Forested landscapes across mainland Africa and Madagascar support an exceptionally diverse assemblage of carnivorous mammals (Happold, Reference Happold1996; Ray, Reference Ray, Webber, White, Vedder and Naughton-Treves2001). Varying in size, food preference and home ranges, these carnivores have important ecological roles and yet are in most cases poorly known (Ray, Reference Ray, Webber, White, Vedder and Naughton-Treves2001). Despite this, they are cited in discussions of ever-growing anthropogenic impacts, including the effects of local hunting practices. Examples are the unsustainable hunting of Cryptoprocta ferox in north-east Madagascar (Golden, Reference Golden2009), local extinctions of Panthera pardus in various locations, such as south-west Cameroon (Abugiche, Reference Abugiche2008), and local extinctions of Civettictis civetta at several sites in Ghana (Ryan & Attuquayefio, Reference Ryan and Attuquayefio2000).
Carnivores are a keystone species in forest ecosystems and so their absence or reduction can distort the trophic balance (Bond, Reference Bond1993). If hunters are exploiting these species to fulfil subsistence or revenue needs, then understanding the full extent of such actions will enable us to prevent total depletion and to plan better for substitute food resources across Africa (Bennett et al., Reference Bennett, Blencowe, Brandon, Brown, Burn and Cowlishaw2007).
Our objectives were to assess whether carnivores are commonly hunted across the forests of Africa and Madagascar for reasons beyond simple nuisance control or for their pelts. Our analysis provides a collation and assessment of information on the effect of hunting practices on carnivores. Our focus is on the purpose and means of hunting, along with the taxa most affected.
The term used to search for relevant articles included countries and carnivore taxa (Supplementary Material 1). Google Scholar (Google, Mountain View, USA) was used as the search engine because it examines all text fields in a large array of databases rather than only titles and abstracts. Google Scholar has been used previously for meta-analyses (e.g. Branton & Richardson, Reference Branton and Richardson2011; Lelieveld, Reference Lelieveld2013). Each search included the phrases ‘bushmeat’ and ‘bush meat’ (bushmeat describes any wild animal used for human consumption; Bennett et al., Reference Bennett, Blencowe, Brandon, Brown, Burn and Cowlishaw2007). This search may miss articles referring to the hunting of carnivores solely for other purposes, such as cultural ceremonies, or articles that do not refer to wild meat as bushmeat. However, through the use of this search and many citation chains we believe that the majority of relevant studies were discovered. The search results included studies focused on alternative uses for carnivores, further suggesting that the search was thorough. Most of the research on local hunting has been referred to as bushmeat hunting and therefore even non-bushmeat studies often reference bushmeat research. The final search was completed on 1 March 2012.
We located and reviewed a total of 62 articles published during 1974–2012 (Supplementary Material 2). These included 337 reports of the hunting of carnivores. All articles were restricted to forest habitats in Africa or Madagascar (Fig. 1). Madagascar was included because its carnivores, although taxonomically separate from those of mainland Africa, occupy similar niches and are hunted for similar reasons. For a few studies conducted across multiple habitat types, such as whole parks, it was difficult to exclude non-forest locations unless these were clearly distinguished.
There was a higher frequency of articles in more recent years, probably a result of increased research efforts and improved research techniques. Information was extracted on every carnivore taxon specifically cited by researchers as being hunted within their study area. Each taxon is considered one record in the analysis. The items noted for each record were: article citation, size class and species of carnivore, purpose of hunting, hunting method, method employed by researchers, political location and habitat of study area.
Only primary data, collected using one of the recorded research methods, were included in the analysis (Supplementary Table S1). Hearsay records, most books, many non-reviewed conference reports, and references to accounts of earlier explorers where the original publication could not be located, were discarded.
Size class was determined by a taxon's mean body weight, defined as small (< 6 kg), meso (6 < 23 kg) and large (> 23 kg; Animal Diversity Web, 2012). Family level groupings (Table 1) were formed with the taxonomy used by IUCN (2012; Supplementary Table S2). Civettictis civetta was separated from Viverridae for the purposes of this study because it is biologically dissimilar from other members of its family. Being large, the ecological role of C. civetta is different from its smaller relatives and it is perceived differently by local hunters. Its high prevalence in our literature search warrants particular interest. The African palm civet Nandinia binotata belongs to a monotypic family and, like C. civetta, is of particular interest because of its high prevalence in reports of offtake. Records that could not be classified at the family level as a result of vagueness in identification were not included in family level analyses.
The variables hunting purpose and method, and research method, were not mutually exclusive. Data were presence/absence and a record could be included in multiple categories. For example, if a researcher used market surveys and hunter interviews then both would be tallied under research method.
Recorded purposes for hunting are given in Supplementary Table S3: consumption, culture (use other than as food), pest (unwanted animals, such as nuisance species), and unknown (reason for hunting not specified). The hunting methods recorded are given in Supplementary Table S4.
Records involving small carnivores were dominated by herpestids and viverrids. Meso carnivores were dominated by C. civetta. Large carnivores, such as felids and hyaenids, were dominated by P. pardus (Table 1). A summary of the purposes of hunting is given in Table 2. Consumption comprised the majority of all records and was reported for all size classes, and was the primary purpose for hunting seven of the nine taxonomic groups. Records of C. civetta were associated primarily with consumption (71.2%), and of records reporting consumption 81.0% were for sale, and of these 47.1% were for urban sale. Records of N. binotata were also associated primarily with consumption (84.6%), and of records reporting consumption 81.8% were for sale, and of these 40.7% were for urban sale.
* Total number of reports differs between the family and overall because some reports (such as those for small carnivores) were not sufficiently specific to be included in the family level analysis but were included in the overall analyses.
The majority of records of consumption were of Herpestes, followed by species of viverrids and C. civetta. Cultural use was the second most commonly reported purpose, with herpestids and then viverrids the most commonly reported families. Nearly a third of the records of large carnivores were taken for cultural purposes (Table 2). The highest percentage involved herpestids and then viverrids. Cultural uses were recorded for > 20% of all observations for five out of the nine groups. Twelve of 33 purpose observations of the large-size class were for cultural use. Sale for fetish uses comprised 18.3% of all reports for cultural use. Pest control was the third most frequently reported reason for hunting, with felids most commonly reported for this purpose, followed by viverrids and mustelids. N. binotata was not hunted as a pest. Records reporting sale of any kind (consumption, fetish or unknown) composed 56.6% of all 488 purpose observations. Carnivores taken for sale were sold for consumption (82.3% of all reports of animals taken for sale and 46.5% of all animals taken for a reported purpose) and for fetish use (7.6% of all reports of animals taken for sale and 4.3% of all animals taken for a reported purpose). Personal consumption was reported in 182 records and consumption for sale in 227 records. Of the sales records 50.7% were for local sale and 47.6% for urban sale. Herpestidae (31.1%) and Viverridae (22.6%) contributed most to records of sale in urban areas. Sale in urban areas for reports published between 1974 and 2009 is shown by decade in Fig. 2. Of the 108 records for urban sale, 99 are from studies conducted in the 1990s and the 2000s.
The numbers of reports by hunting method are presented in Fig. 3. Traps were the most popular hunting method (31.0% of records), followed by guns (23.7%). Traps were the primary method of hunting for five of the nine groups. However, for the other four groups, N. Binotata was reported as taken equally by guns and traps, and Canidae, Hyaenidae and Felidae were all reported as being taken by unknown methods more frequently than by traps.
There were only 28 reports of night hunting but 64.3% of these were from records associated with sale for consumption, and 43.0% were for urban sale for consumption (Table 3). Only two reports of night hunting did not also report the use of guns. Of the 160 reports of the use of guns, 89.4% also reported sale of some kind. Additionally, 51.8% of all sale reports were associated with use of guns, and 48.2% of reports of urban sale for consumption were associated with the use of guns.
1 Includes sale for consumption, sale for fetish uses, and unknown uses.
2 Includes records reporting sale for consumption through local, urban, or unspecified markets.
Carnivores specifically defined by the original authors as rare but still hunted totalled 17 reports in eight articles. Carnivores specifically defined as locally extinct totalled 11 reports in seven articles. Civettictis civetta was the carnivore most recorded as rare and P. pardus was the carnivore most recorded as locally extinct (Supplementary Table S5).
Pest control was in general not the sole reason reported for killing carnivores. Even in areas where carnivores were considered village marauders and poultry killers, they were also still considered a protein source (Campbell, Reference Campbell2009, Ghana; Djagoun & Gaubert, Reference Djagoun and Gaubert2009, Benin; Kotschwar et al., Reference Kotschwar, Gerber, Karpanty, Justin and Rabenahy2014, Madagascar). Where nuisance was the sole reason reported (e.g. Carpaneto & Fusari, Reference Carpaneto and Fusari2000), species were not eaten because it was either taboo to do so or they were poor tasting.
Nearly a quarter of records that gave reasons for taking carnivores reported cultural uses, primarily as skins or in fetish practices. Uses of fetish products for spiritual vs medicinal practices are often difficult to distinguish. For example, skins of genets, civets, and felids are used in utilitarian and ornamental fashions (Maisels et al., Reference Maisels, Keming, Kemei and Toh2001). The Mbuti people of the Democratic Republic of the Congo (DRC) use genet and civet skins in wrist guards for archery and for hats used in initiation ceremonies (Carpaneto & Germi, Reference Carpaneto and Germi1989). Genetta servelina and Genetta pardina skins are used as a mark of importance in Cameroon (Pollard, Reference Pollard1997). Genets are used as medicine in Ghana (Ntiamoa-Baidu, Reference Ntiamoa-Baidu1987), the oil from C. civetta glands is used against respiratory problems in Cameroon (Laurent, Reference Laurent1992), and Crossarchus obscurus, and Herpestes ichneumon are used in animist rituals in Benin (Djagoun & Gaubert, Reference Djagoun and Gaubert2009). In Nigeria leopard parts are believed to increase fertility, protect against or invoke witches, and to counter snake venom (Sodeinde & Soewu, Reference Sodeinde and Soewu1999). Otters are used in Tanzania as medicine and belt buckles (De Luca & Mpunga, Reference De Luca and Mpunga2005), and in Central and West Africa otter fur is believed to render the bearer invisible to enemies (Jacques et al., Reference Jacques, Veron, Alary and Aulagnier2009).
Many supposedly protected carnivores used in traditional medicine are commonly and openly sold throughout West Africa. Ntiamoa-Baidu (Reference Ntiamoa-Baidu1987) notes that carnivore parts are sold for traditional medicine in at least one location in all major towns in West Africa. Sodeinde & Soewu (Reference Sodeinde and Soewu1999) found that 77% of fetish market stalls in Nigeria carried the protected serval Leptailurus serval. Hunters in southern Benin use fetish markets as their primary means for selling small carnivores, especially N. binotata, Lutrinae and Viverridae (Djagoun & Gaubert, Reference Djagoun and Gaubert2009). This prevalence of sales for fetish purposes implies animals once used within a family or village for local traditions are now sold at a regional scale, suggesting a large network of trading for this purpose and growing demand.
Consumption was the most frequently reported reason for hunting but it is difficult to determine who is consuming the meat. Some areas have a religious prohibition against eating certain carnivores whereas in other areas the same species command high prices. Taboos preventing hunting of various carnivores were present in areas such as Equatorial Guinea (Keylock, Reference Keylock2002), Republic of the Congo (Mbete et al., Reference Mbete, Banga-Mboko, Racey, Mfoukou-Ntsakala, Nganga and Vermeulen2011) and Madagascar (Kotschwar et al., in press). Many taboos referred to pregnant women. Genets in Cameroon, for example, are said to cause mother and child to catch a cough, and C. civetta to cause a child to be born with body odour (Ayeni et al., Reference Ayeni, Mgaihli and Ebot2001). In DRC both N. binotata and C. civetta are thought to cause developmental anomalies in children (Carpaneto & Germi, Reference Carpaneto and Germi1989). Taboos can also have a spiritual basis. Where leopards are considered a totem animal, villagers believe they can transform into totem species, and therefore killing a leopard could kill the totem's owner (Nzouango & Willcox, Reference Nzouango and Willcox2000; Abugiche, Reference Abugiche2008; Wright & Priston, Reference Wright and Priston2010).
Although taboos may protect carnivores, the belief that physical and mental ailments are associated with meat intake may be applicable only to a gender or age group, such as child-carrying women. These beliefs are waning (Nzouango & Willcox, Reference Nzouango and Willcox2000; Wright & Priston, Reference Wright and Priston2010) because of increased interaction with outside education and beliefs (Sifuna, Reference Sifuna2012), and perhaps from an increased need for meat or money.
Of reports of consumption 60.5% cited immediate personal use. Noss (Reference Noss1998), Vliet & Nasi (Reference Vliet and Nasi2008) and Keylock (Reference Keylock2002) noted that carnivores are often kept and eaten within the hunter's family. Carnivores are also eaten in the forest during extended hunting trips (Kümpel, Reference Kümpel2006) because they are less favoured by traders and buyers (Noss, Reference Noss1998; Nzouango & Willcox, Reference Nzouango and Willcox2000). Personal consumption of carnivores could negatively affect carnivore populations in forested Africa if these less desirable animals are a major food source for the hunter and his family. Halle (Reference Halle, Mainka and Trivedi2002) reported that in South America the more desirable meat goes to market and the less favoured meat stays within the hunter's family, which invariably results in the ‘urban markets indirectly generat[ing] additional pressure on less preferred species.’
Carnivores compose a small but consistent portion of a hunter's total sales, with 75.4% of reports of consumption associated with sale. Examples of the percentage of carnivores of total sales are: 1.71% (Juste et al., Reference Juste, Fa, Val and Castroviejo1995), 3.8% (all viverrids; Martin, Reference Martin1983), < 5% (Kümpel, Reference Kümpel2006), and 18% (Keylock, Reference Keylock2002). At the other extreme, Kümpel (Reference Kümpel2006) listed N. binotata and G. servalina as two of the most frequently seen species at market. Bushmen of south-east Nigeria traded carnivores and artiodactyls more than any other animal type (Angelici, Reference Angelici1999), and Atilax paludinosus was one of the six most expensive meats per unit mass in eastern Cameroon (Tieguhong & Zwolinski, Reference Tieguhong and Zwolinski2009). Carnivores appear to be increasingly sold in urban areas. Bassett (Reference Bassett2005) reported that carnivores increased in the bushmeat trade in northern Côte d'Ivoire from his first survey in 1981–1982 to his second in 1997–1999. This increase was attributed to depletion of other wildlife, especially of large, preferred herbivores. These examples support the notion that as preferred species are depleted a general shift of trade towards less desired animals, including carnivores, is likely.
The hunting method was unknown for 16.6% of reports (Supplementary Table S4). For known hunting methods however, trapping, predominantly by snares, was the most frequent catch technique (Nzouango & Willcox, Reference Nzouango and Willcox2000; Fa & Yuste, Reference Fa and García Yuste2001; Golden, Reference Golden2009). Trapping may be indiscriminate, capturing any species, but Abugiche (Reference Abugiche2008) found that snares selectively trap C. obscurus, and Fa & Yuste (Reference Fa and García Yuste2001) found that carnivores were particularly vulnerable to snares, suggesting that the discrimination of a trap may depend on its design and/or placement.
Guns were the hunting method in 23.7% of reports. Use of guns is currently expanding to locations such as Cameroon (Willcox & Nambu, Reference Willcox and Nambu2007), Guinea (Ziegler et al., Reference Ziegler, Nikolaus and Hutterer2002) and Ghana (Crookes et al., Reference Crookes, Ankudey and Milner-Gulland2005). The percentage of carnivores killed with a shotgun rose from c. 30% in 2003 to c. 80% in 2010 in Equatorial Guinea (Gill, Reference Gill2010). This rise was attributed to an increase in income, enabling more hunters to afford guns and cartridges. Similarly, Kümpel (Reference Kümpel2006) argued that waxing accessibility to guns and waning trapping success caused hunters to transition from traps to guns. Over half of all reports of sale of bushmeat for consumption were associated with guns, and almost 90% of all gun reports were associated with sale of some kind, suggesting that guns are being employed almost exclusively where hunting is for profit. If hunting for sale is increasing, the percentage of hunters with guns will probably also increase.
Night hunting (jacklighting), in which a hunter uses a light to halt an animal and shoot it (Hennessey & Rogers, Reference Hennessey and Rogers2008), was reported in 4.1% of records. Hunters in the Republic of the Congo reported jacklighting as one of the easiest means of hunting, and civets as one of the two main animals targeted (Hennessey & Rogers, Reference Hennessey and Rogers2008). Likewise, hunters in Gabon reported night hunting as being more efficient and that small nocturnal carnivores such as C. civetta, N. binotata, Bdeogale nigripes and Poiana richardsoni are mostly killed this way (Vliet & Nasi, Reference Vliet and Nasi2008). About 64% of all night hunting records were associated with sale. This is not as strong an association as with all gun methods but rising sales would logically lead to increased night hunting, for if hunters are transitioning to seeking smaller prey then more profitable hunting methods would be desirable.
Government curbs on these hunting methods may be effective in many locations but appear readily ignored in others. Snares, although illegal, are the most popular method of hunting in the Republic of the Congo (Hennessay & Rogers (Reference Hennessey and Rogers2008). Vermeulen et al. (Reference Vermeulen, Julve, Doucet and Monticelli2009) reported snaring, guns, and night hunting as illegal, but used commonly. Wilkie et al. (Reference Wilkie, Sidle and Boundzanga1992) noted that illegal jacklighting and non-traditional snares are ‘openly employed and are acknowledged as the most common, preferred, and effective techniques’.
Taxa affected by hunting
Civettictis civetta is thought to be widely distributed throughout forested Africa and is categorized as Least Concern on the IUCN Red List (IUCN, 2012), although it is commonly taken for sale of its musk to the perfume industry. The relatively large size of this civet (7–20 kg), and its terrestrial habits, make it a desirable and easy target (Carpaneto & Germi, Reference Carpaneto and Germi1989). Laurent (Reference Laurent1992), Lupo & Schmitt (Reference Lupo and Schmitt2002), and Mbete et al. (Reference Mbete, Banga-Mboko, Racey, Mfoukou-Ntsakala, Nganga and Vermeulen2011) report that it is one of the most commonly hunted species. Carpaneto & Fusari (Reference Carpaneto and Fusari2000) found that C. civetta is the only animal to be eaten in hunting of nuisance animals in Tanzania. Djagoun & Gaubert (Reference Djagoun and Gaubert2009) noted the high resale value of C. civetta in the fetish markets of southern Benin. In addition to the five occurrences of known rarity for C. civetta and one of local extinction (Supplementary Table S5), 13 studies within the species range (IUCN, 2012) failed to report it, suggesting that this presumed widespread species is undergoing localized depletion.
The Least Concern N. binotata (IUCN, 2012) is an easy target for hunters because of its conspicuousness and gregariousness (Okiwelu et al., Reference Okiwelu, Ewurum and Noutcha2009). Hunters net large numbers of this frugivore as they descend from fruiting palms. This may explain why it is one of the most common species sold in markets (Kümpel, Reference Kümpel2006), where it is also the only carnivore commonly sold (Foerster et al., Reference Foerster, Wilkie, Morelli, Demmer, Starkey and Telfer2011), and why it is one of the most expensive (Gill, Reference Gill2010). Although N. binotata is considered a common species in many places, 22 of the 58 studies within the species’ range (IUCN, 2012) did not report it as present, suggesting it is also undergoing localized depletion.
Although Herpestidae account for a high percentage of records, harvest relative to other carnivores may be inflated because of the high numbers of this family in a given study area. Three herpestids (marsh mongoose A. paludinosus, black-footed mongoose B. nigripes and dwarf mongoose C. obscurus; Laurent, Reference Laurent1992; Pollard, Reference Pollard1997; Puit et al., Reference Puit, Huart, Leroy and Nsangou2004) comprised 5.0% of all records. Mongoose species are commonly hunted (Lupo & Schmitt, Reference Lupo and Schmitt2002; Carpaneto et al., Reference Carpaneto, Fusari and Okongo2007) but it is not easy to assess any impact because hunters are unaware of species distinctions and because of the difficulty in identifying smoked or butchered specimens.
The family Viverridae (excluding C. civetta) presents similar uncertainties: 15 of 64 references to the genus Genetta were for ‘genet species’, and genets are often grouped together as one taxon. This is doubtless because these often solitary, 1–3 kg species are frequently mistaken for one another (Nowak, Reference Nowak1999). Regardless of this uncertainty, species of Genetta were observed in many studies. Genetta servalina, the most reported species, was noted as the most frequently sold species in at least one location.
The families Felidae and Mustelidae appear to be commonly hunted, and the most hunted felid is the Near Threatened leopard. No species of Mustelidae represented more than 3% of total records but species of the subfamily Lutrinae are commonly used for cultural purposes. There are relatively few records of the family Eupleridae, found solely in Madagascar, being hunted.
Biases in this analysis potentially arise from the keywords used in the search and non-random coverage in the literature of the topics we address. Most studies of localized hunting concern bushmeat and therefore most articles report the purpose of take to be for consumption. Only a few studies focused primarily on cultural uses. However, research that included structured interviews, questionnaires or anecdotal information documented additional non-consumptive purposes for hunting carnivores. Our meta-analysis suggests, nevertheless, that carnivores are hunted pervasively across forested Africa. Research on hunting therefore needs to consider carnivores in more detail. In addition, as much hunting is for personal uses (consumption, culture or pest control), information needs to be gathered on carnivores not intended for sale.
The impacts of hunting on carnivores affect not only the assemblage of species to which these animals belong but also have implications for the people causing these impacts. If local hunters continue to use carnivores for subsistence or revenue then it is imperative that we understand the sustainability of such actions, to prevent a collapse of food supply should this resource become exhausted.
The effects of hunting on the carnivores of forested Africa appear to be escalating and remain poorly understood. The notion that carnivores are hunted solely for pest control is untenable, the taboos preventing their use as food are deteriorating, and there is a lack of conservation attention to their increased use for both personal consumption and sale for consumption.
We thank Marcella Kelly, Butch Brodie and Hank Shugart for their enthusiasm and feedback, Zach Farris for introducing HLD to Malagasy life, and Kat Miles for exploring it with her.
Hunter Doughty is interested in the behavioural ecology and conservation of large mammals, and has field experience in Namibia, Madagascar and Thailand. This article is based on her MA thesis at the University of Virginia. Sarah Karpanty is a vertebrate ecologist and conservation biologist who works on avian and mammalian communities in Madagascar and along the coasts of North America. Henry Wilbur is a population and community ecologist who works in the southern Appalachian Mountains of North America.