themselves out in aquariums, safari parks, zoos, but also in commercial tourism enterprises such as elephant riding, walking with lions (Panthera leo) (Schroeder, 2018), swimming with dolphins, or having coffee with raptors in owl cafes (McMillan, Dingle, Allcock, & Bonebrake, 2021).Previous studies on nonhuman animal welfare in tourism have shown compromised welfare (Carder et al., 2018;Fennell, 2022).Reuter and Schaefer (2016) found evidence of poor lemur welfare in Malagasy hotels, where lemurs are used as photo props and to attract tourists and highlighted the illegality of the trade in lemurs in supplying the hotels.A similar case may occur in Indonesia where civets are taken from the wild to meet the demand from the civet coffee industry and where tourists can observe the process of civet coffee production (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014;Shepherd, 2012).We here focus on civets in civet coffee tourism plantations in Bali, Indonesia's most popular tourist destination receiving over 15 million tourists in 2018 (BPSPB, 2023).Civet coffee tourism plantations offer a unique experience where visitors can observe the process of civet coffee being roasted in a naturalistic setting surrounded by the plants and crops that are grown on the island.They have been referred to as "civet coffee farms", "tourist-oriented civet coffee plantations", "civet coffee farms", and, in Indonesian as "civet coffee agricultural tourism"; the key difference with commercial civet coffee farms, where up to 100s of wild-caught civets are kept to produce civet coffee (Thi et al., 2022) is that the former is open to the public and creates revenue from tourists visiting the facility and purchasing civet coffee onsite, and the latter is closed to the public and revenues come from the wholesale of civet coffee.The facilities in Bali are set in a show-case agricultural setting, with examples of local edible or otherwise useful plants (bananas, ginger, coffee, etc.) grown on site, and civets in cages to allow tourists to understand the process of civet coffee production; hence, we refer to them as civet coffee tourism plantations.
In this study, we aimed to determine numbers and species of civets present in civet coffee tourism plantations in Bali and in the wildlife markets of Bali, Lombok (directly to the east of Bali) and easternmost Java (directly to the west of Bali).With a 2018 quota of just 25 Southeast Asian palm civets allowed to be sold domestically in Indonesia (i.e., 10% of the total annual permissible harvest: KLKH (Kementerian Lingkungan Hidup dan Kehutanan), 2018), we hypothesized that we would see few to no civets offered for sale in the wildlife markets.We aimed to evaluate the animal welfare and living conditions of civets in civet coffee tourism plantations and in wildlife markets.Poor animal welfare in the civet coffee tourism plantations has been shown in previous studies (Carder, Proctor, Schmidt-Burbach, & D'Cruze, 2016;D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014;Thi et al., 2022).Due to domestic laws on animal welfare and regulations on the captive management and husbandry of civets (Sulaiman, 2015) we hypothesized that civet welfare would be sufficient both in plantations and markets.We predicted that lower husbandry scores would result in a higher deviation from optimal body condition scores.We used this information to make suggestions regarding which management factors have the greatest impact on civet welfare.

Study species and civet coffee production
Civet coffee, known locally as kopi luwak, has been described as "the world's most expensive coffee" and is produced predominantly by palm civets, i.e., Asian palm civet (Paradoxurus hermaphroditus, in South Asia and Indochina) and Southeast Asian palm civet (P.musangus, in mainland Southeast Asia, Sumatra and Java).To a lesser extent other species, such as the Philippine palm civet (P.philippinensis) in the Philippines and Borneo, binturongs (Arctictis binturong), small-toothed palm civets (Arctogalidia trivirgata) and masked palm civets (Paguma lavarta) are also used in the production of civet coffee (Roberts, 2020).Coffee cherries are consumed by the civet and then harvested from the feces (Jumhawan, Putri, Bamba, & Fukusaki, 2016).Traditionally, the feces of wild free-ranging civets were used for civet coffee production as it was suggested that the wild civet would only choose the ripest beans.The digestive process of the civet breaks down the tough outer body of the coffee cherry, after which the coffee bean is partially digested in the civet's gastrointestinal tract and the beans are taken from the feces for further preparation (Cahill, 2017).During digestion, digestive enzymes and gastric juices cause surface micro-pitting of the coffee bean, breakdown of storage proteins and results in a darker color bean (Marcone, 2004).The coffee is claimed to have a smoother and less bitter taste when produced this way (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014), with a higher caffeine content (Chan & Garcia, 2011) and therefore has received the reputation of being of superior quality (Jumhawan, Putri, Bamba, & Fukusaki, 2016).This supposed quality combined with the scarcity of the wild civet feces has resulted in high prices for civet coffee (Cahill, 2017).
The traditional method for producing civet coffee poses no welfare implications or conservation threats, but rather creates a mutual beneficial co-existence with an animal often considered a nuisance (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014;Nakashima, Inoue, Inoue-Murayama, & Sukor, 2010).The lucrative nature of the civet coffee business has resulted in caging wild-caught civets to provide manufactured coffee to supply a global economy.A diet of just coffee cherries is considered poor nutrition for civets, which are omnivorous and have a varied diet in the wild and is thought to lead to early fatalities (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014).Poor welfare of these viverrids has been documented in both civet coffee farms (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014;Thi et al., 2022) and in civet coffee tourism plantations (Carder, Proctor, Schmidt-Burbach, & D'Cruze, 2016;Roberts, 2020).

Civet coffee tourism plantations
In May to July 2018, we visited 29 civet coffee tourism plantations in Bali.These are open to the public on a walk-in basis and civets are openly displayed as part of the attraction.We used observational data collection methods to record number and species present and to conduct welfare assessments.We assessed the quantity of civet coffee in the onsite shops by counting the number of bags of civet coffee (grounded and beans) and recording the amount in each package (typically 50 g or 100 g). Thi et al. (2022) established that in captive settings, on average, Southeast Asian palm civets can produce 9.82 kg of civet coffee a year and we used this number to explore the relationship between the amount of civet coffee sold in the shops and the number of civets that we observed in the civet coffee shop.

Wildlife market surveys
Between January and July 2018, we conducted surveys of open trade in wildlife markets in Bali (Satria, Sanglah and Beringkit), Lombok (Mataram) and the eastern most part of Java (Jember, Probolinggo, Bondowoso and Banyuwangi).We chose these wildlife markets due to their close geographical location to the civet coffee tourism plantations.Following the methods described by Nijman et al. (2014) we surveyed all wildlife markets three times, apart from Satria and Beringkit, which we surveyed five times.We noted the number of civets that were openly offered for sale, identified the species, and recorded the age of the civets (adult, non-adult).We requested asking prices and origin from the traders.Prices were obtained in Indonesian Rupiah; these are converted into USD using the July 2018 exchange rate (1 USD = 14,411 IDR).

Welfare assessments
The Indonesian Ministry of Agriculture provides guidelines for the husbandry and welfare of civets in the civet coffee industry (Sulaiman, 2015), and the keeping of all animals should abide by the Penal Code of 1918, which does not allow for withholding necessary sustenance from animals in captive settings.We examined if these requirements were met via three measurements that did not require physical contact with the animals and that we could obtain through visual inspection: presence of absence of the five freedoms, use of a civet husbandry table and body condition scoring.Regarding the five freedoms (Albright, 1998;Webster, 2001), we excluded the "freedom from pain, injury and disease" as we could not confidently determine this via observation alone.For the remaining four freedoms, we used binary yes/no indicators for 1) "freedom from hunger or thirst" -access to appropriate food and water, 2) "freedom from discomfort" -ability to control exposure to elements and access to comfortable area to rest 3) "freedom to express normal behaviour" -ability to move several body strides, to climb, hide and stay clean and ability so socialize with conspecifics/hide from conspecifics 4) "freedom from fear and distress" -ability to hide from stressors including the public.As another welfare measure, we used the civet husbandry factor table developed by Carder, Proctor, Schmidt-Burbach, and D'Cruze (2016), including the same nine factors with scoring guidelines (Table 1).This method was also used by Schmidt-Burbach, Ronfot, Srisangiam, and Sestak (2015) to assess welfare of animals in tourist venues in Thailand.The husbandry factors are scored on a 5-scale system, 0 being the lowest score and 4 the highest.The factors include mobility, hygiene, surfaces, shelter, noise, food, water and social interaction.Each civet in the civet coffee tourism plantations and in the wildlife markets was given an individual welfare score.
When civets could be clearly seen, we gave them a body condition score using a scale from 1 to 9 with 1 being emaciated and 9 being morbidly obese (Figure 1).This scale is similar to the body condition scoring used in equine husbandry and care (Pritchard, Lindberg, Main, & Whay, 2005) and to evaluate the health of domestic dogs, horses, and cats (Burkholder, 2000).To put these scores in perspective, we compared them with a collection of 50 camera trap photographs of wild Southeast Asian palm civets in our long-term study site in Java (Nekaris et al., 2020).

Statistical analysis
We examined whether the number of Southeast Asian palm civets we recorded in the wildlife markets agreed with the annual harvest quota set by the Indonesian government.Based on repeat visits (with intervals between 9 and 42 days) to the same shops we calculated the turnover (i.e., the proportion of animals that were either sold or had died between subsequent visits).From the relationship between time and turnover, we were able to calculate annual number of civets sold in each individual shop (c.f.Bergin & Nijman, 2019).It was not possible to calculate turnover when in between surveys a new consignment of civets had arrived and the number of civets in the shops had increased, and hence, our calculations give a minimum estimate.We summed this for the eight wildlife markets to determine the overall annual number of civets sold in these markets.
We analyzed the combination of two welfare evaluation methods to determine how husbandry factors can affect body condition.We ran generalized linear mixed models (GLMM) to assess the effects of predictor variable (husbandry factors: mobility, hygiene, surface, shelter, noise, food, water and social interaction) on body condition score of 94 civets in markets and 70 civets in civet coffee tourism plantations.The data included four viverrid species, and we considered all viverrid species as a mixed factor in the analysis.We used the logistic family of distribution in both models, which was chosen based on the QQ-Plot of residuals.We tested combinations of predictor variables in the models, from the simplest one (no effect of any predictor variable) to the most complex one (all predictor variables included).Then we used Akaike Information Criteria (AIC) to select the best-fitted models in which the lowest AIC model was selected.For both market and civet coffee tourism plantations, we found no differences between AIC values (ΔAIC) of all remaining models and the model with the lowest AIC equal or higher than 2. We conducted all modeling in R 3.5.1 software (http://www.R-project.org/),using gamlss R-package and tested collinearity in GGally R-package, and we used the function Rsq to calculate the generalized R 2 .No collinearity was found between variables.

Ethics
Part of the research was underpinned by a Memorandum of Understanding between Universitas Gadjah Mada and Oxford Brookes University.In both our United Kingdom and Indonesian Institutes, we did not require institutional permission for observational research in wildlife markets; they were, added to Oxford Brookes University's Register of Activities Involving Animals (2016)(2017)(2018)(2019).Informal discussions with traders followed the ethical guidelines proposed by the Association of Social Anthropologists of the UK and Commonwealth.

Civet coffee tourism plantations
We recorded 99 civets in 29 civet coffee tourism plantations on Bali.With 93 individuals, the Southeast Asian palm civet was the most common species, with many fitting the description of the Javan subspecies, P. m. javanicus.We furthermore recorded four masked palm civets (Paguma larvata) and one small-toothed civet (Arctogalidia trivirgata).We found an average of 13.1 kg of packaged civet coffee in 28 of the 29 shops, with a total of 446 kg.

Wildlife markets
We recorded 136 civets, all Southeast Asian palm civets, for sale in four of the eight markets surveyed.Civets were found for sale in Denpasar, Bali (109 civets in Satria, 8 civets in Beringkit), Mataram, Lombok (8 civets) and Bondowoso market, Java (6 civets).Most dealers had one, two or three civets for sale, but one trader in Satria regularly had 20+ civets for sale.Other small carnivores we recorded include four Javan ferret badgers (Melogale orientalis) and three small Asian mongooses (Herpestes javanicus) that, according to vendors, were suitable for civet coffee production.Nonadults made up the largest proportion of civets (83%).According to traders 61% of the civets were sourced from Bali and 37% from Lombok and 2% from Java, with no other Indonesian islands named as sources.
The mean asking price for a civet was US$ 56.98 ± 36.12.Our lowest estimate for turnover was 33% sold (or died) within a week and the higher estimate was that all were sold within a ten-day period.Using the lower estimate, and assuming the wildlife markets are open 48 weeks of the years, the number of Southeast Asian palm civets sold in the markets over the course of a year is 403.Using the higher estimate, and assuming the markets remain open all days of the year, yields 767 civets sold annually.These number agree well with information provided by the main dealer of civets in Satria market, who informed us that he had sold 52 civets over a 56-day period (which would equate to some 339 civets sold over the course of a year if his shop remained open seven days a week).

Welfare in civet coffee tourism plantations
Civet welfare in civet coffee tourism plantations averaged 2.2 from the husbandry scores (Table 2) and an overall body condition score of 6.05 ± 1.10 (vs 5.50 in the wild in Java).Some scored the highest body condition of 9, "extremely overweight" (Figure 2).Hygiene scores were higher in plantations than in wildlife markets.We regularly recorded cages with metal bar flooring.A third of the civet coffee tourism plantations provided sleeping areas in the form of a box, a log or a shelf; most cages had a roof that offered some protection from the elements (Figure 2).Due to the prevalent poor condition of the surfaces found in both plantations and wildlife markets, we could not statistically calculate the influence of this husbandry factor on the body condition score.
Results of the five freedoms assessment showed that 8.57% of individual civets were not provided with any of the freedoms, 14.28% received one freedom and 77.14% received two freedoms.No civet in civet coffee tourism plantations met all the freedoms.In civet coffee tourism plantations, diet, hygiene and mobility significantly affected body condition score (Table 3; Figure 3e-g).Better hygiene scores resulted in better body condition score.Mobility score significantly and negatively correlated with body condition score.Reduced mobility caused increased body condition scores.Better food scores caused body condition scores closer to the ideal bodyweight of five.

Welfare in wildlife markets
Overall mean husbandry welfare score in the wildlife markets was 0.55 and >90% of animals in the markets were not provided with any of the five freedoms.Mean civet body condition score in the markets was 4.27 ± 0.77.Animals were often situated close to busy roads, crowds, noise and other species.Five civets had access to shelter; others were left in direct sunlight with no access to shade.No designated sleeping areas were provided and only metal barred surfaces were available.Only 15% of individuals had access to water in their cage.Mobility was scored 0 in all of individual cases in the markets thus it was not included in the model.Water had a significant effect on body condition score (Table 3), and a higher score for water correlated with a better body condition score (Figure 3a-d, Table 3).Although not statistically significant, shelter, hygiene, and noise showed a trend that affected body condition.

Civets in civet coffee production and tourism
Despite our hypothesis that we would see no civets for sale, we recorded almost one hundred civets of four species, mainly Southeast Asian palm civet, in 29 civet coffee tourism plantations in Bali, and we found an additional 136 Southeast Asian palm civets in four wildlife markets.Turnover in wildlife markets was relatively high, with estimates ranging between 400 and 800 civets being sold annually.D'Cruze, Toole, Mansell, and Schmidt-Burbach (2014) suggested that, globally, thousands of civets are taken from the wild to supply the civet coffee industry, including those used in civet coffee tourism plantations.Focussing on Indonesia, Nijman et al. (2014) stated that civets are likely sourced from the many wildlife markets that are present throughout western and central Indonesia.
On Bali, the largest wildlife markets are situated in the south, in and near the capital Denpasar and near the main tourist destinations.The majority of the civet coffee tourism plantations are situated within a 60 km radius from these wildlife markets.Shepherd (2012) surveyed Javan wildlife markets between 2010 and 2012, finding the Southeast Asian palm civet the most numerous small carnivore offered for sale, and he argued that many of them were destined for civet coffee production.Eight of the eleven civet species found in Indonesia are not protected by national legislation, and harvest and export quotas are set annually to regulate their trade (Shepherd, 2008).We agree with previous work that the civet coffee industry poses a threat to both civet welfare and conservation of wild populations (D'Cruze, Toole, Mansell, & Schmidt-Burbach, 2014).Previous work on civets in these civet coffee tourism plantations focussed on a smaller number of facilities (16 vs 29) and indeed a smaller  number of civets (48 vs 99) (Carder et al., 2018) but the overall finding in terms of low welfare scores was similar to our observations.The observations of civets in the wildlife markets, with many of them being very young, are consistent with surveys conducted in the 1990s to 2010s by Shepherd (2008Shepherd ( , 2012) ) and Nijman et al. (2014), demonstrating a persistent trade in civets in these markets.With a domestic quota of 25 wild-caught Southeast Asian palm civets, allowed to be harvested annually and to be sold as pets only, our calculations of annual turnover show that these quotas are not abided by.We are not aware of any documentation or guidelines where civet coffee tourism plantations or indeed commercial civet coffee farms can legally source the civets they use, and it appears that all are sourced illegally.

Welfare assessments of caged civets
We found that welfare scores were low and that the civet coffee tourism plantations had poor adherence to the five freedoms, contra to our hypothesis that these facilities would adhere to Indonesia's regulations on civet husbandry.Welfare scores in the wildlife markets were even lower than in the civet coffee tourism plantations, with many not having access to water, no space to move, no shelter and no food.The housing conditions found in our market surveys are similar to those reported by Nijman et al. (2014).These data also support those of Hooper (2022), who compiled TripAdvisor reviews from 25 civet coffee tourism plantations, in which reviewers described the conditions under which civets were housed as generally poor or terrible.The main challenges in the civet coffee tourism plantations were poor diet, limited options for movement or social behavior, and limited choice to hide from the public or the elements.
We found that the cages in which civets were kept were generally small and had little or no environmental enrichment.For a single civet, Indonesian government regulations dictate a cage size of 18 m 3 with a double entrance door (Sulaiman, 2015) but this was never recorded.Animals need space to engage in natural behaviors and enclosure size and configuration is paramount in this.Common and Southeast Asian palm civets typically have home range sizes of in the range of 1.0-3.0km 2 in males and 0.5-2.0km 2 in females (Grassman, 1998;Joshi, David Smith, & Cuthbert, 1995;Nakashima, Nakabayashi, & Sukor, 2013;Rabinowitz, 1991).
We predicted that lower husbandry scores would result in a higher deviation from optimal body condition scores.This was the case, in that better hygiene scores in the civet coffee tourism plantations led to better body condition scores.In general, we found low to very low levels of husbandry, especially in the wildlife markets.Body scores of civets in the wildlife markets were close that what was recorded from photographs of wild civets, which support the conclusion that most, if not all, are sourced from the wild and remain in the wildlife markets for only short periods of time (Nijman et al., 2014;Shepherd, 2008).Food did significantly affect body condition scores in the civet coffee tourism plantations.High body condition scores may be influenced by the lack of mobility granted to civets kept in small cages, reducing their ability to exercise.Obesity was regularly seen in civet coffee tourism plantations where civets were mostly fed a diet of coffee beans, papaya and banana, and occasionally, rice and milk.A species' nutritional requirements need to be understood to provide an appropriate diet (Schuppli & Fraser, 2000).Civets are omnivorous, consuming invertebrates, a variety of fruits and vegetables and small vertebrates (Colon & Sugau, 2012;Fung, Tan, & Sivasothi, 2018;Subrata & Syahbudin, 2016) but this variation was not reflected in the food presented to civets in the plantations or indeed in the wildlife markets.Hooper (2022), on the basis of statements made by tourists having visited civet coffee tourism plantations, suggested that civets were drugged to allow for safe public handling; while we recorded lethargic and sleeping civets, and obese civets barely able to move, we found no evidence of them being drugged.

Genuine civet coffee and the management of civets for civet coffee
At first sight, the amount of packaged coffee we recorded to be present in the civet coffee tourism plantation shops that was labeled "civet coffee" was relatively low, i.e., on average 13.1 kg in each shop.However, the amount of civet coffee that an individual civet can produce at any given time is limited.Based on data from civet coffee farms in Vietnam, Thi et al. (2022) reported that Southeast Asian palm civets on average are able to produce 9.82 kg of civet coffee/year.During our survey we recorded almost 450 kg of civet coffee on display in the shops, or about half the annual production of the hundred or so civets that were present at the time of our survey.We have no data on the turnover of civet coffee in these shops (i.e., how frequently is civet coffee purchased and how long does prepackaged civet coffee remain on the shelves), but we doubt that this is in the order of six months (as this is needed for the civet coffee production and the number of civets to match up).Hence, we expect that a significant proportion of the civet coffee for sale in the shops is either not pure (i.e., it is mixed with non-civet coffee, or it does not contain civet-coffee at all) or, alternatively, a significant amount of civet coffee is produced offsite.The issue of "fake" civet coffee has been brought up frequently in research into what distinguishes civet coffee from ordinary coffee (Burns & Walker, 2020, Jumhawan et al., 2016;Lachenmeier & Schwarz, 2021).
In terms of management of civets used in the production of civet coffee, it is important to distinguish between the civet coffee tourism plantations where a relatively small number of civets are kept on site and where the main revenue comes from the tourists visiting the facility and commercial civet coffee farms where large numbers of wild-caught civets are kept in cages to produce civet coffee.In the former, and the one we report on here, improvements can be made by increasing awareness of the poor welfare standards to tourists and to reduce the demand for civet coffee tourism.Webster (2001) states that it is the responsibility of the consumer to "convert an expressed desire for higher welfare standards into an effective demand."While only a small number of tourists on TripAdvisor raised any welfare/conservation concerns after engaging in wildlife tourism (Moorhouse et al., 2015), in the case of civets and civet coffee tourism plantations this expressed more explicitly (Hooper, 2022).This shows a need for tourist education and the promotion of ethical tourism.Previous campaigns have successfully changed tourist perceptions, for example in the UK there are now no cetaceans held in aquariums with tourists preferring to view this species in a wild setting (Hughes, 2001).Fennell (2013) noted that seeing fewer injured and sick animals will make for happier tourists who are therefore likely to spend more money, and hence animal welfare should therefore be a priority for tourism providers.With relatively small numbers of civets present at each of the civet coffee tourism plantations and the daily presence of tourists, it should be (practically and economically) feasible to improve the welfare standards of civets at these facilities.Small changes to the cage design, and providing shelter, appropriate food and water, are cheap and effective.The management of civets in Indonesian civet coffee farms, and how welfare standards are adhered to, are yet to be properly assessed, but the limited information available suggests small cages and low levels of welfare (Muzaifa et al., 2016;V. Nijman and K.A.I. Nekaris, unpubl. data).For Vietnam, some initial data was presented by Thi et al. (2022).Given that most of the civet coffee production in the civet coffee farms is hidden from view, better adherence to existing rules and regulations and increasing the welfare standards for the civets in these farms, may be more difficult to achieve.Improvement of the welfare of civets kept for short periods in wildlife markets is best achieved by enforcing existing legislation and not permitting this trade to continue.
The impacts of the trade of civets for the civet coffee industry on the population status of Southeast Asian palm civets is not fully understood, due to a lack of accurate baseline data on both the population status in the wild and a lack of data on the number of civets that are taken from the wild (Shepherd, 2008;Duckworth et al., 2016;c.f.;Hughes et al., 2023).The open sale of civets in the wildlife markets in Indonesia, as reported here and elsewhere (Morcatty et al., 2022;Nijman et al., 2014;Shepherd, 2008Shepherd, , 2012) ) necessitates that need for increased monitoring and enforcement of laws within the trade chain.

Conclusion
Civet welfare in civet coffee tourism in Bali, and the trade in civets in Bali, Lombok and East Java does not abide by Indonesian laws and regulations.Better enforcement of existing regulations and laws, both with respect to the keeping of civets in captive settings and the trade in civets, is urgently needed.Continued research into civet behavior and ecology will help to further understand the needs of these species in captive environments (c.f.Clubb & Mason, 2007).Our study highlights the importance of cage size, access to water, diet, hygiene, and noise levels to civet's body condition and welfare.Reducing tourist demand for civet coffee, and the need to interact with the process of civet coffee production, through raising awareness of the welfare and the conservation implications this industry has is urgently needed.Furthermore, education for tourism providers may also improve welfare of civets kept near their premises.

Figure 1 .
Figure 1.Civet body condition score chart used to estimate body scores in civet coffee tourism plantations, wildlife markets and the wild.

Figure 2 .
Figure 2. Civets in wildlife markets (left) and in civet coffee tourism plantations (right).

Figure 3 .
Figure 3. Relationship between husbandry factors on civet body condition score in wildlife markets (a-d) and on civet body condition score in civet coffee tourism plantations (e-g) in Indonesia.

Table 2 .
Mean husbandry scores of civets in wildlife markets and civet coffee tourism plantations and body condition scores of civets in wildlife markets, civet coffee tourism plantations and the wild.

Table 3 .
Details on the best-fitted models of husbandry factor effects on body condition score in civets in wildlife markets and civet coffee tourism plantations in Indonesia, generalized R 2 , and ΔAIC in relation to the second ranked model.For both models the logistic family of distribution was used.ΔAIC in relation to the second ranked model: wildlife markets = 2.71; civet coffee tourism plantations = 4.38.Classes 0 of each predictor variable was used as the references for comparison.