Introduction

Nature is highly regarded for the ecosystem services it provides, yet the disservices which mainly affect people living next to protected areas are often disregarded in the cost–benefit analysis of nature (Blanco, et al., 2019; Larson, et al., 2019). Ecosystem disservices are the ‘perceived or actual negative impacts on human wellbeing’ of ecosystem functions (Shackleton, et al., 2016). They impact on the economic and non-economic aspects of human life (Blanco et al., 2019). Where performed, human–wildlife conflict management by authorities in the African savanna often addresses challenges with predators and mega-herbivores and sometimes small-medium herbivores including baboons (Papio spp.) and wild pigs (Potamochoerus porcus) (Hill, 1997). The major culprits of crop damage, insects and rodents are often dealt with on a farm-to-farm basis (Makundi, et al., 2005) with major conservation efforts focusing on larger species (Arlet & Molleman, 2007) such as elephants (Loxodonta africana) and baboons.

Rodents are one the most numerous invasive species globally (Buckle & Smith, 1994), damaging maize crops (Zea mays) and other crops at different stages of growth, leading to great economic loss, threatening food security and the potential spread of diseases (Mayamba, et al., 2019). Rodents have mainly been controlled through the use of anticoagulants (which are hazardous to the environment and cause secondary poisoning on rodent-eating carnivores) (Mdangi, et al., 2016), traps (whose effectiveness is low as most of them catch a single rodent at any moment) and cats (which are low in numbers, hence their effectiveness on a large scale is low) (Mahlaba, et al., 2017). Further, domestic cats negatively impact wildlife through predation, harassment, disease transmission and hybridisation (Plaza, et al., 2019). Whilst livelihoods, culture and customs influence management of rodents amongst subsistence farmers (Brown, et al., 2018), the management is mainly based on economic reasons; hence, control is undertaken when there are crops in the fields (Makundi, et al., 2005).

There is a cost which comes with the chemical control of rodents, raising a further challenge for communities living adjacent to protected areas as they already suffer the most from living next to protected areas (Vedeld, et al., 2012). Thus, biological control of pests at no cost to farmers and the use of other locally available pesticidal plants with no cytotoxicity may be the most probable solution to rodent-crop damage for such communities (Mdangi, et al., 2016).

Barn Owls (Tyto alba) are widespread and their diet consists of 99.5% rodents (Kross, et al., 2018), thus a pair and their offspring eat at least 3000 rats a year (Endangered Wildlife Trust, 2006). Hence, they are potential controllers of the rodent population and their use reduces the risk of secondary poisoning on non-target species. Farmers in some parts of the world are using Barn Owls to control rodent populations with high success rates (Kan, et al., 2014; Meyrom, et al., 2009). This method has not been used widely in subsistence farms in Africa where owls are believed to be a bad omen, used in witchcraft (Enríquez-Rocha & Mikkola, 1997). This creates a challenge as other control methods, including other Ecologically Based Rodent Management (EBRM) strategies, are considered expensive and the continent lacks adequate expertise in their use (Makundi & Massawe, 2011).

Farmer’s knowledge (Knowledge is a collection of experience and suitable information and expert understanding which offers a structure for assessing and integrating new experiences and information (Mohajan, 2016) of and perceptions (The way an individual observes, understands, interprets and evaluates a subject, an item or act (Bennett, 2016)) shape their attitudes and responses to crop raiding by wildlife (Anand, et al., 2018) and should be considered when introducing control methods (Hill, 2004; Mulungu, et al., 2015). Farmers’ perception of damage determines the level of control that they are willing to engage. In a study of local people’s perceptions of wildlife in northern Tanzania, Bencin et al. (2016) found that negative attitudes towards wildlife were associated with negative past experiences and fear and not associated with costs or socio-demographic variables. Thus, individuals who perceive higher negative interactions with particular species of wildlife want a reduction in their populations (Kross, et al., 2018). Livestock depredation and crop damage influence negative attitudes in farmers towards perceived problem animals (Noga, et al., 2018).

Cultural beliefs however, determine how people want species linked to their beliefs handled, with species linked to witchcraft being treated inhumanely and those linked to allegory, positive myths and totemic stand-ins being treated positively (Anand, et al., 2018; Buckland & Nattrass, 2019). Animals that are linked to witchcraft are feared in many communities (Enríquez-Rocha & Mikkola, 1997) with some of them used in traditional medicine either to ward off bad omen or bring good luck (Cocks & Møller, 2002; Soewu, 2008). Owls are amongst the negatively viewed animals as they are believed to be a messenger of death, to be used on magic medicine, a mythical creature created by witches and wizards and also believed to bring misfortune (Mikkola & Mikkola, 1997). There is clearly a trade-off between the regulatory services by Barn Owls and the cultural values linked to witchcraft and bad omen.

Rodent control to minimise populations, hence, to prevent losses in crop production, is a key strategy to achieving food security, agro-ecological sustainability and economic development amongst small-holder farms (Swanepoel, et al., 2017). In this study, we assessed the community’s perceptions on owls and determined their willingness to adopt the use of Barn Owls as a possible biological control method for rodents. We investigated the effect that gender, age and level of education (Cailly Arnulphi, et al., 2017) and distance from the protected area had on the people’s attitude towards the use of owls in rodent control. The effect of village of origin on attitude towards rodents and owls was also assessed. Parental influence on the responses of teenagers was also tested (Clark, et al., 2017).

We expected that older people and less educated individuals would have negative attitudes towards owls as they are expected to be more influenced by traditional beliefs. The women in the household may also be more susceptible to witchcraft rumors as they seem to be more prone than men to accusation, in particular elderly women (Eboiyehi, 2017). The perceptions of parents were not expected to influence those of teenagers, in particular due to access to education as well as different sources of information. People from villages closer to the protected area were expected to have positive attitudes towards owls, either because they were more exposed to awareness campaigns or because the effect of rodent-eating wild species may be more visible closer to the conservation area.

Methods

Study Area

The study was conducted in the communal areas of Ward 14 and 15, bordering gazetted protected areas within Hwange District in northwest Zimbabwe, an area which is part of the south-eastern section of the KAZA Trans-frontier Conservation Area (TFCA) in Zimbabwe. This area of study borders Hwange National Park, Main Camp area (HNP) (14 561 km²), centred on 19°00'S, 26°30'E (Tarakini, et al., 2018) and Sikumi Forest Area (SFA) (544 km²), a photographic and hunting area, both protected by legislation through the Parks and Wildlife Act (Chapter 20:14) and the Forest Act (19:04), respectively.

Neither protected area is fenced, but nevertheless separated from the communal land by a small buffer zone being a railway line and a road, respectively (Figure 1), thus likely resulting in the high numbers of reported human–wildlife conflict cases (Le Bel, et al., 2010; Loveridge, et al., 2017). Rodent crop damage was not reported as conflict, although rodents generally occur in the area and outbreaks occur almost on a decade cycle (Leirs, et al., 2010; Mazarire, 2016). Although the area is prone to conflict from mega-fauna and primates, local communities still practise subsistence farming. The need for this is also exacerbated by the unstable Zimbabwean economy which has threatened livelihoods and left many persons partly dependent on natural resources to buttress their livelihoods (Chirau, et al., 2014). The Hwange District as a whole has adopted the Communal Areas Management Programme for Indigenous Resources (CAMPFIRE), although the programme has been crippled by repeated political and economic crises in the last two decades.

Figure 1.

Map of the area where the survey was undertaken, showing the homesteads of respondents and protected areas.

The area under study falls within natural eco-regions IV and V, which are semi-arid areas with a mean annual rainfall of 600 mm or less in the wet season (Chamaillé-Jammes, et al., 2006). The area is characterized by Kalahari sands (Childes & Mundy, 2001), and as a result the crop yield is minimal although people in these communal areas rely on subsistence farming with the main crops being maize, sorghum (Sorghum bicolor) and pearl millet (Pennisetum glaucum) (Guerbois, et al., 2013). The threat to livelihoods is thus further affected by crop raiding and livestock preying by wild animals. The area has historically been inflicted with bubonic plague, a zoonotic disease transmitted through rodent fleas from both indigenous and exotic rodents and has experienced rodent outbreaks which have caused problems to subsistence farmers (Munyenyiwa, et al., 2019). Rodent species that have been identified as carriers in southern Africa include Mastomys natalensis and gerbil species (Munyenyiwa, et al., 2019).

While protected areas are viewed as source of wildlife, negatively affecting the crop fields, the economy of the area also relies on these protected areas through tourism activities (Guerbois & Fritz, 2017). The human population has been on the increase (Guerbois, et al., 2013), resulting in the creation of more homesteads and a reduction of grazing lands, thus negatively impacting on the natural vegetation. At the time of study, the areas had 1467 homesteads with a population of 7430 people. As a result, livestock is grazed 3 km into the SFA with permission from the Forestry Commission (Guerbois, et al., 2013).

Data Collection and Processing

Data on perceptions were collected through the stratified random administration of questionnaires and was targeted at people living within the designated study area. Permission to conduct the study was obtained from the District Administrator. Questionnaires were administered through trained assistants, who recorded all responses. A pilot study was conducted with assistants in the team to standardize the technique and thus, to reduce bias. Questions were asked in the language preferred by the respondent (Nambya, Ndebele, Tonga, Dombe or Shona). All visited homes were recorded using a Geographic Position System and their locations are shown in Figure 1.

Responses to open ended questions were recorded exhaustively to avoid recorder bias. In each homestead, everyone above the age of 14 years who was available took part in the survey, and each respondent was asked questions in the absence of other respondents to avoid duplication of responses. This was done to allow for the assessment of the effect of age on their responses. The effect of gender on the homestead’s overall perceptions was also considered as household composition varied greatly (Webbink, et al., 2012). All responses were re-classified a posteriori to reduce the number of modalities in each variable while retaining as much information as possible. Thus, the results may partly be affected by our interpretation of the responses as qualitative analyses are always liable to researcher subjectivity (Guerbois, et al., 2013). The population in the area of study comprised mainly five ethnic groups namely, Nambya, Ndebele, Tonga, Dombe and Shona (Nhongo, 2014).

Socio-demographic data were collected through structured responses which allowed for easy categorisation into different variables in order to analyse their influence on perceptions by the study subjects (Table 1). The respondents were first asked if they were experiencing a rodent problem and the responses were binomial (yes/no). Once this was established, they were then asked if they would be comfortable with the use of Barn Owls in rodent control and this was also a binomial response variable (yes/no). Their responses to this question were categorised into articulated values, a way of classifying ecosystem services values across different analogies of human–nature relationships as suggested by Arias-Arévalo et al. (2018) (Table 1), although these authors only consider articulated values as positive gains from nature; the costs of nature to humans are ignored. In our case, the negatives were considered and were classified as a negative articulated value. The aim was to understand the depth of their myths and beliefs on owls. It was then investigated if age, gender, level of education, village of origin and distance of their homestead from a protected area, that is, either HNP or SFA influenced their perceptions on the use of Barn Owls as agents in rodent control. The effects of village of origin and homestead on attitudes towards owls were investigated. Parental influence on perceptions was assessed by comparing the perceptions of the parent with those of the teenagers in the homestead.

Table 1.

Categorization of Informant’s Socio-Demographic Variables and Attributes.

Data collection was done without prior knowledge of the village demarcations. Their accessibility determined the number of homesteads included in the survey in each village.

Results

Use of Owls in Rodent Control

Of the respondents, 84% claimed to be facing a rodent problem as a result of the presence of rodents in the area. Fifty-seven percent of the surveyed people had no issue with the use of owls in rodent control, 41% had an issue and 2% did not respond to the question (Table 1). More female respondents had an issue with the use of Barn Owls in rodent control than males (χ² = 9.001, df = 1, p = .003), that is, they did not want to use them. Level of education (χ² = 0.203, df = 1, p = . 652) and age (χ² = 2.51, df = 3, p = .474) had no influence on their perception of the use of owls in rodent control. People had an issue with the use of owls in rodent control increasing with distance from the protected area (df = 573, Est. = 0.0001 ± 0.00,004, p = .003).

The articulated values expressed varied with gender (χ² =12.285, df = 4, p = .015) with more females expressing issues of sacredness and more males expressing ecological resilience as their answers on the use of owls in rodent control (Figure 2). Males against the use of owls mostly expressed health and social cohesion (Figure 2). More female respondents were further from the protected area than males (Figure 3).

Figure 2.

The articulated values of the community on the use of owls in rodent control and the factors influencing them. The arrows show where related variables lie in relation to other attributes. Variables linked to certain articulated values lie close to those values. The articulated values were plotted as Resil. (ecological resilience), Sacredness (sacredness), Moral (moral duty), Health (health) and Soc. cohesion (social cohesion). The responses to whether people had an issue or not were plotted as IssuO.Yes (yes) and IssuO.No (no). Age was plotted as Age.U18 (teenager), Age.18–30 (young adult), Age.30–40 (30 to 40) and Age.A40 (above forty).

The probability of both females and males wanting rodents eradicated and having an issue with the use of owls in rodent control increased with distance (Figure 3). There were more females wanting rodents eradicated than having an issue with the use of owls closest to the park. Yet amongst males, most of them had an issue with the use of owls than those wanting rodents eradicated. The probability of females in both cases was higher than that of males with distance.

Figure 3.

The results of a generalised linear model showing the effect of gender on the desire for rodent eradication and perception on the use of owls in their eradication. The lines represent the predicted probability in relation to the distance of the respondent’s homestead to the protected area.

The responses on whether people had an issue with the use of Barn Owls in rodent control or not were positively spatially correlated (statistic = 0.6039), observed rank = 575, p = .0417), suggesting a mild village or neighbour effect. Accordingly, there was a significant difference in how different villages felt about the use of owls in rodent control (χ² = 15.931, df = 5, p = .007). The pairwise nominal test did not reveal the differences (p > .05) between villages. Nesting the variables within the homestead and village, respectively, in the GLMM showed that only gender influenced the attitude towards the use of owls (Table 2), with more females being against the use of owls than males.

Table 2.

Results From a Generalized Linear Mixed Model Showing Factors Influencing the Acceptance of the Owls in Rodent Control Within the Village and Homestead. Where σ² is the Residual Variance, τ00 is the Random Intercept Variance, the ICC is the Intraclass Correlation Coefficient N is the Total Number of Observations (i.e. Villages and Homesteads).

Discussion

This study focussed on understanding the perceptions of people living next to protected areas on rodent damage and the possible use of Barn Owls in the control of the rodents. It is important to consult and involve the people living with wildlife in the introduction of any strategies. Thus, the understanding of local ecological knowledge (LEK) (knowledge obtained through individuals’ observations over their lifetimes (Gilchrist, et al., 2005)) and traditional ecological knowledge (Knowledge, practice and belief that pertains to the relationship of living beings and with their environment which is handed down through generations (Berkes, et al., 2000)) is key in the success of biodiversity conservation (Charnley, et al., 2007).

Witchcraft is a practice which is believed throughout most parts of Africa (Eboiyehi, 2017) and nocturnal animals are believed to be used as tools in the witches’ trade. These animals include owls, Spotted Hyaena (Crocuta crocuta) and the Aardvark (Orycteropus afer) (Adeola, 1992). In believing in the practice of witchcraft, the people also believe in identifying witches, leading to witch hunts which have at times been fatal (Adinkrah, 2004). Elderly women are often on the receiving end of this practice in sub-Saharan Africa (Atata, 2019). A study in Kenya documented witch hunts in a month which resulted in 11 people (8 women and 3 men) being labelled witches (Federici, 2008).

Although environmental education has been around for over 30 years, it is only recently that it has started to involve women (Gough, 2006). Their participation in environmental issues is still lower than that of their male counterparts. This could also be a result of the limited time they have to attend awareness meetings as they are primarily responsible for the day to day running of their homes (Abdelali-Martini, 2011; Graham, et al., 2016). This limits the knowledge they have on environmental issues. Thus, it is not surprising that the women living next to Hwange National Park were reluctant to have owls being used to control rodents in the area. Most of them clearly stated they feared that if the owl was seen in their field or homestead, they would be ‘labelled’ a witch. As a result, more men than women express positive articulated values.

Level of education and age did not influence respondents’ perceptions on owls. Other studies have shown that African people are generally a spiritual people, believing in the power of the spiritual and occult forces. A study focussing on the use of spiritual healing churches in Nigeria revealed that more educated people prefer spiritual churches whilst less educated people prefer traditional healers (Adegoke, 2007). This has also been found to be the case amongst people of other ethnicities. Thus, education does not impact on people’s mythical beliefs and if these are to be overcome, various methods other than education must be employed (Pasachoof, et al., 1970). It has been proven that these beliefs in witchcraft work to the disadvantage of the poor as it affects the implementation of developmental projects and the improvement of livelihoods (Kohnert, 1996). This is seen even with the reluctance to adopt owls in rodent control without considering the benefits. Other studies still advocate for improved environmental education as a way of dealing with perceptions (Williams et al., 2021).

Human settlement areas can be good habitats for owls because they are good for commensal rodents, which can in turn benefit the farmers but the negative attitudes by farmers make it difficult to explore this (Ogada & Kibuthu, 2008). A study by Ogada and Kibuthu (2008) showed that 68% of a community did not identify with beliefs on owls yet their knowledge of owls did not positively impact their behaviour towards owls and the authors attribute this to a lack of ecological literacy and it is important that farmers understand ecological processes. The belief of witchcraft transcends age and the beliefs are held across age groups.

The people living closer to protected areas were more in favour of the use of owls in rodent control. People living closer to protected areas experience more human wildlife conflict (Megaze, et al., 2017) and would be expected to have more negative attitudes towards wildlife. Environmental awareness programmes tend to focus on people living next to wildlife and as a result they exhibit more ecological literacy. The tourism industry in and around protected areas mostly hires from communities living next to protected areas, this not only influences their attitude towards wildlife but could also increase their ecological literacy (Digun-Aweto, et al., 2019). In some instances, people living furthest from protected areas have a positive attitude towards wildlife as they experience less loss due to wildlife conflict (Karlsson & Sjöström, 2007). Human wildlife conflict tends to be highest closest to protected areas and decreases with distance (Karlsson & Sjöström, 2007; Karanth et al., 2013), although in some cases the opposite is true as people living near protected areas are more involved in conservation oriented development projects and other ecotourism based benefits (Infield, 1988). Interestingly, a Human Elephant Conflict (HEC) survey (Sampson, et al., 2019) revealed that farmers living closest to the protected area, thus experiencing more conflicts, were most supportive to elephant conservation. Environmental education thus fosters a positive attitude towards wildlife as it imparts information on environmental values (Guerbois, et al., 2013).

Parents/guardians were found to have some influence on the perceptions of teenagers. The teenagers’ perceptions on owls mirrored those of their mothers. This would suggest that when it comes to mythical beliefs children mirror the beliefs of the parent they spend more time with. Children have been shown to spend more time with their mothers who provide primary care (Pedersen, 2012). As a result, they feel they receive more positive treatment, emotions and closeness from their mothers, allowing them to learn more from them (Miller & Lane, 1991). It may also be that the fear of the unknown is more powerful than reason.

The environment in which people live tends to influence their perceptions. Gifford and Nilsson (2014) described the social influences on pro-environmental concern and behaviour and these include religion, local norms, social class, proximity to the problematic sites and cultural variations. The villages also varied in their attitudes towards the possible use of owls in rodent eradication. This could be a result of the differing levels of environmental and conservation consciousness, mostly influenced by effort put in raising awareness on conservation and proximity to problem sites (Gifford & Nilson, 2014). The pairwise analysis did not show these differences possibly due to low sampling effort in some areas.

The farmers’ perceptions must be considered in the use of owls as they shape the attitudes and responses of the people to crop raiding by different species and to the negative beliefs on owls and witchcraft (Hill, 2004). Increased participation of the community in Ecologically Based Rodent Management through improved knowledge on rodent behaviour, their predators, environmental manipulation to deter rodents and improved attitudes could help deal with perceptions (Makundi & Massawe, 2011; Williams et al., 2021).

Implications for Conservation

There is a wealth of knowledge on wild animals believed to be used for witchcraft and on witchcraft in general. It is clear that these beliefs produce negative attitudes toward the animals (Enríquez-Rocha & Mikkola, 1997) to the detriment of the species or leading to a failure to harness ecosystem services offered by the species in question. It is therefore important that conservation cautiously approaches these misconceptions in an effort to break them. Communities in Chile have positive perceptions of owls which are based on aesthetical and ethical values but not serviceable ones (Godoy-Güinao et al., 2017). The expressed values allow for the conservation of the birds, hence allowing the communities to benefit also from the ecosystem services they provide without necessarily being aware of them. This highlights the need for improved education and awareness even in areas where owls are not persecuted to create a better appreciation of their value.

The study highlights the factors influencing attitudes towards owls (with Barn Owls used as a proxy in this study), thus indicating to conservationists the important aspects to consider in efforts to confront the myths on these birds. The study shows that in Zimbabwe, like other countries women are the most vulnerable to witchcraft labels (Atata, 2019; Eboiyehi, 2017) and any conservation efforts aimed at dealing with these issues needs to be strategized accordingly. The study also highlights the need to effectively distribute awareness programmes as educating only those closest to the protected area does not effectively protect the wildlife. The findings call for a deeper understanding of people’s relationship with certain species. Values and relationships are complex and need to be investigated further. Other factors that need to be investigated include ethnic group and religion.