1. Introduction
Crocodiles are apex predators in freshwater ecosystems throughout tropical regions. Conflicts between crocodiles and human populations date back to the Plio-Pleistocene when these reptiles preyed on earlier hominids in Africa [
1]. Nile crocodiles (
Crocodylus niloticus) are among the main threats to fisherfolk and riverine communities in Africa [
2,
3,
4,
5,
6]. It is estimated that hundreds of deadly attacks attributable to this species occur yearly in Sub-Saharan Africa including South Sudan, with a mortality rate ranging between 50% and 100% [
7,
8]. Nile crocodiles not only pose a direct threat to people and livestock, but they can also indirectly affect the quality of life of people in more remote locations and economically impoverished areas [
9]. On the other hand, Nile crocodiles are also valued prey for humans because of their meat and skins [
10]. Although considered a Least Concern species by IUCN [
11], this species is exposed to local declines [
12,
13] and, therefore, needs continued monitoring, including using innovative survey techniques [
14].
Thus, the relationships between Nile crocodiles and human populations are often complicated and need careful examination and monitoring even at the local scale, especially in the scientifically poorly explored, thus data-deficient regions and the poorest areas of the African continent. South Sudan in East Africa is a prime location to study crocodile–human conflicts as it is a scientifically poorly explored region because of the civil war that has run throughout the country for more than 20 years. Considering the importance of the vast South Sudanese wetlands for both human food security through artisanal fishing and as a prime habitat for Nile crocodiles, the potential for crocodile–human conflicts is very high given that (i) the density of people is relatively high around the Nile, (ii) most of the economic activities are concentrated along the Nile, (iii) the country is a low-income food-deficit country with a per capita GDP limited to USD 1570 [
15], whereby 21% of the population was at level four emergency under the Integrated Food Security Phase Classification in 2018 [
16], and (iv) expansive areas are characterised by riverine and freshwater marshlands (for instance, the Nile river and the Sudd wetlands), thus providing potentially excellent habitats to Nile crocodiles [
17,
18,
19]. A previous study documented a high frequency of attacks by Nile crocodiles on both humans and livestock in the Sudd wetlands, with substantial mortality for fisherfolk and riverine communities [
8].
Given the severity of conflicts between humans and crocodiles, changing human behaviour through environmental education could mitigate these conflicts. Such behavioural changes could also positively impact attitudes towards biodiversity conservation more broadly. This, in turn, could help ensure that future economic development and human population growth are more compatible with conservation needs. Our objective was to lay the groundwork for this effort by gaining a deeper understanding of human perceptions and attitudes towards crocodiles. In this paper, we investigated crocodile–human conflicts in South Sudan through a standardised interview survey on fisherfolk and other people living in the riverine area to respond to these key questions:
- -
How do Nile crocodiles affect the lives and livelihoods of local communities?
- -
What are the attitudes of local communities towards Nile crocodiles?
- -
What strategies are used by local communities to minimise/mitigate the risks of Nile crocodiles’ attacks on human and livestock?
- -
Can the knowledge of the attitudes of local communities be used for applying any management and conservation strategies for Nile crocodiles in South Sudan?
2. Materials and Methods
South Sudan lies within the tropical zone between latitude 3.5° and 12° north and longitude 25° and 36° east, and it occupies an estimated area of 633,906 km
2 (in 2018, [
20]). It is a landlocked country in the Nile River Basin in east–central Africa. The regional climate is tropical with a wet season in April–October (with an average of 100 mm rainfall per month) and a dry season between November and March (5–35 mm per month). During the dry season, the maximum temperature of 38 °C is typical in February.
Poverty is prevalent, with approximately 80% of the population living on less than USD 1 per day [
21,
22]. Nearly 80% of the population relies on smallholder agriculture, farming, and fishing. Fish is crucial for both the food security and livelihoods of communities in and around the Sudd wetlands, thus creating potential human–crocodile conflicts [
8].
Designated as a Ramsar site in 2006, the expansive Sudd wetlands rank among the largest and species-rich wetland areas globally [
23]. Comprising lakes, marshes, and extensive floodplains, the Sudd is renowned for its biodiversity, serving as a critical habitat for numerous endangered species such as the Nile lechwe and the shoebill stork. Its biodiverse aquatic habitats provide essential habitats for fish, offering ideal grounds for their spawning, rearing, and feeding. These habitats remain largely untouched by industrial development, preserving their ecological integrity [
24].
The Sudd wetlands are also vital for local livelihoods, providing ecosystem services and supporting water resources. However, they also present important challenges to humans living in the area, one of which are conflicts between humans and crocodiles, as the wetlands are a prime habitat for Nile crocodiles. In the south, the wetlands are bordered by the Badingilo National Park (also spelled Bandingilo National Park), which is one of the most important migration areas for wildlife in Africa, second only to the Serengeti Mara ecosystem. The park serves as a critical corridor for the migration of various antelope species, including white-eared kob and tiang, which undertake one of the largest terrestrial migrations in the world.
We studied 21 different villages (
Appendix A) and their 85 associated fishing camps from the following five administrative areas of the Southern Zone of Sudd wetlands, mainly situated in Central Equatoria State (
Figure 1): (1) Terekeka (N050 27.1555″ and E0310 45.268) and (2) Northern Terekeka (N050 38.829″ and E0310 43.108″) on the western bank of the Nile; (3) Mangalla on the eastern bank (N050 11.5350″ and E0310 46.164″); (4) Gemeiza on the western bank (N050 44.2738″ and E0310 47.1021″); and (5) Tombek (N050 47.369″ and E0310 42.285″). All these locations are situated surrounding the western and eastern corridors of the Badingilo National Park and include the large swamps 40 km east of Mangalla Payam in Central Equatoria State.
The Badingilo National Park (about 8400 km
2) is characterised by grassland and woodland savanna and is known to be the earth’s second-largest hotspot for ungulate annual migration after the Serengeti. These abundant populations of multiple species of ungulates support large crocodile populations who prey on them [
8,
25].
The diversity of natural resources has contributed to the expansion of human settlements near the Badingilo National Park, whereby the main economic activities fishing, livestock grazing, illegal hunting/poaching of wildlife, charcoal production, collection of reeds and other building materials, collection of fuel wood, and the production of crafts are important aspects in the rural economy. Vegetation and sandy areas along the bank of the Nile provide suitable nesting sites for the Nile crocodiles [
8]. The total human population of the study areas was 47,718 in 2018 [
26]. All study sites are characterised by near proximity of humans and Nile crocodiles (
Figure 2).
Figure 1.
Study area. The map was created using QGIS version 3.20.2-Odense (qgis.org) from public domain map datasets from Open Street Map (
www.openstreetmap.org, accessed on 24 January 2024), diva-gis (
diva-gis.org), Humanitarian Data Exchange, HDX (data.humdata.org), and UNEP-WCMC [
27] for the boundaries of the Badingilo National Park.
Figure 1.
Study area. The map was created using QGIS version 3.20.2-Odense (qgis.org) from public domain map datasets from Open Street Map (
www.openstreetmap.org, accessed on 24 January 2024), diva-gis (
diva-gis.org), Humanitarian Data Exchange, HDX (data.humdata.org), and UNEP-WCMC [
27] for the boundaries of the Badingilo National Park.
The interview survey was carried out from 2018 to 2020 by a team of three researchers. We randomly selected 28 of the 85 fishing camps for interviews. The random selection was across the five administrative areas without a proportional representation of the administrative areas. Male participants were randomly chosen from fisherfolk, farmers, and shepherds who lived in the fishing camps and villages around and inside the Badingilo National Park. Only males live in the fishing camps, according to local culture, since women visit their partners to cultivate vegetables and crops around the camps during the dry season when water levels are low. In our surveys, women were under-represented because they required their male partner’s permission to speak, which was rarely granted. Potential participants were randomly approached when encountered when interviewers visited the fishing camps. A total of 378 local community members were interviewed, of which 21 were women and 357 were men. Each interview was conducted individually and lasted between 20 and 25 min. Only adults age 19 or above were questioned; the age distributions were 19–29 years old, N = 103; 30–49 years old, N = 189; 50–69, N = 76; and 70 years and above, N = 10. Interviewees were given the option to decline the complete interview or single questions. After each interview, the name of the interviewee was requested (optional) and a reference number recorded to avoid replication. Five main questions were asked:
- -
How do Nile crocodiles affect the lives and livelihoods of your local community regarding (Q1.1) restricting freedom of movement at the river bank, (Q1.2) attacking livestock, (Q1.3) destroying fishing equipment, and (Q1.4) attacking community members?
- -
What is the attitude of local communities in your area towards Nile crocodiles regarding (Q2.1) danger, (Q2.2) honour, (Q2.3) hate, (Q2.4) fear, and (Q2.5) enemy?
- -
How are Nile crocodiles being used by the local community in your village regarding (Q3.1) local meat trade, (Q3.2) illegal farming practices, (Q3.3) skin/leather trade, (Q3.4) ornamental purposes, and (Q3.5) religious purposes?
- -
What are the perceptions of local communities towards eating Nile crocodile meat at your village regarding (Q4.1) favouring longevity, (Q4.2) enhancing sexual performance, (Q4.3) for anti-witchcraft, (Q4.4) for medicine purposes, and (Q4.5) boosting business and good luck?
- -
What strategies are preferred by your local community to minimise/mitigate the risks of Nile crocodile attacks on humans and livestock regarding (Q5.1) creating a sanctuary/protected area for crocodiles with no access to people (while allowing free hunting outside it), (Q5.2) reducing the livestock activities along the rivers, (Q5.3) destroying the crocodile habitat to make them leave the area, (Q5.4) emigration of the whole community to safer places with no crocodiles around, and (Q5.5) promoting the hunting of crocodiles?
These questions were asked in the local language and verbally phrased such that answers could be given on a Likert scale: (a) strongly agree, (b) agree, (c) neutral, (d) disagree, or (e) strongly disagree (thereafter, “agreement options”). For example, Q2.1 was formulated as “Do you perceive that crocodiles pose a danger to you?” The information provided by the interviewees was augmented through group discussion in each location to validate the individual interviews on a group level. Following these questions, we provided the interviewees with an opportunity to express their attitudes towards crocodiles and biodiversity conservation, aiming to gather general background information.
After completing the interviews, in-depth telephone interviews were held with six senior government officials (two from the HQ Office in Juba, two from the State Department of Wildlife, and two from the Terekeka County Authority). Qualitative data were gathered from key informants including thirteen village chiefs or community leaders and sixteen heads of fishing camps to understand their views towards the conservation of Nile crocodiles. All this information guided us in interpreting the collected quantitative data.
All work was undertaken under a research permit from the Ministry of Internal Affairs, National Government, Juba, South Sudan. Because of the COVID-19 pandemic, the research team members used the relevant safety measures defined by the Ministry of Health of South Sudan when making the face-to-face interviews, including social distance measures and the use of face masks throughout the research period.
The various frequencies of answers were analysed by contingency tables χ
2 tests. To evaluate whether answers were influenced by the intensity of the threats due to crocodiles, we used the dataset provided by Benansio et al. [
25] who reported, for the period 2018 to 2020, a total of 23 fatal crocodile attacks on humans and 355 attacks on livestock, of which 166 were killed. Stratified by district, these data were used to assess their effect on answers.
Generalised Linear Models, GLMs [
28,
29], were used to test the relationship between questionnaire answers (from Q1.1 to Q5.5) and crocodile attacks on humans and livestock in the five surveyed village districts. GLMs can handle ordinal response variables including Likert scale responses by using appropriate link functions. GLMs do not assume that the response variable has a normal distribution, which is often violated in Likert scale data. The values of the estimation parameters in the Generalised Linear Model are obtained by maximum likelihood (ML) estimation through iterative computational procedures. Tests for the significance of the effects in the model can be performed via the Wald statistic. Detailed descriptions of these tests can be found in McCullagh and Nelder [
29]. In the models, answers were used as dependent variables, and the numbers of crocodile attacks on humans and livestock were used as predictors. All models were computed with the all-effects procedure, the identity link function, and a normal distribution of errors [
30]. The estimates of the models indicated whether the probability of a given answer increased/decreased with the increasing number of crocodile attacks. Attacks on humans and livestock were considered separately as predictors, but the number of attacks on all types of livestock (cattle, goats, and sheep; see Benansio et al., [
8]) was cumulated for model evaluation.
Software STATISTICA 13.0 was used to perform all the analyses, with alpha set at 5%. Nonparametric tests were used when the variables were not normally distributed.
4. Discussion
We investigated attitudes and perceptions towards crocodiles by applying structured questionnaires. The design of interview questions in structured interviews may bias answers or limit the range of response [
31]. Although our survey questions appear biased towards measurements of conflict, we believe, based on our experience working with fisherfolk in the area, that the question reflects fisherfolk attitudes. Despite the emphasis on conflict, several results indicate nuances and complex attitudes. For example, the perception of crocodiles as enemies is relatively low (when asking how crocodiles are used. Our study focuses on conflict in the context of human and livestock mortality by crocodiles [
8] but does not address how common or uncommon such conflicts are. Whilst focussing on the conflicts, we also probed positive attitudes to crocodile conservation. It became clear that local communities were not strictly against crocodiles but also sought their conservation through the establishment of sanctuaries. We recommend a follow-up study addressing this specific question, e.g., questioning how often interviewees see crocodiles without conflict. This is important to fully understand the crocodile–human conflict in the current global climate where conflicts—such as crocodile and alligator attacks in Florida or brown bear conflicts in national parks and residential areas—are often sensationalised and predators are villainised for exhibiting natural behaviour where humans encroach into predator habitats, for example, by ever-increasing urbanisation.
Fisherfolk in the Sudd area living near the River Nile perceived that their lives and livelihoods are seriously affected by Nile crocodiles through attacking humans and livestock, not only by the loss of human lives and livestock but also restricting the freedom of movement of communities along the river bank and causing damage to fishing equipment, which is the main operative expenditure of the fisherfolk [
4,
32]. Damaged fishing nets can be fixed, but it takes time and money [
4,
33]. While the fishing nets are being repaired, the fisherfolk cannot fish unless they purchase or hire fishing nets. For the fisherfolk to mitigate the impact of crocodile damage on fishing equipment, fisherfolk need to equip themselves with more than eight fishing nets annually, but the extra expense is economically challenging considering the overall poverty. The damage to fishing nets occurs in the flooding season when most fish species migrate to the floodplain for breeding and the crocodiles migrate to the floodplain to feed on these fish species. Similarly, fisherfolk in the Okavango area of Botswana view crocodiles as a serious threat to their lives and equipment, and they delay entrance into the floodplain when water levels drop due to the risk of crocodile attacks [
34]. Moreover, crocodiles eat fish from nets. In the dry season between November and January, most crocodiles migrate to breed in the sandy area along the river bank. The reproductive season is considered dangerous as reported by crocodiles in South and Central America [
35,
36]. During this time, parents strongly warn children not to play in the sandy area along the river bank, and women and girls are advised not to fetch water in dangerous areas. These general views of the various interviewees are set against the backdrop of 23 people being killed and 355 livestock attacked by crocodiles in the same villages between 2018 and 2020 [
25]. The GLMs show the number of attacks on both humans and livestock significantly influenced several answers made by the interviewees. Although the attitude of local communities towards the conservation of crocodiles is negative because of the above-mentioned reasons, crocodiles are also actively used by people, and the answers to some of the questions revealed complex and nuanced patterns. Local communities were promoting the eating of Nile crocodile meat to enhance longevity, drug, and sexual potency. On the other hand, certain families and clans have a longstanding tradition of revering crocodiles for cultural, ritual, and spiritual purposes, similar to practices observed in other parts of Africa [
37,
38,
39,
40] and elsewhere [
41]. The key informant interviews conducted in the Sudd wetlands revealed the presence of over 34 such families, with 14 located in the Southern Zone between Mangalla, Gemeiza, and Terekeka.
Conservation of the Nile crocodile populations and the effective management of their habitats depend on understanding these human attitudes and our ability to recognise and predict species–habitat interactions [
42]. People are known to destroy any nesting they come across while collecting resources in the floodplains [
34]. Human activities in the immediate surroundings of the wetlands cause significant stress for female crocodiles and can lead to the abandonment of nesting sites and breeding habitats. The rapid population growth demands more resources, and this has a strong link with the expansion of human settlement in a detrimental way to wildlife habitats [
43]. Thus, cultural attitudes, the fear of crocodiles because of the real danger, and human habitat encroachment will likely increase the pressure on the species in the future given the increasing human population density and the increasing pressure for development after the end of the civil war. However, the issue is complex as promoting crocodile killing and destruction of their habitats were not considered as crucial strategies to reduce the risks of Nile crocodile attacks by the majority of the local communities.
This complexity is highlighted by the agreement of local communities on the need to destroy Nile crocodile breeding habitats on the one hand and the need to establish crocodile sanctuaries on the other hand. The polarised views on promoting illegal hunting also underscore the intricate challenges in managing human–crocodile conflicts, necessitating nuanced, community-driven strategies. The low perception of crocodiles as enemies suggests that communities may adopt a balanced or integrated approach to coexisting with crocodiles, recognising their significance beyond the immediate threats they pose. Promoting awareness creation and outreach programs to local communities on conservation education are positive strategies to increase awareness of the real risks and crocodile behaviour, thus reducing human–crocodile conflicts. The nuanced attitudes revealed in certain questions provide a valuable foundation for raising awareness and designing targeted promotional campaigns. The creation of Nile crocodile sanctuaries in selected parts of the Sudd wetlands can minimise the risks of illegal hunting by the local communities especially as this option is being positively considered and accepted by local communities. Such sanctuaries might come with economic benefits from ecotourism.
During interviews with key informants, the issue of promoting crocodile farming for commercial purposes as a source of income and livelihood was raised as a possible sustainable strategy for the management of this species at the local scale. However, the increasing development of Nile crocodile farming has been associated with diseases transmittable to wild crocodiles and humans [
44,
45,
46,
47]. For example, crocodile pond water can have high concentrations of
Salmonella spp. [
48,
49,
50,
51,
52,
53]. Crocodile meat is an important source of
Salmonella contamination with a human exposure risk, especially during slaughter and dressing operations [
53]. Crocodile meat consumption may lead to infections with a variety of bacteria (
Salmonella spp.,
Vibrio spp.), parasites (
Spirometra,
Trichinella,
Gnathostoma, pentastomids), and intoxications by biotoxins [
54]. Amon viruses, there is a zoonotic disease risk through arboviruses such as Rift Valley Fever virus [
55]. Contaminated water can play a role in the transmission of the Rift Valley Fever virus. The virus can be present in the blood, tissues, and bodily fluids of infected crocodiles, and when these contaminated fluids come into contact with water sources, there is a risk of transmission to humans and animals that drink or come into contact with the water. Thus, the introduction of crocodile farms, as widespread elsewhere in Africa including neighbouring Ethiopia [
56,
57], needs careful consideration of the possible negative effects through increased disease risk to humans and wild crocodiles.