Introduction

Carnivores incur energetic costs and risk injury when hunting and killing their prey [1]. Each stage of a hunt e.g., locating, stalking or chasing, and killing prey requires considerable energy expenditure [2–8]. Prey pursuit and capture entails physical risks to predators [9], including the possibility of being injured [10] or killed [11, 12]. Following a successful hunt, a carnivore may face the added cost and risk of defending its prey from being stolen (kleptoparasitized [13]) by conspecifics, other carnivore species [14–18], or by humans [19, 20]. While considerable attention has been paid to humans deliberately killing carnivores (e.g., trophy hunting [21–23], and lethal removal as a form of management [24]), non-lethal impacts have received far less consideration [25–28].

Human activities such as ecotourism viewing can disturb and stress carnivores [25, 29, 30], including while the animals are hunting [31–32]. Humans also target carnivore kills to obtain meat for personal consumption and to sell or trade [19, 20, 33–36].

Although early hominids are believed to have, at times, hunted game [37], the practice of early hominids obtaining meat from carnivore kills, referred to as human kleptoparasitism, has been well-documented [38–41]. Human kleptoparasitism is known to still occur in modern times (i.e., during the past 60 years) [19, 20, 33, 34, 36]. In their examination of data from human-carnivore conflicts in Uganda between the years 1923–1994, including human kleptoparasitism, Treves and Naughton-Treves (1999) highlighted the interlinked evolutionary histories of humans and early hominids that shared their habitats with large carnivores. However, to date, the geographic scale, frequency, and the potential impact on carnivores of the modern practice of human kleptoparasitism have not been adequately investigated [20]. A paucity of reports 35 years ago led Sunquist and Sunquist (1989) to conclude that while human kleptoparasitism appeared to pose no significant threat to individual carnivores in the short-term, repeated losses could negatively impact carnivore populations in the long-term. Lacking an in-depth exploration of the modern practice, it is likely that both the costs to individual carnivores, and the population-wide effects, have been underestimated.

Humans are now considered a “super-predator” [42], and human disturbance in the proximity of kills has been shown to induce fear in large carnivores [43–45], likely resulting in stress, reducing the overall time spent at the carcass [44], and reducing the total amount of meat consumed [36, 46, 47]. Even a perceived risk of encountering humans may cause carnivores to alter their behaviors and movements [27, 44, 45, 48], including abandoning their kills [49], or shifting to increased nocturnality [50].

Loss of kills to humans increases the frequency at which a carnivore must hunt [36, 43, 47], thereby increasing the costs and risks for an animal to obtain the equivalent amount of energy i.e., meat [44]. For carnivores that return to their kill following human disturbance, loss of a partial or complete carcass decreases or eliminates the amount of meat remaining for subsequent feeding bouts [33], reducing the energy gained per kill [44]. Removal of carrion also reduces or eliminates food resources for scavengers [36] potentially disrupting scavenger communities and associated ecosystem services e.g., nutrient cycling [51–54].

Humans that actively chase carnivores from their kills to obtain meat (confrontational scavenging) may injure carnivores, or may themselves be injured or killed in the process [19, 55]. Carnivores that injure humans in defense of their kills may subsequently be destroyed as problem animals [19, 55]. Additionally, once kills have been located, humans may poison carcasses [56–58] to kill carnivores and scavengers under the auspices of animal control, in retaliation for livestock depredation [59, 60], or to obtain body parts for illegal trade [61, 62].

Despite this broad range of known impacts, our understanding of the extent and frequency of modern human kleptoparasitism is lacking. The few previous reports available have focused primarily on anecdotal or regional accounts [19, 20, 33, 34]. However, other factors e.g., geographic location of the kill within or outside of protected area boundaries, legal ramifications of possessing game meat, and cultural or religious taboos, may influence the occurrence of human kleptoparasitism.

Given that the majority of modern threats to large carnivores are anthropogenic [63–65], and that anthropogenic threats are likely to be cumulative, further investigation into the potential impacts of modern human kleptoparasitism on carnivores is warranted.

To address this data gap, we conducted a questionnaire survey among lion experts and stakeholders with knowledge of the African lion, Panthera leo, in the current African range states to elucidate the geographic extent, frequency, and socio-economic drivers of modern human kleptoparasitism. Due to their conspicuous social behaviors and reliance on large-bodied prey [66], lions may be particularly vulnerable to being detected on their kills [67]. While our survey focused on the kills made by lions, respondents were asked about the occurrence of human kleptoparasitism on other large carnivores in their areas i.e., leopard, Panthera pardus, spotted hyena, Crocuta crocuta, wild dog, Lycaon pictus, cheetah, Acinonyx jubatus.

This study represents the first effort to systematically characterize and assess modern human kleptoparasitism across the range of the African lion, and to illustrate the suite of potential costs faced by carnivores exposed to this practice.

Methods

We developed a questionnaire survey with a goal of quantifying the geographic extent, rate of occurrence, trends, and other associated factors e.g., legality, cultural or religious taboos, of humans taking meat from the kills of large carnivores in sub-Saharan Africa. To prospectively recruit participants, we distributed a pre-structured questionnaire (S1 Appendix) among lion experts in the African Lion Working Group (ALWG), a working group of approximately 90 members under auspices of the IUCN Cat Specialist Group. The ALWG represents active lion researchers, and other stakeholders, whose work may not involve field studies. Thus, we did not anticipate receiving responses from every ALWG member. Additionally, we circulated the survey among approximately 35 other stakeholders with knowledge of wild carnivores (with a focus on P. leo) i.e., conservation organizations, lodge owners, and professional hunters in lion range states. The goal was to query lion experts’ and stakeholders’ existing knowledge i.e., direct observations, received reports, and opinions of the practice within their geographic area(s). Although some survey recipients did not conduct scientific studies per se, the geographic area described in a respondents’ survey is hereafter referred to as their study area.

On 31 May 2020, we circulated the questionnaire in the text of an e-mail and as an attached document with a request for respondents to reply through whatever means was most convenient. We sent a follow-up reminder on 15 June 2020, and set a cut-off date of 2 July 2020 to receive responses. If respondents had knowledge about more than one study area, we requested that they complete one questionnaire for each study area. In addition to completed questionnaires, some respondents also provided comments via e-mail; others provided only e-mail comments but did not complete the questionnaire. Some respondents who completed the survey did not answer every question, thus reported sample sizes vary. Additionally, some research teams designated one person to reply on behalf of the team with information from their study area; such replies were tallied as one respondent. All answers, whether obtained from questionnaires or in e-mail format, were incorporated into the results to the extent possible. Apart from maintaining the geographic context (country) of the data received, responses were combined and the specific locations of the study areas, as well as the identities of the responding individuals, were kept confidential.

The pre-structured survey form contained twelve questions focused on whether kleptoparasitism by humans had been observed or reported to occur in the respondent’s study area, how information was obtained, and details of the practice e.g., how often and by whom (S1 Appendix). Further, we examined whether the kleptoparasitism events involved wild prey or livestock, and how prey type influenced human attitudes and responses. Respondents were asked how long they had worked in the study area (tenure), and, whether in their expert opinion, kleptoparasitism represented a potential threat to lions in their study area (Question #8) or across the species’ range (Question #9). For respondents who completed surveys for more than one study area, we counted their answer to Question #9 only once regardless of how many surveys they completed. All other answers were tallied as representing individual responses. Because the answer options to the frequency of occurrence were presented as a range of values e.g., 1–5, 6–10, etc., (S1 Appendix), we used the average value within each option when considering frequency of occurrence on respondents’ tenure. For example, for option a) 1–5 times/year, we used a value of 2.5 times/year when examining frequency of occurrence on tenure.

We performed descriptive statistics on the different responses to each question, and performed a Welch two-sided t-test to examine whether knowledge of the presence of human kleptoparasitism in a study area was influenced by the respondents’ tenure in that area. We also performed chi-square tests of independence to examine whether a respondent’s knowledge of kleptoparasitism in their study area influenced their opinion regarding whether or not kleptoparasitism represented a potential conservation threat to lions across the species’ range. We considered a result statistically significant at p < 0.05.

Results

We received 56 completed questionnaires from 33 individual respondents (14 respondents completed questionnaires for more than one study area). An additional 5 respondents provided e-mailed comments but did not complete a survey. The 38 total respondents were comprised of 31 ALWG members and 7 stakeholders (response rate 34% and 20%, respectively). Thus, our results represent experts’ knowledge of extant lion populations from 56 studies located in 15 of 22 African lion range states (Fig 1). The 15 surveyed range states include all resident lion populations in West Africa and nearly all East and Southern African range states that, combined, account for > 95% of Africa’s lions [68] (S1 Table). The occurrence of human kleptoparasitism on the kills of wild carnivores was reported in 37 of the 56 questionnaires (66%) received, confirming modern occurrence of the practice in 12 of the 15 range states surveyed (Fig 1; S1 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Map of the African lion range states surveyed for modern human kleptoparasitism.

Highlighting indicates countries in which occurrence (green) or no occurrence (brown) was reported. The geographic regions of West, Central, East, and Southern Africa are demarked by dotted black lines. Figure generated with QGIS software by author Laura Bertola using layers (e.g., country borders) which are copyright-free to re-purpose for non-commercial use.

https://doi.org/10.1371/journal.pone.0308068.g001

The number of years (tenure) spent by individual respondents in their respective study areas ranged from 1 to 30 years (X = 11.9 years). There was no significant difference in tenure between respondents who reported occurrence of kleptoparasitism in their area (X = 10.8 years) and those who reported no occurrence (X = 11.6 years, Welch two-sample t-test (26.9) = 0.3, p = .76). There was no significant association between a respondent having reported human kleptoparasitism within a study area and whether, in their expert opinion, this practice represented a threat to lions across the species’ range (Pearson’s chi-square test χ2 (1, N = 46) = 0.004, p = .94).

A respondent’s knowledge of whether kleptoparasitism occurred in their study area was most often gained through reliable reports (40%) and direct observation (35%), followed by local gossip (25%) (Table 1). A reliable report was subjectively defined as a field assistant, wildlife officer, professional hunter, or conservation organization who themselves had directly observed the practice and subsequently relayed that information to the survey respondent.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Respondents’ source for knowledge on the occurrence of kleptoparasitism, strategy for obtaining meat from kills, who procured meat, and nature of activity.

https://doi.org/10.1371/journal.pone.0308068.t001

Respondents indicated that in half (50%) of all reported occurrences, humans actively chased lions from kills; 36% of the time humans scavenged meat from kills when lions were not present. An additional 13% of cases involved other strategies of humans obtaining meat, including responding to playbacks broadcast by researchers to attract carnivores (Table 1). In the latter case, because the playback imitated the sounds of a dying prey animal, it could be assumed that the intention of the approaching humans was to obtain meat by actively chasing the lions off of a fresh kill. However, because this was an assumption, these cases were grouped under “Other”.

Meat procurement was reported to be undertaken mostly by local residents (59%), or game scouts (29%). Game scouts are here defined as employees working for a wildlife authority or private organization who are trained to combat wildlife crimes and whose duties include apprehending poachers, confiscating wire snares and firearms, etc. A smaller percentage (12%) of cases involved non-residents (people who did not reside full-time in the community) i.e., transient or nomadic people, refugees, or seasonal workers (Table 1). The activity was reported as being largely opportunistic (88%), although a few respondents (10%) reported that the act involved deliberate targeting and following lions that were engaged in hunting (Table 1).

Of those respondents reporting kleptoparasitism, most (53%) had encountered human kleptoparasitism in their study area 1–5 times; others reported 5–10 occurrences (8%), or more than 10+ occurrences (14%). An additional 25% of respondents were aware of kleptoparasitism occurring in their study area, but did not know how often it occurred. Because length of tenure varied by individual respondent, we calculated frequency of occurrence per year of respondents’ tenure so that rates were comparable across study areas and time periods (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Frequency of occurrence of kleptoparasitism in relation to respondents’ tenure.

Square points show the tenure of nine respondents who reported kleptoparasitism occurring in their study area at unknown frequency.

https://doi.org/10.1371/journal.pone.0308068.g002

Most respondents (49%) did not know whether the activity was increasing, stable, or decreasing; 30% reported steady occurrence of the practice, with smaller numbers reporting a decrease (14%) or increase (8%) during their tenure (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Trend of occurrence, level of perceived threat within a study area and across the species’ range, cultural/religious significance, and prey type involved with human kleptoparasitism on lion kills.

https://doi.org/10.1371/journal.pone.0308068.t002

Most (62%) respondents perceived human kleptoparasitism as representing no serious threat i.e., having no negative impact to lions within their study area. However, 16% of respondents perceived human kleptoparasitism as a possible threat with a potential to negatively impact lions within their study area (Table 2); 22% reported insufficient knowledge on the topic.

From a cultural standpoint, the majority reported the practice of obtaining meat from carnivore kills as being neutral (41%), or socially acceptable (35%). Two respondents reported that kleptoparasitism went against local religious beliefs, although one indicated that the practice occurred despite the taboo. The remaining respondents (19%) reported insufficient knowledge on the topic (Table 2).

Results showed that human kleptoparasitism was geographically widespread (Table 3). Prey type (wild game vs domestic livestock) was an important component in the likelihood of humans taking kills. Of the 37 respondents, slightly more than half (54%) reported carnivores losing wild game to humans, while the remaining 46% reported that humans only took the meat if carnivores had killed livestock.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Trend, perceived threat of kleptoparasitism within study areas, and prey type by region in surveyed range states where human kleptoparasitism was reported.

https://doi.org/10.1371/journal.pone.0308068.t003

Because livestock (numbers, species) and culture/religious practices vary regionally, we also considered trend, perceived threat within a study area, and prey type by consolidating responses into one of three regions: West and Central Africa (regions combined due to small sample sizes), East Africa, and Southern Africa. In West and Central Africa, livestock was the predominant prey type involved in human kleptoparasitism, whereas in both East and Southern Africa, humans took wild game and livestock prey in close to equal measures (Table 3).

In all regions, a high percentage of respondents reported insufficient knowledge regarding trends of occurrence and whether, in their expert opinion, the practice represented a threat to carnivores (Tables 2 and 3). Regionally, a greater percentage of respondents from Southern Africa (25%) perceived the practice as possibly representing a threat to lions within their study areas, compared to respondents from West and Central (17%) or East (7%) Africa (Table 3). More than one-quarter (28%) of those responding perceived human kleptoparasitism on lion kills as representing a threat across the species’ range (Table 2).

Human kleptoparasitism on kills made by other large carnivore species also was investigated. More than half (20/37, 54%) of respondents reported insufficient knowledge on species other than lion. However, because the survey targeted experts and stakeholders with knowledge specifically on lions, and because not all other species of large carnivores occur throughout the entire lion range, the number of reports of meat taken from other species are not directly comparable. In summary, twelve respondents had knowledge of leopard losing kills to humans, seven reported occurrence of meat theft from spotted hyena, while fewer (4) and (3) respondents reported meat theft from cheetah and wild dog, respectively.

Discussion

For millions of years, large carnivores have competed over access to carcasses, adapting their behaviors and social systems to maximize benefits while reducing costs of intra and interspecific kleptoparasitism [15, 66, 69–72]. Concurrently, early hominids are believed to have included meat in their diet both from scavenging and actively chasing carnivores off kills [41, 73–76]. This practice, which may have represented a survival strategy for ancient humans, still exists in present-day Africa as a means for people to obtain animal protein. Traditional hunters, such as the Hazda, still obtain meat by scavenging from lions [73, 77, 78]. Schaller (1972) [79] demonstrated how readily humans can displace lions to scavenge their kills [80]; other cases of modern human kleptoparasitism have been documented [19, 20].

If large carnivores have evolved to fight for possession of their kills, sometimes losing that battle to competitors, why should modern human kleptoparasitism be a cause for concern?

The dynamics between modern humans and large carnivores have changed. The majority of modern threats to large carnivores are anthropogenic e.g., habitat loss and alteration [64], reduction of prey populations [63], and retaliatory killing [65]. Growing human populations and associated development adjacent to shrinking protected areas is increasing human-wildlife conflicts [81, 82] including direct competition for wild prey [83, 84], although not all studies agree that human-carnivore conflict is increasing [85] (but see [86]). As “super-predators” [42], humans now increasingly impact large carnivores inducing fear [43, 44], causing stress [25, 29], and altering carnivore behavior, activity patterns, and movements [27, 43, 44, 45, 48, 50, 87, 88].

Consequences of human disturbance may be most severe when carnivores are feeding [26, 32], resulting in a reduction of the total amount of meat consumed [36, 46, 47]. The mere perception of human threat i.e., the sound of humans conversing, caused pumas, Puma concolor, to shorten their feeding periods [44]. Carnivores that lose substantial portions of their kills must increase their hunt frequency and kill additional prey to replace the resources lost [36, 43, 44, 47]. Stress, energy expenditure, and risk of physical injury associated with replacing lost resources may especially impact vulnerable individuals e.g., pregnant or nursing females [89], young cubs, and animals in poor condition, as well as the most vulnerable populations e.g., lions in West and Central Africa where prey densities are very low and, thus, the energetic costs of replacing lost prey are especially high [90, 91]. Moreover, the resultant increase in predation can potentially exacerbate human-wildlife conflict if carnivores target livestock to replace their lost kills. Ultimately, sub-lethal physiological effects of human disturbance may reduce individual fitness or survival [92, 93], and human-induced behavioral changes may alter the functional role of large carnivores as apex predators [44, 94].

Our study found modern human kleptoparasitism to be geographically widespread in sub-Saharan Africa. Surveyed experts confirmed the modern practice in 12 countries in Africa that constitute the majority of extant lion populations [68]. This represents the minimum geographic extent; kleptoparasitism may occur in lion range states not surveyed in our study. Our results may also contain reporting bias if those aware of the practice were more likely to report it; this applies both to survey recipients and to local people. Further, the likelihood of local people detecting and reporting kleptoparasitism to our experts may have been greater in areas of higher human density. Conversely, reluctance of local people to disclose the activity may have resulted in underestimates of occurrence. Thus, it is unlikely that non-response at either the expert or local level directionally biased our results. Nearly one-half of the survey respondents cited insufficient knowledge as to whether human kleptoparasitism was increasing, stable, or decreasing, underscoring a dearth of information on the modern practice.

Human detection of kills was largely opportunistic. However, once a kill was found, humans actively chased carnivores off of the carcass 50% of the time. More than half of all reported cases involved humans taking kills of wild game. The taking or possession of game meat is a largely illegal activity in most African countries [95], and the risk of disciplinary action may prompt discretion among those involved. This includes enforcement personnel; in about one quarter of the cases, anti-poaching game scouts were reported to be secondary participants, often keeping confiscated meat for themselves. Similarly, Treves and Naughton-Treves (1999) noted that game wardens in Uganda likely under-reported successful instances of humans scavenging on carnivore kills. Fenced reserves, or reserves with strict rules against kleptoparasitism appeared to have reduced (reported) incidents, or absence of incidents. However, where enforcement is lacking, kleptoparasitism has been associated with other unlawful activities i.e., illegal diamond extraction in remote parks in the Central African Republic [20].

In contrast to game meat, people have legal ownership of livestock and viewed repossession of depredated livestock carcasses as ‘taking back’ the meat of an animal they already owned. Thus, prey type is an important factor in assessing kleptoparasitism of carnivore kills because not only do people have the legal right to repossess a livestock carcass, they also may be more motivated to do so from a herding (cultural) standpoint. As a result, both humans and carnivores may face increased risk of injury in confrontations over possession of livestock carcasses.

Culture, religion, and social status influenced whether communities or individuals condoned kleptoparasitism. Therefore, it is possible that, within the range states where kleptoparasitism was confirmed, the occurrence was localized or sporadic. For example, meat theft from carnivores is common practice within the nomadic Mbororo in Northern Cameroon [96]. In contrast, Muslims, a prevalent religion in Sudan and Tanzania, hold strict taboos regarding how animals must be slaughtered (i.e., halal) [97], making meat obtained through kleptoparasitism unfit for human consumption. Some viewed kleptoparasitism as socially beneath them, while others who stood to gain a share of the meat were more accepting of the practice. Unsurprisingly, economic status was reported as an important factor; community members with low income, including park staff with low remuneration, were reportedly more likely to partake in the activity often to supplement their protein supply. Because people who kleptoparasitize carnivores gain an immediate and tangible benefit i.e., meat to consume or sell for cash, it raises the question as to why this practice is not more ubiquitous. It is possible (and highly likely) that insufficient knowledge even among experts and stakeholders, attributable at least in part to a reluctance among local people to report the practice, resulted in significant underestimates of the frequency with which modern human kleptoparasitism occurs.

Conclusion

Our survey found modern human kleptoparasitism on kills made by lions, and possibly other large carnivores in Africa, to be geographically more widespread than previously reported. Humans disrupted carnivores at their kills of wild game and livestock to procure game meat and compensate for losses, respectively. Economic hardship was a main driver of the practice, with game meat important both for personal consumption and trade. Although modern human kleptoparasitism may appear inconsequential in comparison to other anthropogenic threats, the frequency at which carnivores lose kills to humans is likely underestimated. The fact that our surveyed experts repeatedly cited a lack of information, and the strong likelihood that cases go unreported, suggests that the impacts of modern human kleptoparasitism may be higher than are currently recognized. Whether chased or scavenged, the loss of meat or entire carcasses to humans, particularly wild game, represents additive costs to carnivores that should not be dismissed. We argue that human kleptoparasitism of carnivore kills poses an important yet largely cryptic anthropogenic threat that negatively impacts not only the individual carnivores that lose their meal, but also scavenger populations and the broader ecological community. Further, the potential health threats to humans from taking meat from carnivore kills have not been investigated. We propose that modern human kleptoparasitism warrants greater scrutiny.

Supporting information

Acknowledgments

We thank the members of the African Lion Working Group, conservation organizations, professional hunters, and other stakeholders who participated in this study. Sincere thanks to A. Mateos and A. Treves for their constructive reviews that improved this manuscript. We dedicate this paper to the memory of our colleague, student, and friend, the late Kennedy Kariuki, a passionate conservationist and wildlife researcher, who compiled the original survey data into a preliminary report.