Diversity and conservation status of the mammalian species sold on wildlife markets in southern Benin
Human activities related to bushmeat hunting and wildlife markets are significant factors of depletion of the mammalian fauna in tropical Africa (Fa et al., 2015), and wildlife markets from the Dahomey Gap are no exception (D'Cruze et al., 2020). Our integrative approach allowed identifying a total of 94 species-level mammalian taxa sold on the wildlife markets in southern Benin, which is slightly higher than previously reported (87 species; Djagoun et al., 2013). Such number is c. twice greater than what was ever found in large-scale bushmeat trade surveys conducted in the biodiversity-rich countries from western and central African forests (e.g., Avila Martin et al., 2020; Fa et al., 2014; Mbete et al., 2011). Our results are especially striking as Benin is situated in the Dahomey Gap, a supposedly diversity-poor savannah-forest mosaic zone relative to the two rain forest blocks that it separates (Booth, 1958). However, such a high score may be explained by the fact that we surveyed two different types of markets (TMM and BM) having different purposes, functioning and ranges of influence. BM by itself gathered 25 species, which is in line with what was found in a recent bushmeat survey from the Dahomey Gap (Sonhaye-Ouyé et al., 2022). Carnivora, Rodentia and Cetartiodactyla were the dominant mammalian orders in both TMM and BM. Across western and central Africa, antelopes and wild pigs (Cetartiodactyla) together with rodents (Rodentia) were also found to be dominant contributors of the species diversity sold on the markets. However, primates, not carnivorans, are usually within the top three contributing orders (Fa et al., 2005; Gonedelé Bi et al., 2017; Juste et al., 1995), whereas in southern Benin Primates were fourth contributors and carnivorans first.
TMM constituted the main collector of mammalian species diversity (94) and orders (11), three of which (bats, elephant and aardvark) were not sold on BM. Given their wide range of recruitment across taxonomic orders, bioclimatic zones and countries (Djagoun et al., 2013), their dense network (> 40 markets counted within the study area; SZ, pers. obs.) and the large demand for traditional medicine and religious practices based on wildlife (D'Cruze et al., 2020), TMM likely are a major threat to the conservation of mammals in Benin and the subregion. Carnivorans were unusually dominant, with 27 species representing the seven families (Nandiniidae, Herpestidae, Viverridae, Felidae, Hyaenidae, Canidae, Mustelidae) occurring in the subregion. Carnivorans have been reported as prevalent in medicinal and spiritual practices across tropical Africa (Doughty et al., 2015), which is extensively confirmed by our study. Small carnivorans constitute a valuable income for hunters in southern Benin, where almost half of the animals are sold to TMM for incomes ranging from US$ 2.5 to 33.7 (Djagoun and Gaubert, 2009). Large carnivores such as big cats and wild dogs, almost entirely extirpated from the country, may also be persistently imported to feed the demand for their use in traditional medicine practices (see Williams et al., 2017), the trade here cumulating cultural purposes with large benefits related to selling rare species (Djagoun et al., 2013).
TMM sold the greatest number of species of high conservation concern (19 NT to CR species [IUCN] and 51 NT to EX species [RLB]) or regulated by international conventions (22 species listed on appendices I and II of CITES). Some marking elements included the western gorilla (Gorilla gorilla), the African savanna elephant (Loxodonta africana), the cheetah (Acinonyx jubatus), the lion (Panthera leo), the leopard (Panthera pardus), the African wild dog (Lycaon pictus) and the hippopotamus (Hippopotamus amphibious), all implying sourcing from northern Benin and / or foreign countries. A series of pangolins and primates were also of great conservation concern, including the white-bellied pangolin (Phataginus tricuspis; also found on BM), the white-thighed colobus (Colobus vellerosus), the red-bellied monkey (Cercopithecus erythrogaster) and the red-capped mangabey (Cercocebus torquatus). The latter being considered extinct (RLB) or potentially extinct (IUCN) in Benin, its presence in TMM implies either transnational trade with the neighbouring Nigeria (see Maisels et al., 2019) or the presence of an undetected population in Benin at a high risk of extinction. Our analysis also showed that a series of 19 species had their national conservation status not evaluated (RLB), mostly consisting of small mammals (rodents, bats, hare, shrew) found on both types of markets. In the next future, it will be of prime importance to gather information on their population trends, especially relative to their trade in Benin.
A few species sold on TMM were considered uncertain or not present in Benin both by IUCN and RLB, namely the southern reedbuck (Redunca arundinum), the big-eared swamp rat (Malacomys longipes), and the Jackson's fat mouse (Steatomys jacksoni). Because the identification of the southern reedbuck was confirmed by DNA-typing, our results might be evidence for long-range wildlife trade from southern Africa (see IUCN SSC Antelope Specialist Group, 2016) to Benin, exemplifying again the wide range of action of TMM. The two species of rodents may prove difficult to identify from morphological grounds (Happold, 2013), so molecular identification will be needed to definitely confirm their taxonomic attribution. However, the species’ ranges roam at the vicinity of southern Benin, and their –yet unreported– occurrence in the country remains conceivable (see Cassola, 2016; Gerrie and Kennerley, 2016).
Overall, BM appeared as a subsample (c. 27% of the species) of the taxonomic diversity found on TMM. This emphasizes the narrow range of recruitment of the BM in southern Benin, predominantly sourcing species from the Guinean (southern) forest zone or ubiquitous species. As a matter of fact, BM are mostly supplied by hunters operating in and at the vicinity of LF (Sogbohossou and Kassa, 2016). BM recruited a local community of medium- to small-sized mammals equally dominated by antelopes and wild pigs, rodents and carnivores, representing 72% of the total species diversity. The mammals sold on BM were generally of lower conservation concern according to international status (IUCN and CITES), with the exception of the white-bellied pangolin, the wild cat (Felis silvestris), the mona monkey (Cercopithecus mona) and the patas monkey (Erythrocebus patas). According to national status (RLB), four additional species were of high conservation concern, including the red river hog (Potamochoerus porcus), the sitatunga (Tragelaphus spekii), the African palm civet (Nandinia binotata) and the Benin tree hyrax (Dendrohyrax interfluvialis; evaluated under D. dorsalis). Although BM recruited less species diversity than TMM, our results suggest that BM can also be considered as a likely threat to the conservation of mammalian diversity in southern Benin.
Four taxa from both markets could not be genetically identified to the species level, including an antelope (Kobus), a scaly-tailed squirrel (Anomalurus), an African striped squirrel (Funisciurus) and a sun squirrel (Heliosciurus). Because cryptic diversity is likely at stake in several groups of small to medium-sized mammals in tropical Africa and notably in the Dahomey Gap (e.g., Colyn et al., 2010; Gaubert et al., 2016; Oates et al., 2022), further investigations will have to be conducted to decipher whether those represent new taxa or yet unrepresented species in public databases.
Domestic species can also be part of the species spectrum sold on wildlife markets in the tropics (Karesh et al., 2005). In southern Benin, domestic cats and European rabbits were found both on TMM and BM, whereas domestic dogs were only sold –frequently– in TMM. DNA-typing proved useful in resolving the issue of the cranial artefacts sold as “leopard” or “wild dog”, returning domestic dog as the disguised species (skull and skin) from which those artefacts were created. Similar cases of domestic species sold as wild species to gain a higher benefice have been reported in western African BM (Olayemi et al., 2011), but not reaching that level of sophistication as in TMM from Benin. Here, the jaws of a domestic dog were maintained open with a wooden stick to make it look “ferocious”, and a piece of dog skin spotted with black markings was stuck around the skull so to imitate the leopard head. A similar strategy has also been observed to produce fake serval skins from dogs (PG and CD, pers. obs.). Artefactual wild dogs’ heads seem to be created from a transformative protocol where the dog’s head is placed in an unknown mixture that would result in inflating the tissues and remaining organs (e.g., eyes). In the neighbouring Togo, dogs may be killed as part of sacrificial rituals (Verdier, 1981) and can be found on TMM stalls (https://togo-tourisme.com/culture/marches/le-marche-aux-fetiches). In southern Benin, dogs seem to occupy a double function where they are both used in traditional medicine for specific purposes (as in Nigeria; Dongnaan Gurumyen et al., 2020) and as “substitutes” to some rare species on the market such as the leopard and African wild dog.
Functional diversity of the mammalian community sampled by wildlife markets in southern Benin
In the tropics, habitat loss and hunting activities are the main drivers of the loss of functional diversity (FD) (Matuoka et al., 2020), challenging equilibrium among ecological functions and thus the durability of ecosystem services (Brodie et al., 2021; Flynn et al., 2009). In central Africa, hunting is predicted to exacerbate the degradation of ecological functions such as seed dispersal and forest regeneration quicker than the expected effects of climate change (Abernethy et al., 2013). As a matter of fact, a few case studies conducted in western and central Africa have shown the negative impact of hunting on the functional diversity of mammals (Tagg et al., 2020; Vanthomme et al., 2010). However, to our knowledge this is the first time that functional diversity is directly assessed through the prism of the African wildlife trade.
Wildlife markets in southern Benin sampled across a large dietary spectrum, englobing the whole trophic space found in western African mammals (see Fa and Purvis, 1997). A total of nine diet specializations and five main ecological functions were gathered by TMM, with herbivores being slightly dominant over predators, whereas scavengers represented a single species (Crocuta crocuta). Herbivores are likely the functional guild the most hunted in the tropics, which has important consequences for plant communities and nutrient cycling (Brodie et al., 2021). The unexpected, balanced representation between herbivores and predators sold in TMM is not in line with such general trend. Here, TMM from southern Benin appear as a wildlife trade network potentially impacting the full functional diversity of the sourced ecosystems, jeopardizing at the same time the regulatory, structural and production roles that the traded species represent, involving both herbivores and predators (Schmitz, 2009).
TMM collected a majority of frugivorous species (31%; mostly bats, primates, antelopes and rodents). The consequences related to the over-hunting of medium to large frugivorous mammals on forest regeneration has been extensively studied in the tropics (Abernethy et al., 2013), including western Africa (Effiom et al., 2013). The depletion of frugivores has negative effects on seedling diversity, seed dispersal and, as a downstream consequence, forest carbon storage (Brodie et al., 2021; Kurten et al., 2015; Vanthomme et al., 2010). Across the tropics, forests affected by high levels of long-term hunting have seen the establishment of fast-growing pioneer species and lianas together with the decline of large hardwood trees, and the likely alteration of the dynamics of ecosystem processes (Abernethy et al., 2013; Lamperty et al., 2020).
Folivorous species (26%; mostly antelopes and primates) were the second most diverse trophic group sold on TMM stalls. Comparatively to frugivores, the ecological function of mammalian folivores has been poorly assessed, although they might have an important impact on forest composition by acting on death and recruitment of the consumed trees, and enhancing nutrient cycling through fertilization (Chapman et al., 2013). Folivores also act on the vegetation biomass, and their depletion might imply a cascading effect where stem densities increase and the recruitment of large trees reduces to result in homogeneous forest structure and lower carbon stocks (Poulsen et al., 2018).
Predators –invertivores and carnivores– were the third trophic group sold on TMM (23%; mostly carnivorans). Invertivores play an important ecological function as they regulate pest abundance and may suppress arthropod outbreaks (Maas et al., 2016), thus limiting the propagation of disease vectors (Ghanem and Voigt, 2012). Mammalian carnivores have an important role in regulating ecosystems, acting as biocontrol agents against –for instance– the pullulating of rodents (Williams et al., 2018), which are both disease vectors and seed predators. Carnivores also have a cascading effect on the trophic chain and thus ecosystem equilibrium and productivity, and may enhance carbon storage by limiting the number of herbivore preys (Ripple et al., 2014).
As a token of the broad impact that TMM may have on functional diversity and ecosystem services, five additional trophic groups, representing 20% of the species sold, were present on the stalls, including omnivores (carnivorans and rodents), granivores (rodents), lignivores (mostly rodents), piscivores (otters) and nectarivores (bats). Omnivores and granivores are considered ecosystem engineers positively acting on soil ecosystems but at the same time as seed predators likely influencing vegetation assemblages through targeted seed predation (Asquith et al., 1997; Mills et al., 2018). Lignivorous species act on nutrient cycling and ecosystem fertilization together with piscivores, the latter also participating to prey regulation and, as a top-down effect, shaping prey communities and aquatic habitats (Peterson and Schulte, 2016). Eventually, nectarivorous mammals, by pollinating flowers of many tropical and subtropical plant species, promote forest regeneration and support timber production (Ghanem and Voigt, 2012).
In accordance with the taxonomic nestedness that we observed between the two markets, BM was also found to be a sub-representation of TMM functional trait space (c. 28%). This indicates that the range of trait variation found in BM is lower than that observed in TMM. However, it is worth noting that the FRic index is strongly influenced by species with extreme trait values and consequently does not provide a full picture of trait variation within each functional trait space (Legras et al., 2018). By considering the FDis index, which is less affected by extreme values (Laliberté & Legendre 2010), we showed that the two markets displayed similar FD values. In other words, although BM contain species with less extreme trait values than TMM (see Fig. 4), BM is also representative of a large spectrum of functional traits. For instance, seven out of nine trophic groups were found in BM, including folivores (33% of the species), carnivores (21%), frugivores (16%), granivores (12%), invertivores (9%), omnivores (7%) and lignivores (2%). Such patterns remain broad in terms of trophic niche assemblage. Folivores and predators (carnivores and invertivores; 30%) were the dominant groups, contrary to TMM where frugivores prevailed. Such discrepancy may be partly explained by the protection status of some frugivorous primates from the LF, such as the red-bellied monkey Cercopithecus erythrogaster and the white-thighed colobus Colobus vellerosus (Djègo-Djossou and Sinsin, 2009; but see Nobime et al., 2009). However, because BM are sourcing animals locally, the relative lack of frugivores may be better explained by local extinctions within such functional group through centuries of habitat fragmentation and hunting in southern Benin (Amadji and Roesch, 1990). BM acting on an already depauperate trophic spectrum representation, we posit that they also represent a significant threat to ecosystem functioning in southern Benin.
The spectrum of ecological functions impacted by TMM was wide, including seed dispersion (37% of the species; mostly frugivores, folivores and omnivores), prey regulation (36%; mostly carnivores and invertivores), browsing (22%; mostly folivores), grazing (3%; omnivores) and fertilization (2%; nectarivores). The functional community found in BM exhibited similar patterns, being also dominated by seed dispersers (35%), prey regulators (30%) and browsers (26%), and followed by grazers (9%), whereas fertilizers were not represented. Beyond the largest functional spectrum sampled by TMM, the fact that both types of wildlife markets sampled across the main ecological functions at stake in western African terrestrial ecosystems is a concern for the ecological health of the sourced habitats. As said above, the depletion of seed dispersers has an impact on tree species composition and downstream consequences on nutrient cycling, carbon storage capacity and browsing (Abernethy et al., 2013; Poulsen et al., 2018), all the more since several species were both grazers and seed dispersers (e.g., rodents, antelopes and elephant). The depletion of prey regulators is obviously linked to agricultural and health risks (pest pullulating, zoonotic spillover; Sinclair, 2003), but also involves cascading effects on the ecosystems such as density-compensation by grazers and seed predators, likely affecting patterns of forest regeneration (Effiom et al., 2013; Scabin and Peres, 2021). The interplay between browsing and grazing also plays a major functional role as involved in the dynamics of nutrient cycles and habitat structure (tree-grass equilibrium), ecosystem productivity and resilience (Milchunas and Lauenroth, 1993; Staver et al., 2021; Terborgh et al., 2016). Eventually, fertilization via pollination by bats is key for fruit production and forest ecology (Ramirez-Francel et al., 2022).
Conservation implications for mammalian species and ecosystems in southern Benin
Thanks to our integrative approach crossing market interviews with direct observations, morphological-based identification and DNA-typing, we were able to provide an unprecedented list of 94 species-level mammalian taxa sold on wildlife markets in southern Benin. Such an exhaustive taxonomic list will serve as a basis for future refinement, market surveys and re-assessment of conservation status across the Dahomey Gap region, where similar market networks acting on similar mammalian communities are at stake. Given its high level of resolution, DNA-typing will have to be routinely applied in the future wildlife trade surveys for species identification (Dipita et al., 2022; Gaubert et al., 2015; Gossé et al., 2022) and eventually the geographic tracing of sourced habitats (e.g., Wasser et al., 2004). Several taxa could not be given a species name, and we suspect additional levels of cryptic diversity in the Dahomey Gap mammalian fauna that will deserve further scrutiny.
Overall, wildlife markets from southern Benin likely constitute a threat to both biodiversity conservation and the resilience of forest and savannah ecosystems where they source from. TMM heavily recruited across the full spectrum of (i) diet specializations and ecological functions and (ii) functional traits including body weight, generation length and litter size, available in western African mammalian communities. TMM notably sampled the most endangered species with the greatest body weights and generation lengths, and smallest litter sizes, including the African savanna elephant, the hippopotamus, the buffalo (Syncerus caffer), the roan antelope (Hippotragus equinus), the lion, the western gorilla and the topi (Damaliscus lunatus).
Browsers and prey regulators found on TMM, including folivores, carnivores and omnivores, were especially heavier than in BM. Because the depletion of large and medium-sized herbivores and keystone predators have serious downstream consequences on ecosystem (see above), our results identify TMM as a main threat for ecosystem conservation in Benin. However, it remains challenging to accurately estimate the ecological impact of TMM as the market network is able to source into various geographic regions and habitats, sometimes in foreign countries. Moreover, the volumes of traded mammals remain uncertain as carcasses sold on TMM –skins, bones, skulls, appendages, organs– can stay for weeks, months and even years on the stalls (SZ, pers. obs.), contrary to BM where animals are generally sold within a few hours.
BM in southern Benin are locally fed –mostly by LF– and as such, are a representation of the depleted mammalian fauna from the area, where the large- and most of the medium-sized mammals have been extirpated by agriculture and hunting. Our findings are of concern for the integrity of LF, which is the last large patch of semi-deciduous tropical forest preserved in southern Benin (the “Noyau central”), because the mammalian community sold on BM represented seven diet specializations and four main ecological functions. The sampling of such a large functional diversity ultimately jeopardizes the ecological integrity of the forested habitat, and is the signature of non-selective hunting in the area. Given that BM have restricted, identified sources, the system can be more easily characterized than TMM, notably regarding the volumes (numbers and biomass) of the trade. However, given the lack of available data, it is urgent to undertake long-term surveys of BM to quantify the threat they constitute for mammalian biodiversity and the sustainability of ecosystem services in and around LF.
Our study had the merit to highlight for the first time the potential impact of the wildlife trade on the FD of mammalian communities in western Africa. Together with taxonomic diversity, FD is a marker of evolutionary heritage and ecosystem productivity (Ahumada et al., 2011; Ernst et al., 2006; Oliveira et al., 2016) that needs to be considered when tackling the issue of the wildlife trade. With the taxonomic and functional database that we provide as a starting point, it will be important to pursue the effort to quantify functional diversity in African mammals using continuous traits, given the existing knowledge gaps and the promise of such traits to better apprehend FD (Kohli and Rowe, 2019). Such dedication will ultimately help investigate the appropriate scales to which functional diversity metrics may reflect changes in ecosystem functioning in western Africa (see Hatfield et al., 2018). This is particularly important to set up management strategies able to maintain ecosystem integrity, biodiversity and livelihoods in the subregion, anticipate the resilience of such ecosystems to global change, and provide public authorities with scientific evidence for the impact of the wildlife trade on ecosystem services.