Introduction
Crocodilians are charismatic megafauna that act as keystone species and often serve as indicators in ecosystem monitoring and restoration programmes (Ross, Reference Ross1998; Mazzotti et al., Reference Mazzotti, Brandt, Moler and Cherkiss2007). They are widely viewed as flagship conservation species and have the potential to galvanize efforts in wetlands conservation. Three species are native to West Africa: the Nile crocodile Crocodylus niloticus, the slender-snouted crocodile Mecistops cataphractus, and the African dwarf crocodile Osteolaemus tetraspis. A rich cultural history has included both worship and overexploitation (Moiser & Barber, Reference Moiser and Barber1994; Kpera, Reference Kpera2003; Toonen, Reference Toonen2003) but there is an almost complete deficit of knowledge regarding ecology and population status of crocodiles across the region (Ross, Reference Ross1998).
Although the Nile crocodile is one of the most widely studied crocodilians it has been little studied in West and Central Africa (Pooley, Reference Pooley1980; Ross, Reference Ross1998). Most of the limited information available consists only of presence/absence information (Wilson, Reference Wilson1978; Pooley, Reference Pooley1980; Kofron, Reference Kofron1992; Behra, Reference Behra1993, Reference Behra1994; Moiser & Barber, Reference Moiser and Barber1994; Dore, Reference Dore1996; Even, Reference Even1996; Akani et al., Reference Akani, Luiselli, Angelici and Politano1998; de Smet, Reference de Smet1999; Luiselli et al., Reference Luiselli, Politano and Akani2000; Shine et al., Reference Shine, Bohme, Nickel, Thies and Wilms2001; Kpera, Reference Kpera2003). Recent debate on the taxonomic status of this crocodile in West and Central Africa has highlighted the importance of understanding its population status (Schmitz et al., Reference Schmitz, Mansfeld, Hekkala, Shine, Nickel, Amato and Böhme2003; Hekkala, Reference Hekkala2004). It is categorized as Lower Risk/least concern on the IUCN Red List (IUCN, 2008) but all West African populations are in CITES Appendix I (UNEP-WCMC, 2008). The IUCN Crocodile Specialist Group noted that surveys throughout this region are the highest priority for Nile crocodile conservation (Ross, Reference Ross1998).
The slender-snouted crocodile is regarded by the Crocodile Specialist Group as the least known crocodilian (Ross, Reference Ross1998). Until recently it was considered to be a true crocodile, in the genus Crocodylus, but has now been placed in the monotypic genus Mecistops (McAliley et al., Reference McAliley, Willis, Ray, White, Brochu and Densmore2006). In certain parts of its range (Gabon and Republic of Congo) it may still be relatively common (Thorbjarnarson & Eaton, Reference Thorbjarnarson and Eaton2004), although the most recent surveys in West Africa (Waitkuwait Reference Waitkuwait1985a,Reference Waitkuwaitb, Reference Waitkuwait, Hall and Bryant1989) suggested that populations are seriously declining. The slender-snouted crocodile is categorized as Data Deficient on the Red List (IUCN, 2008) but the Crocodile Specialist Group has expressed concern that it may meet the criteria for Endangered (Ross, Reference Ross1998), and all populations are on CITES Appendix I (UNEP-WCMC, 2008). The highest priority action for this species is surveys of population status throughout its range to determine its Red List status and establish conservation programmes (Ross, Reference Ross1998).
The African dwarf crocodile is a poorly known species ranging throughout the lowland forest zone of West and Central Africa and inhabits forest waterways, swamp forests and isolated savannah pools (Kofron & Steiner, Reference Kofron and Steiner1994; Ross, Reference Ross1998; Riley & Huchzermeyer, Reference Riley and Huchzermeyer1999). Like the Nile crocodile, its taxonomic status is uncertain, with a second taxon, Osteoblepharon osbornii, described for animals from the western Congo basin (Thorbjarnarson & Eaton, Reference Thorbjarnarson and Eaton2004). The dwarf crocodile is cateogrized as Vulnerable on the Red List (IUCN, 2008) and all populations are on CITES Appendix I (UNEP-WCMC, 2008). Range-wide status surveys and evaluation of the threat from the bushmeat trade are priority actions for this species (Ross, Reference Ross1998).
The objective of this study was to survey the distribution, population status and threats facing the slender-snouted and Nile crocodiles in Ghana and Côte-d'Ivoire, and to collect data opportunistically on the African dwarf crocodile.
Study area
From 19 July to 27 November 2006 we surveyed the status, distribution and threats faced by crocodiles in Ghana and Côte-d'Ivoire (Fig. 1). The southern portion of both countries is mostly Upper Guinea Forest, whereas the northern portion is largely wooded savannah. In both countries the major wet season is in March–June, a short dry season in August, and a second, reduced wet season from September to mid November. There is a unimodal rainfall regime in the northern savannah zone from May/June to November (Grubb, Reference Grubb1998). The average day and night temperatures during the study were 26°C and 23°C, respectively. Our surveys were conducted during the latter half of the normal wet season, although drought conditions characterized 2006 and crocodiles were still limited to areas of permanent water.
Fig. 1 Locations surveyed for crocodiles in Ghana and Côte d'Ivoire (large dots). Small dots represent cities.
Surveys were planned systematically according to information from published records, anecdotal reports, and presumed suitable crocodile habitats. Significant effort was made to sample broadly within each country to obtain information over a wide range of habitat types, land uses, and protection status. Political instability prohibited us from surveying the northern half of Côte-d'Ivoire (Fig. 1).
Methods
Initial surveys detected low crocodile densities and we therefore decided to spend less time at each site and sample more sites (Thompson, Reference Thompson2004). Surveys were conducted from available water craft and on foot. Diurnal surveys were utilized to detect basking crocodiles and plan nocturnal survey routes. Nocturnal spotlight surveys are a standard crocodilian survey method (Webb & Smith, Reference Webb, Smith, Webb, Manolis and Whitehead1987); crocodiles were detected with LED headlamps or a 100,000/200,000 candlepower flood/spotlight. Length of survey routes was calculated using a global positioning system. This survey method results in an index of relative encounter rate because on any given survey not all individuals of a population are detectable (Bayliss, Reference Bayliss, Webb, Manolis and Whitehead1987; Hutton & Woolhouse, Reference Hutton and Woolhouse1989; Thorbjarnarson et al., Reference Thorbjarnarson, Platt and Saw Tun Khaing2000).
Survey sites were characterized as river, flood plain (i.e. river or stream in wooded savannah habitat that flooded into the surrounding plain), natural lakes, artificial dams, coastal lagoons, wetlands (e.g. swamp forest), and urban or other artificial habitats. When spotted, crocodiles were approached to determine species and estimate total length in the following size classes: hatchling, < 1.0 m, 1.1 > 3.5 m in 0.25 m increments, and EO (eyes only) for unapproachable individuals. While it is impossible to assign EO individuals to a species with complete confidence location, behaviour, size and other sightings allow some to be provisionally assigned. Unless otherwise stated, all crocodile encounters (both counts and rates) are presented as the number of non-hatchling crocodiles.
Information on crocodile hunting was obtained from market and vendor/hunter surveys, principally at the Kumasi Central Market in Ghana, with additional information collected at the Tamale Market in Ghana, the Dimbokro Central Market in Côte-d'Ivoire, and villages during the surveys. Hunters, fishermen and village councils were informally interviewed regarding perceptions of present and historical crocodile abundance, relative value of, or preference for, crocodile products (consumption, traditional medicinal and religion), and human-crocodile conflict.
Results
Seventy-one crocodile surveys were conducted at 46 different sites covering 726.4 km of crocodile habitat (Table 1): 25 sites (n = 46 surveys, 333.7 km) in Ghana and 21 sites (n = 25 surveys, 402.2 km) in Côte-d'Ivoire. Over 57% of surveys (n = 41, 440.3 km) were in riverine habitat yet only 36.5% (n = 100) of crocodile sightings were in rivers. In contrast, 11% of surveys (n = 8, 137.3 km) were coastal lagoons, yet 31% (n = 85) of crocodiles were sighted there (Table 1). No crocodiles were observed in 31.5% of surveys, and < 5 crocodiles were observed in 71% of surveys (median encounter rate = 0.26 km-1).
Table 1 Survey details by habitat type. Survey distance is the combined length of surveys, with average survey length and percent of total distance for each habitat type. Crocodile sightings are presented as the number of non-hatchlings, with average number of non-hatchlings per survey and the percentage of sightings of non-hatchlings per habitat, and total including hatchlings. Encounter rate is the average for non-zero encounter surveys, with average for all surveys per habitat. Zero encounter is the number of surveys where only hatchlings were encountered, with number of surveys with no encounters and percentage of no encounter surveys of the total surveys per habitat.
We encountered 454 individual crocodilians (Tables 2 & 3) that ranged in size from hatchlings to < 3.5 m in length (Tables 2 & 3). Over 84% of crocodiles whose sizes were determined were ≤ 1.5 m (i.e. not sexually mature). The combined, non-species specific mean encounter rate for all surveys was 0.90 km-1 (1.184 and 0.433 km-1 in Ghana and Côte-d'Ivoire, respectively). However, for sites where only non-hatchling crocodiles were detected the mean encounter rate was 1.41 km-1 (1.821 and 0.712 km-1 in Ghana and Côte-d'Ivoire, respectively).
Table 2 Survey results for confirmed C. niloticus sightings and unidentified sightings presumed to be C. niloticus in Ghana and Côte-d'Ivoire. Encounters are in estimated size class by site. Total effective represents the number of non-hatchlings seen in each survey. Average encounter rates include only surveys that detected crocodiles.
1 Htch, hatchlings
2 Unidentified sightings presumed to be C. niloticus
3 Surveys removed from the total average encounter rates due to their non-natural site status.
C. niloticus was the most frequently encountered species, with 301 confirmed individual sightings and 124 of the assigned EO sightings (Table 2). Individuals ranged in size from hatchlings to < 3.5 m. However, only 8.8% (n = 37) of individuals fell in potentially sexually reproductive size classes (> 1.5 m) and most were detected at just a few sites. Where detected, C. niloticus was encountered on average at a rate of 1.05 km-1 (1.193 and 0.882 km-1 in Ghana and Côte-d'Ivoire, respectively). C. niloticus was found almost exclusively in northern, savannah woodland rivers and dams or in coastal lagoons and rivers.
M. cataphractus was rarely encountered, with 10 sightings and four presumed EO sightings (Table 3). Individuals ranged in size from hatchling to < 2 m, and only two confirmed individuals were in a potentially reproductive size class (> 1.5 m). M. cataphractus was encountered at 0.029 km-1 over all non-zero encounter surveys, 0.468 km-1 where both C. niloticus and M. cataphractus were detected sympatrically, and 3.89 km-1 where only M. cataphractus was found. M. cataphractus was found primarily in forested rivers in the southern forest belt of each country, although its presence in northern rivers with gallery forest indicates broader habitat use than formerly presumed.
Table 3 Survey results for confirmed M. cataphractus and O. tetraspis sightings and unidentified sightings presumed to be one of these species. Encounters are in estimated size class by site. Total effective is the number of non-hatchlings seen in each survey. Average encounter rates include only surveys that detected crocodiles.
1 Htch, hatchlings
2 Unidentified sightings presumed to be either M. cataphractus or O. tetraspis
3 Percentage known sightings for each species from the total number of crocodiles detected
4 Average encounter rate for each species over all surveys (including zero-encounter surveys)
5 Average encounter rate for each species over all non-zero surveys (i.e. where any of the three crocodiles were detected)
O. tetraspis was also rarely encountered, with 11 sightings and two EO sightings (Table 3). Individuals ranged in size from hatchling to 1.75 m, and 69.3% of encountered individuals were in a potentially reproductive size class (> 1.25 m). Where detected, O. tetraspis was encountered at 0.311 km-1. These survey results are not representative of the true abundance of O. tetraspis as the survey strategy was directed at finding the other two species and O. tetraspis was mainly observed during incidental excursions into tributaries and forest habitat unsuitable for the other species. Anecdotal evidence suggests that O. tetraspis remains abundant throughout both countries.
Interviews with market vendors suggested that crocodiles are rarely sold in major bushmeat markets. Market patrons and community members offered biological, cultural, and social reasons and indicated that crocodiles are rarely hunted and even more rarely sold for consumption when captured. Four crocodile hunters were encountered, although all claimed to have not hunted crocodiles in at least 10 years (in one case > 20 years) because of crocodile scarcity. Only three instances of attacks on humans were encountered, only one of which was fatal, and there were only two records within the past 15 years. Human-crocodile conflict is limited to rare predation on domestic livestock.
Discussion
The result of most immediate concern is the few encounters with M. cataphractus. Both Ghana and Côte-d'Ivoire are within the reported species range and historically had dense patches of tropical forest and well-forested river margins (Ross, Reference Ross1998). However, because of rapid loss and fragmentation of the Upper Guinea Forests (Hawthorne, Reference Hawthorne, Swaine, Alexander and Watling1996; Fairhead & Leach, Reference Fairhead and Leach1998; Oates, Reference Oates1999; Myers et al., Reference Myers, Mittermeier, Mittermeier, da Fonseca and Kent2000) it is now likely that these countries represent the western edge of the modern range of M. cataphractus. The reported range extends west to Senegal and, whilst this may still be the case, populations are probably heavily depleted west of Côte-d'Ivoire and we suggest that the core range for this species is contracting east into the Congo Basin. It is therefore likely that the observed low densities of M. cataphractus and seemingly abnormal habitat use (i.e. savannah woodland) are range limit effects. Additionally, it is possible that the more abundant C. niloticus limits the distribution and abundance of M. cataphractus, a phenomenon unlikely to occur in Central Africa because of the low abundance of C. niloticus there (M. Eaton, pers. comm.).
The spotlight surveys found that crocodiles were usually present at low densites. Encounter rates from this study are among the lowest reported for crocodiles throughout Africa, and of particular note is the high proportion of zero-encounter surveys. Compared to surveys in East and southern Africa, where some encounter rates may be low, few surveys detect no crocodiles (Table 4). While it is impossible to quantify current population trends because of a lack of historical data, information from the interviews suggests that populations are in decline. Local reports indicated that as recently as the 1980s crocodiles could be encountered frequently, and at most sites populations were robust enough to be hunted. The apparently depleted crocodile populations in Ghana and Côte-d'Ivoire are in contrast to reports from East and southern Africa where populations have been evaluated for their harvest value and in many areas have recovered significantly (NPWS, 1993; Changwe, Reference Changwe1996, Reference Changwe2004; Jachmann, Reference Jachmann1996; Leslie, Reference Leslie1997; Nyirenda, Reference Nyirenda1997, Reference Nyirenda2004; Games & Severre, Reference Games and Severre1999; Brown et al., Reference Brown, Stander, Meyer-Rust and Mayes2004; Kampamba, Reference Kampamba2004; Simwanza, Reference Simwanza2004; Zimbabwe Parks and Wildlife Management Authority, 2006).
Table 4 Encounter rates (individuals km-1) for C. niloticus from across its eastern and southern African distribution for comparison with the results of this study. Encounter rates are comparable because standardized survey methods were used.
Of the major threats facing crocodilians in West Africa habitat loss and encroachment probably have the greatest effects. Habitat loss is probably having the greatest impact on the forest-river dwelling M. cataphractus, with loss of the Upper Guinea Forests limiting available habitat. Habitat encroachment, where human settlements invade potential crocodile habitat, affects all three species by limiting available nesting and basking habitat (only a single crocodile was seen basking during the day).
The second major anthropogenic threat facing crocodilians is overfishing. Fishers in this region report small catches, in weight and volume (Brashares et al., Reference Brashares, Arcese, Sam, Coppolillo, Sinclair and Balmford2004; M.H. Shirley, pers. obs.), suggesting a heavily depleted fishery that will affect crocodiles by depleting prey resources and increasing the rate of incidental mortality through drowning in fishing nets. Finally, while hunting does not seem to be a current phenomenon (see below), commercial hunting was at one time common and the depauperate crocodile populations today could, at least partially, be a result of this historical trade that, in combination with the other current threats, limits the ability of populations to rebound despite evidence of breeding from the presence of hatchlings and juveniles.
Throughout Central Africa (Gabon, Democratic Republic of Congo, Republic of Congo and Equatorial Guinea), crocodiles are currently hunted for the bushmeat trade. In Ghana and Côte-d'Ivoire, crocodile hunting appeared to be a limited phenomenon until the 1980s and early 1990s in certain localities, although primarily for skins, traditional medicine and religious uses. Crocodile harvest is now rare because of a combination of scarcity of crocodiles, collapse of the local market for crocodile products, traditional beliefs, and stricter international trade regulations (e.g. CITES).
Our results show that crocodiles are rarely encountered and when encountered the predominant size class is hatchling and juveniles, making individuals too small to be worth hunting on a regular basis. Although we do not expect equal detectability amongst all size classes, our inability to detect adult and subadult crocodiles is probably also a problem for bushmeat hunters. This scarcity of crocodiles has also led to a decline in the supply for crocodile products, which has led to a large-scale market reduction and social de-emphasis on crocodile hunting. Interviews with people from throughout the bushmeat supply chain and with bushmeat consumers confirmed both the lack of supply and demand for crocodile products (M.H. Shirley, unpubl. data). This is corroborated by other bushmeat studies that only found insignificant and irregular records of crocodiles in major markets (Ghana Wildlife Division, 1998a,b; Holbech undated, Reference Holbech2001; Conservation International, 2002; Mendelson et al., Reference Mendelson, Cowlishaw and Rowcliffe2003; Brashares et al., Reference Brashares, Arcese, Sam, Coppolillo, Sinclair and Balmford2004; Bassett, Reference Bassett2005; Cowlishaw et al., Reference Cowlishaw, Mendelson and Rowcliffe2005a,Reference Cowlishaw, Mendelson and Rowcliffeb; Crookes et al., Reference Crookes, Ankudey and Milner-Gulland2005; Damania et al., Reference Damania, Milner-Gulland and Crookes2005; Rowcliffe et al., Reference Rowcliffe, Milner-Gulland and Cowlishaw2005; E.J. Milner-Gulland, pers. comm.; Ghana Wildlife Society, undated.).
The cultural history associated with crocodiles in this region is complex, and most ethnic groups have at least some religious or historical attachment to them. Many people believe the persistence of water is tied directly to crocodiles, and if the crocodiles vanish the rivers will run dry. Crocodiles were worshipped as totems and used as guardians of community grain stores. This is exemplified in the village of Paga in Ghana, which has a well known community-based tourism programme centred on the cultural relationship with crocodiles. Additionally, crocodile bile is widely believed to be a potent toxin and because of this many people are wary of buying crocodile meat. However, we did encounter a small number of crocodiles for sale in the form of skins, skulls, and teeth in traditional religion and medicine market stalls, and three households with live pet crocodilians.
We recommend continuing research programmes to examine the distribution, population status, ecology and conservation status of crocodilians in the region. Priorities are to identify key localities for long-term monitoring and sites for future ex situ conservation actions (e.g. repopulation from captive stock). Surveys in Ghana and southern Côte-d'Ivoire need to be repeated in the dry season to determine the impact of environmental conditions on survey results, and new surveys need to be carried out in protected areas throughout Côte-d'Ivoire (Parc Nacional du Tai, Parc Nacional du Comoe and Parc Nacional d'Azagny), as well as in Nigeria, Liberia, Sierra Leone, Guinea-Bissau, Guinea, The Gambia and Senegal. Emphasis should be placed on discovering populations of M. cataphractus and understanding the status of O. tetraspis throughout the region.
Our results suggest that the IUCN Red List category for M. cataphractus should be re-evaluated. Under the IUCN criteria, a species qualifies for Data Deficient ‘where data are so uncertain that any category is plausible’ (IUCN, 2001, 2007). Surveys in Liberia (Kofron, Reference Kofron1992), Nigeria (Dore, Reference Dore1996; Akani et al., Reference Akani, Luiselli, Angelici and Politano1998; Luiselli et al., Reference Luiselli, Politano and Akani2000), Senegal and Gambia (M.H. Shirley, unpubl. data), and our stud-ies in Ghana and Côte d'Ivoire, countries that hold the last remaining blocks of Upper Guinea Forest, suggest that M. cataphractus in West Africa qualifies for Endangered under criteria A1abcd+2+3+4 C1+2a(i) and may even qualify for Critically Endangered under criteria A1abcd+2+3+4 (IUCN, 2001). We propose that a formal regional assessment is conducted because populations in West Africa are isolated both biogeographically and behaviourally from breeding populations in other parts of the continent (IUCN, 2003). Additionally, we recommend that future survey efforts focus intensively on populations in Central Africa to facilitate a global assessment of this species.
Strategies for crocodile conservation can include sustainable use and many such programmes for harvest and commercial trade exist in East and southern Africa. We recommend, based on our information on current interest and demand, that any use be focused on the local traditional medicine and religious users. Additionally, we recommend research to understand the traditional medicine and religious market dynamics as a platform for establishing sustainable use programmes.
The successful conservation of crocodiles in Ghana, Côte-d'Ivoire, and West Africa as a whole, will require raised awareness amongst conservation and enforcement personnel. The Crocodile Specialist Group convened the 1st Working Meeting for West African Crocodiles in Parc W, Niger, over November 12–15 2007, with the emphasis on generating a base for future efforts by involving local people throughout West Africa. These activities should be continued, with an emphasis on local involvement and capacity-building efforts, without which any conservation initiatives will not be successful in either the short- or long-term.
Acknowledgements
The success of this project was made possible by our Ghanaian and Ivoirian colleagues and support staff at KNUST in Ghana, and SCNCI, University of Cocody-Abidjan, and Akoi Kouadio in Côte-d'Ivoire. We thank Mike Adu-Nsiah of the Ghana Wildlife Division, Ghana, and Col. Kouame Amani Denis of the Wildlife Directorate, Ministry of Water and Forests, Côte-d'Ivoire, for their support and permission to conduct this research. Earlier versions of this article were improved by comments from J. Thorbjarnarson, J. Austin, F. Mazzotti, J.P. Ross and two anonymous reviewers. Protocols for animal and human use were reviewed by the University of Florida IACUC (Protocol #E423) and IRB (UFIRB #2006-U-586) committees. Funding was provided by the Wildlife Conservation Society, the Conservation, Food, and Health Foundation, the Columbus Zoological Park Association, Inc. Conservation Fund, IDEA WILD, St Augustine Alligator Farm, and private donations. Many students in Ghana and Côte-d'Ivoire made this project successful through their enthusiasm to learn about crocodiles and to become better conservation biologists.
Biographical sketches
Matthew Shirley is interested in the conservation of wildlife in developing nations and the challenges that this entails. In his research he utilizes multiple tools including model-based population estimation and molecular ecology and systematics. William Oduro has been an advocate for the conservation of Ghanaian, and West African, wildlife throughout his career. He is interested in the international, sustainable trade of reptile species, and how conservation efforts affect local communities. Hilaire Yaokokore Beibro is principally inter-ested in the conservation of forest birds in the Upper Guinea Forest. He is a founding member of the Society for Nature and Conservation in Côte d'Ivoire.
