Theoretical Framework

The theory of planned behaviour (TPB) (Ajzen, 1991) attempts to explain behaviours over which people have the ability to exert self-control. The TPB tries to predict an individual’s intentions to engage in a particular behavior at a specific time and place. The TPB has been used in a wide range of disciplines including health (covering intensions of smoking, drinking, breastfeeding etc.), agriculture such as the adoption of soil erosion control measures (Wauters et al., 2010), and evaluation of perspectives in wildlife conservation (Miller, 2017). Within the context of the current study, the TPB framework is being used to investigate willingness by people to consume meat from wildlife species and visit wildlife centres given the perceived risks of contracting the COVID-19 disease. The study investigates if the behavioral achievement of visiting wildlife centers depends on both motivation (i.e. intention to eat meat from wildlife species and enjoy various tourism adventures) and ability (i.e. control of their action given the risks of contracting COVID-19). It was also envisaged that the same framework could be useful in identifying groups of people that are more vulnerable to the COVID-19 related infodemics.

This study used the case of Zimbabwe, a southern African country to investigate possible effects of COVID-19 on the tourism industry. Zimbabwe has more than 14% of its areas set aside for wildlife conservation (Zimbabwe Park and Wildlife Authority, 2019) and about 8% of its gross domestic product from tourism activities (Chigora et al., 2019). While Zimbabwe has experienced pandemics in the past (mostly related to hygiene and water e.g. Cholera and Typhoid), none of the previous cases caught media attention in similar magnitudes as the COVID-19. Any negative perceptions towards wildlife following the COVID-19 pandemic might negatively affect wildlife based tourism and ultimately the economy. However, without knowing people’s perceptions towards wildlife among different socio-demographic groups (e.g. gender, age, education, and employment status), it is difficult to know whether awareness programs are needed, and if they are needed which groups to target. It was hypothesised that people’s perceptions of wildlife management will be influenced by sources of COVID-19 information, age, employment status, and education.

Sampling

We used a questionnaire administered online during March and April 2020 which took approximately five to eight minutes to complete (Appendix). We purposely targeted various groups of people in Zimbabwe to acquire their views on COVID-19 and impacts on conservation. Our online questionnaire was therefore distributed to online platforms (emails, whatsapp, Facebook, and twitter) in high schools, religious institutions, environmental organisations, the health fraternity, policy makers, those working in various private and public entities as well as those who are self-employed. Using information on the geographic location from the questionnaire, we excluded one response from outside Zimbabwe. Due to the increasing availability of internet services in Zimbabwe, especially mobile networks (Amon et al., 2020), it was possible to have responses from most districts in the country. It was impossible to conduct complementary direct interviews (especially to those people who did not have appropriate gadgets to respond online – such as smartphones and computers). This was due to the lockdown restrictions that were still in place at that time in Zimbabwe (i.e people were not allowed to go outside of their homes except for essential services). The questionnaire was written in English as it is the universal language used in the country to communicate across tribes. Structured questions were used to provide quick and relatively accurate ideas on people's ideas concerning our subject. The “other” option was provided for some questions to enable respondents to explain other views besides those provided. We collected five socio-demographic variables from the respondents. These were 1) gender (male/female), 2) age in years (16–20; 21–25, 26–30, 31–35, 36–40, 41–45, 46–50, and 50 years and above), 3) the town/city/district of the respondent’s residence, 4) the highest level of education attained, and 5) the respondent employment status (employed, entrepreneur and unemployed).

We also wanted to assess the people’s perceptions on the source of the COVID-19 disease, being guided by what we obtained in the press during the time of survey design (i.e wild animals, domestic animals, manufactured in a laboratory, and God’s punishment to humanity). For the respondents who thought that the COVID-19 disease originated in wildlife species, we probed them to suggest a possible action for the concerned species. Responses included total destruction of the species, prevention of further contact with humans, stopping the consumption of its meat, ban wildlife trade. For the two above mentioned questions, we included “not sure” and “others’’ as a response. We also asked if the respondent would be affected in the way in which they eat meat from wildlife species given the reported associations of SARS-CoV-2 with some wild animals. We restricted the respondents to give a yes/no answer.

Since some media suggested close links between COVID-19 disease and animals in wildlife sanctuaries/zoos, we asked respondents if they were still willing to visit such enclosed wildlife centres and wildlife protected areas (options for “yes” and “no” were provided).

Lastly, we asked the respondents to state the dominant source of information they had access to (responses included radio/television, newspapers, Ministry of Health website, Whatsapp, Twitter, Facebook, and other).

Data Presentation and Analysis

Due to paucity of data in some classes of our age, source of COVID-19 information, employment, and highest level of education attained, we re-categorised them. Age was recategorised to young adults (16–25 years), middle aged (26–45 years) and older respondents (46 or more years). The sources of COVID-19 information was re-categorised into formal (including Ministry of Health websites and radio/television), and social media (including Facebook, Whatsapp and Twitter). For the employment variable, we re-classified “entrepreneur” as employed to retain two classes of employed and unemployed. Lastly, the highest level of education was categorised as no tertiary education (all levels from Diploma, Advanced level and below); undergraduate; and postgraduate qualification (those who obtained a Master’s degree or further).

We tested for associations between gender, education, employment status, and age with major sources of COVID-19 information, perceived origin of COVID-19 and effects on respondent consumption of wildlife meat using Chi-squared tests. Spatial associations were also tested for similarities in responses given. Considering the socio-demographic variables, we used binomial regression models to explore the factors that influenced respondents to 1) support the ban in wildlife trade; 2) visit an enclosed wildlife centre after the COVID-19 pandemic and whether or not they would support a ban in wildlife trade; and 3) change the way they eat meat from wildlife species considering their perceived option of dealing with the species from which COVID-19 originated. For the regressions, we used the Akaike Information Criterion (AIC) to select the model that best suited our data. The R statistical package was used to conduct the binomial regression models (R Development Core Team, 2020).

We used spatial techniques to explore the relationships between responses and other spatial phenomena such as distance of a respondent residence to a protected area. Hotspot analysis was carried out to establish geographical similarities in the responses given through the survey. Hotspot analysis of the geographical locations of responses was done using a combination of two indicators of spatial association, the Getis-Ord and the Moran I index. The spatial statistics were used to distinguish between two different kinds of spatial effects, that is, spatial interaction (spatial autocorrelation) and spatial structure (spatial heterogeneity). A positive Getis-Ord score as well as a positive Moran I value are indicative of clustering of near-similar responses given (Mitchell, 2005). The ArcGIS version 10.7 was used to generate statistics related to spatial correlation of data sets in terms of dispersion or clustering.

Ethical Considerations

The study received an ethical clearance from Chinhoyi University of Technology.

Public’s Reactions to the Pandemic

Most of the respondents in this study advocated for a ban in the eating of wildlife meat as a response to the pandemic. The probability of advocating for a ban in wildlife meat consumption was higher for those who perceived wildlife species to be the source of the disease. If such suggestions of stopping the consumption of meat from wildlife are coming from the people themselves, there could potentially be a reduction in the demand of wildlife meat which may be a positive development for wildlife conservation (Roe & Booker, 2019). These perceptions may also inspire support of stricter measures on preventing poaching and authorities may get public support since poaching will be conceived not only as a wildlife management issue but a global human health threat (Ntuli et al., 2020). A significant proportion of respondents also advocated for a ban in wildlife trade. This was expected as has been observed in previous outbreaks of zoonotic diseases. Following the outbreak of SARS, China temporarily imposed a ban on the trade of wildlife (Xie et al., 2020). With the outbreak of COVID-19, the Chinese government temporarily banned the trade of wild animals again (Hou et al., 2020; Xie et al., 2020). The general public in China reportedly supported the government in its efforts to reduce contact between humans and with wildlife species (Hou et al., 2020). This also has positive implications on wildlife management as reduced interests in wildlife trade by the public will reduce exploitation of wildlife for money. However, the high probability of visiting wildlife centres by those who supported a ban in wildlife trade may indicate that advocacy for ban in wildlife trade is not necessarily influenced by COVID-19 but other factors. There is a possibility that those who had always advocated for a ban in wildlife trade for one reason or another, may maximise on the COVID-19 pandemic opportunity to be heard. This confirms previous findings by Mejova and Kalimeri (2020) who noted how the pandemic has been greatly used by different sectors of society to push individual agendas such as political attacks, donation solicitations, business promotion and animal rights campaigning. Revived advocacies for banning wildlife trade by many people may result in governments introducing stricter rules to prevent illegal wildlife trafficking; minimize or eliminate wildlife trade which will ultimately improve wildlife conservation (Roe et al., 2020).

Some respondents, though the proportion was smaller, advocated for total destruction of species perceived to be sources of the SARS-CoV-2. This suggests that there are dangers of human persecution of some species in response to the COVID-19 pandemic. Conservationists might therefore consider educating the public on dangers of eliminating the species and efficient methods of preventing future infections. Outreach programmes may focus on those people with low levels of education as advocated by Tarakini et al. (2020). Considering that the threat of species total destruction mainly came from the less educated, a relatively higher proportion could have been detected if it were not for the study’s limitations to those who have internet access, those with appropriate information technology gadgets, and those who are English literate. Destruction of species which are sources of COVID-19 by people may be a significant threat which wildlife conservationists need to assess and also work towards preventing. Indeed, the Convention on the Conservation of Migratory Species of Wild Animals (CMS) (2020) has reported that there are communities that have already sought to cull bats in an “effort” to combat the disease.

The likelihood to visit enclosed wildlife centres was significantly influenced by age and inclination to banning wildlife trade. The likelihood of visiting such wildlife centres was significantly lower in the older respondents when compared to the young and the middle-aged respondents. The older people may feel more vulnerable as has been emphasized by health authorities and governments that they are at greater risk of dying of COVID-19. However, the older people are crucial to tourism as they constitute a significant portion because of their level of wealth, higher discretionary income and greater free time to travel (Littrell et al., 2004). A significant reduction in the older people visiting wildlife centres may negatively impact the tourism sector at national levels. Most of these wildlife centres’ (e.g. captive breeders, anti-poaching, research and environmental education) are sustained by money paid by tourists. Wildlife centres may need to educate the elderly populations on how their safety from COVID-19 is guaranteed in these centres. Furthermore there may be a need to redesign iconic wildlife tourism experiences for the good of both wildlife and humans. For example, in the Democratic Republic of Congo, tourism programme managers are re-evaluating tourism activities because of the need to reduce human contact with habituated apes (Newsome, 2020). Conservation groups indicate that COVID-19 presents an increased risk in the conservation status of African gorilla populations which are already critically endangered. However, noting the skewed distribution of our respondents across the age categories (i.e 21 young, 101 middle-aged and 17 older respondents), we cannot deny that there is a small likelihood that the age effects we detected are merely due to the small size. More data is therefore needed to verify these trends, more so through utilisation of other data collection methods and also engaging people who are offline and those that do not have appropriate gadgets to participate online.

The Role of Media in Framing COVID-19 Information

The framing of a topic usually involves the selection of certain aspects within a perceived reality, moulding and mainstreaming them in a way that promotes a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described (Entman, 1993). Age, gender, employment status, and perceived source of disease were all not significantly associated with major sources of COVID-19 information. These findings suggest that all age groups, gender and employment categories were equally exposed to all sources of information for COVID-19. This is possibly due to family structure in these communities (consisting of different age groups, gender and employment types) through which people live and share information. This makes it easier for wildlife conservation outreach programs and similar campaigns to reach various categories with their information not being biased to specific groups in society. However, social media was used by the majority of respondents as a source of information on COVID-19 in comparison to traditional methods like radio, TV, newspapers, and ministry of health reports. Apart from being an invaluable method for tracking epidemics in the 21st century, various types of social media have been used by experts to investigate public awareness, attitudes and reactions about specific diseases thus informing the decision making process (Smith et al., 2016). The observation also confirms previous findings that people use social media in seeking information about crises (Spence et al., 2006; ). This implies that traditional methods of information dissemination were less effective and people could trust social media for important information delivery. It is therefore highly recommended that wildlife conservationists make use of social media platforms to enhance their impact. This realisation has also seen the World Health Organisation (WHO, 2020) signing up on TikTok social media to reach a wider audience. This further underscores the need to regulate information posted on social media considering the huge numbers depending on it for important information. The public may fail to distinguish authentic information to misinformation. This finding highlights the important role played by social media in influencing societal opinions, perceptions and behavior.

The most common perceived source of COVID-19 was that it was made in a laboratory (by 40% of respondents in this study). Such perceptions highlights the importance of proper media framing of the virus. In the same regard, Macfarlane and Rocha (2020) has argued that much of the ongoing media framing has been poorly crafted especially for bats and pangolins. Message framing of zoonotic diseases requires close collaboration between different stakeholders to avoid inadequately contextualized associations which can mask the intrinsic, ecological and economic importance of wildlife.

Study Limitations

We acknowledge that the data we collected were mainly from people who had access to the internet and appropriate devices which allowed them to respond to the online survey. Such criteria could have left out the views of those without access to such devices, and those who are not English literate. However, due to the countrywide lockdown restrictions, it was difficult for us to use other data collection means. Also, as the level of internet penetration has greatly improved to 60% of the population according to (POTRAZ, 2020) and that many smartphones could be used for this online survey, we therefore feel that these results offered us a good opportunity to assess the immediate response of people to the COVID-19 pandemic.

Implications for Conservation

The findings of the study revealed that perceptions about COVID-19 may both positively and negatively impact wildlife conservation. Advocacy for the stopping of wildlife meat consumption is expected to reduce demand for wildlife meat and poaching through reduced market demand and reduction in poaching. Support for a ban in wildlife trade may also protect wildlife from illegal wildlife trafficking and transfer of diseases across wildlife. However, the advocacy for total destruction of species implicated as sources of COVID-19 may result in human persecution of the species which may ultimately affect their survival. Wildlife conservationists are therefore recommended to conduct outreach programs to prevent persecution of species like pangolins and bats. Outreach programs should target less educated communities as they were the biggest threat. The reduced interest in visiting wildlife centres by older people may negatively impact wildlife based tourism which is sustained by money paid by tourists. Wildlife conservationists might need to rethink some of the activities in these centres that may cause human contact with wild animals exposing them to diseases. Outreach programs are also critical in informing the older people on measures put in place for their safety. Finally, conservation outreach programs need to consider using social media as it proved more efficient than traditional methods. However, there is a need for further research on mechanisms that should be put in place to regulate information on social media platforms to prevent spread of misinformation which may ultimately dilute desired facts.