Introduction

The livestock industry in the People’s Republic of China (hereafter referred to as China) has rapidly developed over the past two decades [1]. In 2014, China produced approximately five billion chickens, 480 million pigs, 202 million sheep and 117 million cattle, making it the world’s largest livestock producer [1]. It was the world’s biggest meat producer (approximately 83 billion tonnes in 2013) [2], with major exports of beef [3], pork [4], chicken and other meats [4]. In addition, its population is one of the world’s largest meat consumers [5]. Therefore, the livestock industry is of considerable economic importance in China, with an estimated output value of RMB 25.8 trillion (approximately 4 trillion USD) in 2011, which was 31.7% of the total value of Chinese agriculture (including crops, forestry, livestock, and fishery) [6].

The scale of livestock production and its associated intensification has resulted in growing animal welfare concerns, both in terms of numbers of animals and the welfare conditions [7, 8]. Chinese traditional ethics from Confucianism, Taoism, and Buddhism support compassion for non-human animals [8]. However, with the current industrialisation of China’s livestock production, many Western farming practices (such as gestation crates, castration, tail docking and ear-clipping of pigs, beak-trimming and battery cages for hens, castration and early weaning of calves, and forced feeding of ducks and geese for foie-gras production) are being adopted and utilised with increasing frequency. At the same time, some Western countries (particularly in the European Union) are phasing out some of these practices, such as battery cages and stalls for sows during gestation [9]. Due to the large numbers of livestock produced in China, this is probably one of the major animal welfare concerns in the world today [8]. Within China the concept of animal welfare is little understood, and in a recent survey of public opinions toward farm animal welfare in China, approximately two thirds of respondents did not know what the term meant [10]. In Chinese, the usual translation is ‘dongwu fuli’. Fuli literally means interests in life, especially care taken for employees by providing food, shelter and medical care, thus creating a concept of a high level of provision for needs which has then been applied to animals (dongwu)[11]. The high level of provision described in this concept may lead some to question the validity of concern for animal welfare, when humans are not afforded all of these provisions. Confusion also exists between animal welfare and animal rights causes [12]. Motivation in China to improve animal welfare has been lacking due to concern that farm profitability would be adversely affected as a consequence [8, 11]. Nevertheless, there is a growing recognition that trade with overseas partners will require international animal welfare standards to be met [8, 11, 12, 13, 14]. China’s trade partners in both the import and export markets have expressed concern regarding Chinese animal welfare standards [15, 16, 17]. However, there are an increasing number of Chinese animal protectionists, and they are one of the most active interest groups in China [8]. The proposal, in 2009, to introduce a Prevention of Cruelty to Animals Law in China was an important development in animal welfare in the country, but the progress on advancing animal welfare has been slow and there is still no national farm animal protection legislation in effect [11, 12, 13, 14].

In terms of the regulations or legislations related to animal welfare in China, a pig slaughtering management regulation was first introduced in 1998 and in December 2007 it was revised [18], but neither of these required stunning before slaughter. In 2008, an act outlining requirements for humane slaughter was drafted [12]. Since 2014, there have been several proposals to develop more comprehensive livestock regulations for animal welfare, posed by for example the China Association for Standardization [19], the Chinese Veterinary Medical Association, and in 2017 the Chinese government [12].

The welfare of animals in livestock industries is a responsibility shared between farmers, the livestock companies and industry service providers, such as transporters, slaughter facilities, and veterinarians [20]. Livestock industries must comply with animal welfare legislation but are also indirectly responsible to the public and consumers for the welfare of the farm animals [20]. Consumers in many countries have shown increasing interest in animal welfare issues in recent years, and improvements in animal welfare standards have often been driven by concern from consumers and the public [21]. As well as putting pressure on industry to improve farm animal welfare, consumers are increasingly willing to choose products based at least partly on the perceived welfare of the animals [10, 20, 21]. For example, consumers in Shanghai and Beijing are now showing a preference for pork with better animal welfare claimed in the labelling, not just because they thought it was tastier and safer but for the benefit of the animals [22]. Successful management of issues with a human dimension requires involvement of stakeholders in the definition of the problem to be addressed, decision-making, and development of relevant and targeted strategies informed by research [23, 24, 25]. The importance of understanding and engaging local stakeholders in strategies to address issues has been demonstrated in other comparable areas of human-animal interaction, such as human-wildlife conflict, wildlife and environmental conservation, and management of introduced animals [23, 26]. Marketing theory stresses the importance of understanding the target audience in order to make marketing campaigns relevant and engaging [27]. Engaging stakeholders in the Chinese livestock system to improve animal welfare first requires an understanding of what the perceived welfare issues are, but they are little understood for Chinese livestock transport and slaughter.

Livestock transport and slaughter are two areas in which animal welfare may be of considerable concern. Animals in China are transported over long distances (for example, chickens travelling 2500 km from Jilin, near North Korea, to Shenzhen, just north of Hong Kong [28]), which could potentially have serious adverse effects on their welfare [28]. In addition, humane slaughter is a relatively new concept in China, with currently no requirements for this in many slaughter facilities [8, 10].

Choice tests can be used to gauge stakeholders’ attitudes to animal welfare, in which they select which of two options provides the better welfare for animals [20, 29, 30]. Such conjoint analysis can be adaptive (Adaptive Conjoint Analysis, ACA), in which case a computer customizes the questions according to the respondents’ answers, giving advantages over traditional survey techniques [20]. This technique was designed for market research to determine consumers’ product preferences [31, 32] but has also been adapted to determine opinions related to the importance of different welfare issues in Australian livestock production [20,33]. The strength of this technique is its ability to ask respondents realistic questions which mimic the trade-offs that people need to make in the real world. For example, it is not always possible to ensure that every aspect of animal welfare is ideal in a given scenario. Therefore, it is necessary to decide which aspect of an animal’s welfare is most important so that it can be prioritized. Web-based interactive ACA questionnaires can be constructed using intuitive software (e.g. Sawtooth Software Inc., Sequim, WA, USA), which allows emphasis on issues that require additional questions because previous responses are too similar to indicate the respondent’s preferences. This adaptive capacity facilitates a rapid and accurate definition of the preferences or opinions of the questionnaire respondent [33]. The computer generates utility values, which are measures of the respondent’s relative strength of preference for each level of welfare provision on each issue. For example, for the welfare issue of stunning before slaughter, the levels might be electrical stunning, penetrating captive bolt, percussive captive bolt, blunt trauma and no stunning. In our opinion, a web-based interactive ACA questionnaire was a better technique for the purposes of this study than a conventional questionnaire.

The objectives of this study were to establish what the perceived animal welfare issues in livestock slaughter and transport are in China, using ACA and utility scores, analysed through multivariable analyses. When coupled with in situ measurements of welfare, this information is anticipated to improve understanding of animal welfare issues in transport and slaughter in the Chinese livestock industry and inform the direction of future animal welfare improvement.

Material and methods

Establishing the welfare issues

An initial one-day focus group meeting in December 2015 addressed the question ‘what are the most readily identified farm animal welfare concerns for commercial animal welfare professionals in China?’ Participants excluded fish as a farm animal of interest. The participants consisted of eight senior livestock transport and slaughter company representatives and livestock production academics operating in and around Guangzhou, China. They were invited by a senior academic representative of South China Agricultural University based on their pre-eminence in the livestock sector. The participants also had self-identified interests in improving animal welfare for commercial reasons, including gaining access to export markets. The area of Guangzhou was chosen because it is one of the most important regions in China in terms of livestock production, including that supported by foreign investment [34]. Attendance was voluntary and unrewarded.

In the focus group discussion, which took the form of a wide-net participant-directed brainstorm, respondents systematically commented on potential welfare challenges within their industries. After discussing the issues generally in the first half of the day, the group was asked to collectively decide on the 15 most pressing animal welfare concerns during livestock transport and slaughter in China. The major issues were listed, and discussed again to ensure that the list was an accurate representation of their collective view.

The issues identified for livestock during transport were:

1. 1) Experience of livestock transport workers
2. 2) Attitude of livestock transport workers
3. 3) Method of catching chickens
4. 4) Part of body used to catch fully developed chickens
5. 5) Livestock loading facilities
6. 6) Type of road vehicle used to transport livestock
7. 7) Overcrowding of livestock in transport vehicles
8. 8) Length of journey to slaughterhouse
9. 9) Maintenance of animal comfort during transportation (temperature)
10. 10) Maintenance of animal comfort during transportation (provision of rest and water)
11. 11) Stress to livestock during journey

Issues identified for livestock during slaughter were:

1. 1) Pre-slaughter accommodation
2. 2) Stunning procedures for poultry and pigs
3. 3) Stunning procedures for cattle and sheep
4. 4) Achievement and maintenance of unconsciousness during the slaughter process

Water injecting of pigs was raised as a potential issue although it was not included in the distributed list as it is currently illegal in China. The practice involves forcing water into the stomach of pigs shortly before slaughter, or injecting it into their hearts, to increase their weight. Each issue was further described by two to five levels (or different livestock management scenarios) with different welfare implications (Table 1). These were developed and refined during the meeting.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Welfare issues and their utility components, indicating the relative strength of preference for each level of welfare provision on each issue.

Each issue had 2–5 utility levels demonstrating the perceived importance for animal welfare. The issues are listed in declining order of importance value (the difference that each issue made to the total of all the utility values for each respondent). The average importance and zero-centred utility values are listed for each issue, as rated by respondents (n = 267), who were stakeholders in the animal transportation and slaughter sectors, to a questionnaire on the major welfare issues in the Chinese livestock transport and slaughter industry.

https://doi.org/10.1371/journal.pone.0197028.t001

After the focus group meeting, two new issues were added by the authors of this paper

to ensure that the issues were balanced and consistent: ‘attitude of livestock transport workers’ and ‘level of experience of livestock transport workers’. In addition, the ‘maintenance of animal comfort during transportation through provision of rest and water’ was added to ensure clarity and consistency with a question about stress, and ‘stunning procedure for cattle and sheep’ was added to match the issue that was raised for pigs in the focus group meeting.

The questionnaire (S1 and S2 Tables)

The outcome of this focus group was used as the basis for a survey using intuitive software, which was distributed electronically to a wider group of stakeholders in China, to include a more diverse work force within the animal transportation and slaughter sectors and increase the number of respondents. The questionnaire was developed using the list of issues, relevant literature, and consultation with industry and academic stakeholders. Using a web-based interactive ACA questionnaire format, it was piloted with volunteers and changes were made based on their feedback. After finalising the questionnaire in English, it was translated into Mandarin by a professional translator. The meaning and translation were checked for accuracy and consistency of meaning by a third party fluent in both Mandarin and English. In addition, our academic partners in China were sent both the English and Mandarin versions of the questionnaire to compare in order to double check the accuracy of the questionnaire translation.

The finalised questionnaire was then loaded into the online software (Sawtooth Software version 8.4.8., Sawtooth Software Inc., Sequim, WA, USA). This has the advantages of being able to intelligently adapt the questionnaire in real-time to each respondent based on their answers and engaging the respondent by using a computer-interactive method [33].

Distribution of the questionnaire to potential respondents.

The questionnaire was distributed to people involved in livestock transport and slaughter in the China during February and March 2016. We used a virtual snowballing technique, starting with our academic and industry partners and stakeholders in the Chinese livestock industry. A recruitment e-mail with information about the study was sent to the focus group attendees; they were then asked to forward the recruitment email to all Chinese livestock industry stakeholders in their social and professional networks with request for them in turn to forward the recruitment e-mail to other relevant people. A link in the e-mail connected them directly to the survey site. Those who volunteered to participate clicked on the link and were directed to the first page of the survey, where participants were shown information about the questionnaire. This stated that: participation was voluntary, participants needed to be 18 years of age or older, the study had been cleared in accordance with the ethical review guidelines and processes of The University of Queensland, and that participants could discuss their participation with project staff or the School Ethics officer (contact details were given). Respondents confirmed on-line that they were over 18 years of age, that they understood that their participation in the study was voluntary, consented to participate, and to their responses being used as outlined. Only participants who gave consent answered questions from the survey. Demographic questions were used to gain information about type and duration of the Chinese livestock industry, gender, age, education level, geographical location of work relating to the Chinese livestock industry, and knowledge of livestock transport and slaughter in China. Finally, respondents were provided with a definition of animal welfare (based on the World Organisation for Animal Health definition in the Terrestrial Code [35]): “Animal welfare means how an animal is coping with the conditions in which it lives. An animal is in a good state of welfare if it is healthy, comfortable, well-nourished, safe, able to express innate behaviour, and if it is not suffering from unpleasant states such as pain, fear, and distress.”

The sections of the questionnaire, including adaptive conjoint analysis.

Part 1: Rating of levels. First, issue levels were screened by asking the respondent to rate the importance of each, which allowed the software to learn enough about the respondent’s preferences to construct initial part-worth estimates. Respondents were given an introduction before they were presented with the issues: “In this section you will be asked about different livestock management scenarios and, in your opinion, how acceptable they are in relation to animal welfare. Please consider only the immediate impact of each scenario on the animals’ welfare, regardless of the associated practical and economic factors.” The responses were recorded on a 5-point Likert scale from “not acceptable” to “very acceptable”.

Part 2: Importance questions. Respondents were asked to rate how important they thought the difference was between their perceived worst and best levels of animal welfare for each of 15 issues (Table 1). They were given an introduction before they were presented with the perceived best and worst levels for each issue: “In this next section, you will be presented with two livestock management scenarios. Please consider whether the scenarios are different in terms of their immediate impact on animal welfare. If you think that animal welfare does not differ between the two scenarios, please choose ‘not important’. If you think there is a difference between the two scenarios, then use the scale to show how important you think the difference is (please assume that all other aspects of the livestock management are acceptable)”. The 5-point Likert scale was from “not important” to “extremely important”.

For example:

Respondents were asked to rate the importance, in terms of animal welfare, of the difference between an animal remaining unconscious throughout the slaughter process and an animal regaining consciousness during the slaughter process after stunning.

Part 3: Customized paired comparison trade-off questions. Respondents were presented with ten sets of two combinations of livestock management scenarios, based on their previous answers. Since it was not practical to have respondents evaluate all possible animal welfare combinations, the ACA approach allowed the efficient collection of detailed information on animal welfare issue importance by presenting respondents with only a carefully chosen set of hypothetical welfare scenarios for evaluation. Questions provided the most additional information by selecting concepts that were known to be nearly equal in importance to the respondent, given what was already known about the respondent's values from their previous responses. The respondent was then asked which of the pairs of hypothetical welfare scenarios was preferable to them assuming that all other aspects of the livestock management were acceptable and considering just the immediate impact of each procedure on the animal welfare. Following each response, the estimates of part-worthy were updated to allow the software to determine what pair of concepts to present to the respondent next. Certain pairs of issue levels that rationally could not appear together in the same concept were specified as part of the questionnaire logic (for example, welfare scenarios that are specific to poultry with a welfare scenario that is specific to cattle). In contrast to direct questioning methods simply asking how important each issue is, conjoint analysis forces respondents to make difficult trade-offs, as in a real world decision-making process [36].

One combination was presented on the LEFT side of the screen and the other on the RIGHT side (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Example of combination scenario.

https://doi.org/10.1371/journal.pone.0197028.g001

Part 4: ACA calibration questions. Six calibrating scenarios were composed by the software, using those issues determined to be most important to the respondents, who were asked to rate how acceptable the scenario was to them from an animal welfare point of view. They selected a number between 0 (definitely unacceptable from an animal welfare point of view) and 100 (very acceptable from an animal welfare point of view). Scenarios were chosen to occupy the entire range from very unacceptable to very acceptable for the respondent. The concept the respondent was expected to find least acceptable among all possible concepts was presented first, and the one expected to be most acceptable was presented second. Those two concepts established a frame of reference and the remaining concepts were selected to have intermediate levels of acceptability. Each animal welfare scenario had two to five components, for example:

Livestock transport workers with a poor attitude.

Fully developed chickens are manually caught by the legs and feet.

Poultry are caught manually in the light.

Finally, respondents were given the opportunity to comment in a free text field whether there were any other welfare issues associated with livestock transport and slaughter in China that they believed were important but were not covered in the questionnaire.

Results

The questionnaire was started by 943 respondents. Of these, 332 dropped out after the question asking what industry involvement they had (presumably because they did not fit into any category), 180 dropped out after reading the instructions, 28 dropped out during the paired comparison questions, three after other issues, one after the last system question. Although 399 respondents finished the survey online, not all of them had utility values, and only 267 respondents had answered the demographic, ACA rating and importance questions, and customized paired questions contributing to the utility value.

The most common role of stakeholders was in livestock transport (Table 2), which represented 33%, and the second most common was in livestock production research or a teaching role, which both represented 27% of respondents. Only 2.2% of respondents were not involved in any way, and few were involved in an environmental protection role related to livestock, and none in a veterinary role or in livestock slaughter. Government roles relating to livestock production, livestock farmers, and livestock welfare research or advocacy were all represented (12.7%, 11.2% and 7.9% respectively).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic characteristics of respondents, who were stakeholders in the animal transportation and slaughter sectors, to a questionnaire on the major welfare issues in the Chinese livestock transport and slaughter industry, as defined in an initial expert meeting (n = 267).

https://doi.org/10.1371/journal.pone.0197028.t002

Demographic effects

As respondents’ self-reported knowledge of transport increased, their rating of the acceptability of many issues increased (Table 3). Similarly, as respondents’ self-reported length of involvement in the industry increased, they increased their rating of the acceptability of several issues (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Effects of the knowledge of respondents (n = 267), who were stakeholders in the animal transportation and slaughter sectors, about transport on their ascribed utility values, which measure the perceived importance of different issues for animal welfare, when the relationship was examined by linear regression in a general linear model.

The relationship is reported as a regression coefficient, SE coefficient and probability (P) value of the regression being significant.

https://doi.org/10.1371/journal.pone.0197028.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Significant (P< 0.05) effects of the duration of involvement of respondents, who were stakeholders in the animal transportation and slaughter sectors, with the industry on the mean utility values of issues, which measure the perceived importance of different issues for animal welfare, when the relationship is examined by linear regression in a general linear model.

The relationship is reported as a regression coefficient, SE coefficient and probability (P) value of the regression being significant.

https://doi.org/10.1371/journal.pone.0197028.t004

Stakeholder involvement.

Respondents in the livestock production teaching and research role rated stunning of cattle and sheep electrically and livestock remaining unconscious throughout the slaughter process as more acceptable, and not stunning pigs and poultry less acceptable, compared to livestock transporters (Table 5). They also, in conjunction with those in a government role and livestock farmers, rated catching poultry manually in the dark as more acceptable than livestock transporters.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Significant (P< 0.05) effects of type of industry involvement of stakeholders, who were stakeholders in the animal transportation and slaughter sectors, on the mean utility values of different levels, which measure the strength of preference for animal welfare.

High values indicate high acceptability on animal welfare grounds. There were no veterinary or livestock slaughterer respondents, and a small number of respondents that did not align themselves with any of the stated industries were excluded.

https://doi.org/10.1371/journal.pone.0197028.t005

Other demographic factors

Better educated stakeholders (with a university degree) perceived stunning poultry and pigs by carbon dioxide to be more important for animal welfare than those educated only to high school or a technical college level (Table 6). Similarly, those with a postgraduate degree perceived electrical stunning of poultry and pigs to be more important for animal welfare than those educated only to a school or a technical college level. Women perceived transporting livestock by an open-sided vehicle to be more important for animal welfare than men (mean utility values 0.46 and 0.096, respectively, SED = 0.157, P = 0.04).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Significant (P< 0.05) effects of level of education of stakeholders, who were stakeholders in the animal transportation and slaughter sectors, on the mean utility values of issues, which measure the perceived importance of different issues for animal welfare.

High values indicate high acceptability.

https://doi.org/10.1371/journal.pone.0197028.t006

Discussion

In this study of Chinese stakeholders in the livestock transport and slaughter industries, stunning for cattle and sheep was rated as the most important animal welfare concern, amongst the different options given, followed by stunning for pigs and poultry. In contrast, the animal welfare concern considered least important was the method of catching chickens. This suggests that Chinese stakeholders were more concerned about slaughter issues than transport issues. This is in accordance with a recent study of the attitudes of Asian students of veterinary medicine and animal science towards slaughter and transport of livestock, in which the students in China were more concerned about slaughter issues than transport issues [40]. A greater concern for the slaughter process, rather than transport, is also in accordance with views in Australia towards slaughter standards in Indonesia, where, in particular, the absence of stunning has caused major concern amongst the public [41]. However, a survey in Australia showed that road and ship transport were of greater concern than pre-slaughter stunning to stakeholders in the Australian livestock industries [20]; this is probably because the pre-slaughter stunning is the norm in Australian abattoirs. Concern about slaughter practices in China has been heightened recently by the link between slaughter of poultry in markets and the emergence of bird flu (H7N9) [42]. This occurred initially in 2013 and by January 2017 had caused the deaths of 79 people, with 192 people infected [42]. In some provinces markets have been required to end the slaughter of poultry [43].

It was expected that government officials would demonstrate greater concern for welfare issues associated with international trade as some livestock and poultry products have been refused by the European Union and the USA due to welfare concerns [11]. However, this was not observed in our results, perhaps because the officials were not dealing directly with international trade. Most respondents were either involved in livestock and poultry production or in livestock teaching and researching roles, and most (75%) had more than one year’s experience in their work. As respondents’ level of education increased they rated stunning before slaughter to be more important, and self-reported professional knowledge and experience were also positively associated to the importance attributed to stunning. In China, the relationship between meat quality and welfare prior to slaughter has been widely reported in academic literature [44, 45, 46, 47]. Despite this, the maintenance of good conditions during transport and slaughter is often neglected [48]. In the absence of any transport legislation or animal welfare legislation generally, and a strong market imperative to grow the livestock industries, there is little incentive to improve welfare at this point in the animal’s life [28].

Chinese consumers are showing more concern about livestock and poultry products’ quality and safety than previously [11]. Issues relating to unsafe food have occurred frequently in the country, and this has had an impact on public health. It has also negatively impacted on the national economy as products are imported at high cost (e.g. milk powder) from overseas as they are considered to be safer [49]. As food safety is currently of significant concern to consumers in China, one way to encourage good conditions for livestock during transport and slaughter is to emphasize the links between improved conditions and food safety [11]. In addition, according to a recent survey of Chinese public preferences for pork, nearly 77% of consumers prefer to purchase meat produced under good welfare conditions even if it costs more [11].

The survey also highlighted some of the aspects of transport and slaughter that are related to the geographical conditions in the country. For example, respondents in this study rated open-sided vehicles more acceptable in terms of animal welfare than semi-closed and closed-sided vehicles. Livestock and poultry are generally transported at a high density in China, and in the summer transport may occur in high temperatures, which will cause heat stress and sometimes even death [48]. Good ventilation during transport is considered one of the most important factors, which may improve the meat quality of the carcass [48]. Most (77%) of the respondents were from South China where the ambient temperatures are often high. Ventilation is generally better in open-sided than closed-sided vehicles [48], and good ventilation will reduce the temperature to which animals are exposed during transport [50]. Conversely in the northern Europe where there are often cooler conditions and lower stocking densities, closed-sided vehicles are usually preferred due to the ability to control ventilation, and, consequently avoid the chilling of livestock, e.g. [51] for chickens. Journey length was not considered very important but this may have been because the options were <3, 3–6 and > 6 hours, which reflects the fact that in China pigs and chickens may be raised relatively close to their point of slaughter. There are however many exceptions where livestock are transported long distances across this large country, mainly from the inland regions to provide meat for consumers in the east coast cities [28].

Limitations of the study

The benefits of virtual snowballing include the ability to collect data from a large number of people rapidly and inexpensively, and false reporting may be reduced by the anonymity of this sampling technique [63, 64]. In contrast, there is significant potential for selection bias with this sampling technique and it is not possible to estimate response rates. It has, however, been reported that internet surveys using this method give similar results to traditional sampling methods and can make valuable contributions to research [64, 65]. We acknowledge that the respondents probably represented a well-educated and geographically limited sector of the stakeholders in this industry. It was also clear that many people started the survey but did not complete it, which may have been due to their eligibility as stakeholders, the time taken to complete it, or their lack of interest in the subject matter. This is not uncommon in adaptive conjoint analysis surveys, which require respondents to think carefully, but it suggests a bias, that those that completed the survey were more likely to be those more interested in, and able to provide, detailed evaluation of livestock welfare. There was evidence of this since well-educated respondents issued a higher utility value to stunning procedures, which may be due to their being more aware of the welfare benefits these provide. An over-representation of well-educated respondents is valuable for the survey since reliable answers are likely to be obtained, but it does not necessarily represent the majority view. Less well-educated respondents may be more likely to rely on intuition, which may favour more emotive issues such as avoiding causing pain to livestock.

The survey required stakeholders to have access to the internet, which only 52% of the Chinese population has [66]. Since it is mainly the older generation and the rural poor that lack access, it was expected that most stakeholders in this industry would have access, but it must be acknowledged that some of those most regularly in contact with these rural industries did not have access to, or regularly use, the internet.

Due to the unbalanced geographical distribution of respondents, with 76% of the respondents from the South of China, the distribution is unlikely to represent the overall perspectives of all Chinese people. People from different regions in China may have different attitudes because of different levels of economic and social development [9]. In addition, the absence of livestock slaughterers and veterinarians in the sample may have influenced the results, as people in these roles have very specific and important functions and may have different perceptions of the importance of welfare issues in China. Similarly, less well-educated farmers and those that are actively engaged in, rather than just organising, livestock transport, were not adequately represented, and they may have had a different, more pragmatic view on welfare problems during livestock transport and slaughter. However, the importance of welfare enhancement is arguably restricted to the high value supply chains for livestock products. There is generally much less opportunity for less well-educated farmers to obtain increased prices for value-added products, such as welfare-enhanced livestock products, compared with the larger scale suppliers to the major food chains, particularly if they are exporting their product to the EU [67]. Nevertheless, it is important to obtain views from a widespread set of representatives in the livestock sector. Future studies could use community-focused group meetings to elicit responses directly from stakeholders that are hard to reach. These would need to provide incentives for participation, including veterinarians, as the benefit to them is not as obvious as, for example, teachers/researchers, who need to have a detailed understanding of how the industry functions. Farmers are particularly hard to reach, especially as they are older than city dwellers on average, with 55% over 40, or they are rural migrant workers, working both in the cities and on their farms and expanding at about 4% per year [68]. In the latter case, much of the daily agricultural management of the small plots allocated to rural dwellers is done by older members of families. These would be very hard to reach in participatory research. At the same time social science research could gather public opinion on the welfare concerns which have been identified, and whether they would welcome animal products that come from high welfare systems addressing these issues, even if they cost more [69]. The combination of providing information to the public on welfare issues and allowing them to express their opinions is gathering momentum in solving animal welfare problems [69].

It is also possible that our initial focus group meeting failed to identify some of the biggest problems, and therefore stakeholders were unable to select the ones that they believed to be most important. Although water injecting was raised as a potential issue, but omitted because it has been declared illegal in recent years, many Chinese researchers have recently reported it as a current problem [48, 70, 71, 72, 73]. The extent of the practice is unknown but it has obvious welfare concerns. It was not investigated in this study due to it being an issue of great sensitivity in China, its illegality and concerns about the influence this would have on potential responses and the risk that such an investigation might pose for respondents. Future studies should investigate the prevalence of this practice and stakeholder and public attitudes towards it. It is also possible that some of our descriptions of the levels of each issue were potentially leading, by describing attitudes as good or poor, for example, for which benign or malevolent would have been more appropriate.

Conclusions

This study reports the perceived major welfare concerns for livestock during transport and slaughter in China, as identified by a sample of stakeholders in the industry that rated issues that had been proposed by a focus group. Although the stakeholders may not be entirely representative of the industry, there was reasonable agreement on the most serious welfare issues. Slaughter-related issues represented the major concerns, in common with previous surveys addressing transport and slaughter in Asia. This may reflect an increased importance placed on the critical moments of taking life, as compared to other potentially more chronic concerns. Livestock transporters rated catching poultry manually in the dark as less acceptable and not stunning pigs and poultry as more acceptable than many other stakeholder groups, demonstrating some differences between stakeholder groups in the perception of welfare issues. Future research should address key welfare issues identified here with a broader range of stakeholders and examine the potential impacts of improvement of welfare on farmers in particular.

Supporting information

Acknowledgments

This work was conducted as part of the ‘Animal Welfare Standards Project’ under the auspices of the World Animal Health Organization (OIE).