Close encounters: responses of road users to wild chimpanzee road crossings in Uganda

Expanding road networks increasingly encroach on wildlife home ranges, intensifying human-animal interactions. While ecological consequences like habitat fragmentation or wildlife mortality from vehicle collisions are well-studied, immediate behavioral responses of human road users during wildlife encounters remain underexplored despite their critical impact on safety and conservation. This study investigates behavioral responses of vulnerable (pedestrians, cyclists) and non-vulnerable (motorists) road users during wild chimpanzee (Pan troglodytes) crossings in a human-dominated landscape in Bulindi, Uganda. We analyzed 376 opportunistically recorded interactions over 38 months to identify factors influencing road users’ caution (stopping or slowing down without interfering with the chimpanzees, or retreating) and interest (stopping or slowing down to watch, including approaching chimpanzees). Local pedestrians and cyclists displayed significantly higher caution than motorists. Caution and interest did not differ significantly by age or sex among pedestrians. Road users were generally more cautious and curious when chimpanzees actively crossed the road compared to waiting at the roadside. Conversely, encounters with larger chimpanzee groups correlated with reduced caution. Interactions with local field assistants, employed to monitor the chimpanzees, increased both caution and interest among road users. Although no collisions occurred, seven near-miss incidents were documented, primarily linked to poor visibility and reckless motorcyclist behavior. These results show that road user behavior is influenced by physical vulnerability, chimpanzee visibility, and contextual cues, highlighting potential risk factors for both humans and chimpanzees. Despite chimpanzee adaptability to anthropogenic environmental change, roads and traffic remain significant threats to this endangered species. Our findings emphasize the need for measures to limit vehicle speed (e.g., speed humps, signs, traffic controls) and improve visibility on roads traversing chimpanzee habitats, coupled with targeted awareness campaigns for local communities, tourists, and non-resident workers, as well as the integration of locally based and trusted personnel into road safety strategies.

In response to the increasing demands of a growing human population and thriving economies, road networks have expanded extensively, reaching even the most remote regions of the globe (Laurance et al., 2014). Nevertheless, the extension of road networks has significant impacts on wildlife populations, such as habitat loss and fragmentation (Barber et al., 2014; Sánchez-Fernández et al., 2022), altered movement patterns (Shepard et al., 2008; Asensio et al., 2017; Andrasi et al., 2021), increased pollution (Bignal et al., 2007; Krief et al., 2020), and mortality from vehicle collisions (Schwab and Zandbergen, 2011; Cunneyworth and Duke, 2020; Tayebwa et al., 2025). These negative effects highlight the importance of integrating ecological considerations into transportation planning and implementing mitigation measures (Jaeger et al., 2005; van der Ree et al., 2011; Bennett, 2017).

Although previous research on the impact of road traffic on wildlife has primarily centered on mortality and habitat fragmentation (Trombulak and Frissell, 2000; Bennett, 2017; Mtega et al., 2025), there has been a lack of investigation into the immediate behavioral reactions of human road users and wildlife to road crossing encounters (Vidya and Thuppil, 2010). Most studies have focused on vehicle collisions with ungulates, a serious road safety concern across the United States and Europe (Groot Bruinderink and Hazebroek, 1996; Jägerbrand and Antonson, 2016; see Stanley et al., 2006 and Grace et al., 2015 for simulated environments). To our knowledge, one study examined motorists’ reactions to elephant crossings in India (Vidya and Thuppil, 2010), but no literature is available regarding the responses of pedestrians to animal crossings. This gap is critical: understanding how different road users react when encountering wildlife crossings holds valuable insights for conservation management strategies, such as raising awareness, minimizing disturbance to wildlife, and preventing traffic collisions with them (Kioko et al., 2015; Shilling et al., 2020).

This study helps address this gap by providing the first quantitative assessment of road-user behavioural responses during real-world road crossings by wild chimpanzees (Pan troglodytes) in an unprotected, human-occupied landscape in Uganda. It is also the first study to examine such responses in relation to great apes, incorporating both motorists and non-motorised road users. We focused on chimpanzees because they are endangered (Humle et al., 2016), increasingly impacted by road development (Arcus Foundation, 2018), and particularly likely to arouse people’s interest as a ‘charismatic’ flagship species of sub-Saharan Africa (Plumptre et al., 2010; Albert et al., 2018). Chimpanzees are also large-bodied mammals with high terrestrial mobility, which gives them a strong physical presence on the road and makes them easy to detect but also more susceptible to vehicle collisions (Rytwinski and Fahrig, 2015; McLennan and Asiimwe, 2016; Borza et al., 2023). For motorists, colliding with large mammals is likely to cause damage to vehicles and risks injury to occupants (Bjornstig et al., 1984; Huijser and McGowen, 2010; Kioko et al., 2015), unlike other taxa such as reptiles or other smaller animals (Mesquita et al., 2014; Crawford and Andrews, 2016). Besides, chimpanzees are potentially dangerous animals that may threaten and occasionally attack people, especially children (Reynolds, 2005; Hockings et al., 2010; McLennan and Hill, 2010; McLennan and Hockings, 2016). Thus, encounters with chimpanzees can endanger pedestrians’ physical safety and evoke fear (McLennan and Hill, 2012). Research in transportation psychology has shown that road users adapt their behavior to perceived hazards and unexpected events on roads. For example, drivers commonly adjust speed, stopping behavior, and attention in response to pedestrians, obstacles, and other roadway hazards (Ventsislavova et al., 2016; Sheykhfard and Haghighi, 2017; Bella and Nobili, 2020), whereas pedestrians and cyclists generally exhibit heightened caution (Harrell, 1991) because of their greater physical vulnerability. Consequently, the sudden appearance of chimpanzees on the road is expected to provoke various reactions from road users, whether pedestrians or motorists.

This study aims to identify patterns in road user behavior during chimpanzee crossings on a busy paved road at Bulindi, western Uganda, to inform chimpanzee conservation measures in increasingly anthropized African landscapes undergoing road development. We sought to identify the factors influencing the behavior of road users, particularly their level of caution in response to the potential physical risk posed by chimpanzees (i.e. risk of collision or perceived risk of aggression), as well as their curiosity upon seeing this charismatic great ape species. Our predictions are as follows:

Road user vulnerability: Pedestrians and cyclists (defined as ‘vulnerable road users’) are expected to behave more cautiously (e.g., stopping, retreating) than motorists, as they are more physically exposed. Among pedestrians, women and children are likely to be more cautious than men, given local knowledge that male chimpanzees may threaten these groups (McLennan and Hill, 2010, McLennan and Hill, 2012; McLennan and Hockings, 2016). Additionally, individuals traveling alone should be more cautious than those in groups, as the presence of others can reduce perceived vulnerability and increase risk-taking (Stoner, 1961; Dion et al., 1970; Tedeschi et al., 2021).

Chimpanzee visibility: Road users are expected to respond more strongly–either with caution or curiosity–when chimpanzees are on the road surface rather than at the roadside, due to higher visibility and perceived threat. Similarly, larger groups of chimpanzees are likely to attract more attention or caution than smaller ones.

Contextual influences: Road users are expected to show greater caution and curiosity when local field assistants–known locally as ‘Chimpanzee Monitors’–interact with them during chimpanzee crossings. Such interactions may increase awareness of the animals and draw attention to the crossing event. Motorists are also expected to be more cautious and curious near speed humps, where lower vehicle speeds provide more time to observe and react to chimpanzees.

Study area

The study was conducted in Bulindi (1°28′N, 31°28′E), which is a town council located in Hoima District, western Uganda, between the Budongo and Bugoma Central Forest Reserves (Figure 1). Since the 1990s, most unprotected forest outside these reserves has been converted into farmland (McLennan, 2008; Mwavu and Witkowski, 2008; McLennan et al., 2020) and the human population increased from 156 individuals per km 2 in 2014 [Uganda Bureau of Statistics (UBOS), 2016] to 192 per km 2 in 2024 (Uganda Bureau of Statistics, 2024)). Despite these extensive anthropogenic disturbances, a regional population of over 260 chimpanzees, residing in more than 10 social groups (‘communities’), occupies small remnants of unprotected forest amidst farmland and villages outside the main forest blocks (McLennan, 2008; McLennan and Plumptre, 2012; McCarthy et al., 2015). Bulindi hosts one such community.

Agriculture is the primary economic activity regionally, with tobacco, maize, cocoa and sugarcane among the main cash crops (Mwavu and Witkowski, 2008; McLennan and Hill, 2015). In Bulindi, forest clearance peaked between 2006 and 2014, when over 80% of forest used by the chimpanzees within their 20 km 2 home range was cleared entirely for farming (McLennan et al., 2020). This severe habitat loss affected the chimpanzees’ behaviour and ecology (e.g., feeding and nesting: McLennan et al., 2020; van Dijk et al., 2021) and increased their interactions with humans, as they now travel and forage daily in and around village gardens and homesteads (McLennan and Hill, 2012; Satsias et al., 2022). While local Banyoro culture forbids primate consumption, human–chimpanzee conflict is frequent due to habitual crop foraging by the apes (McLennan, 2013; McLennan et al., 2020). Many residents view the chimpanzees not as kin, food, or endangered animals, but as troublesome wild animals that threaten livelihoods and safety (McLennan and Hill, 2012). Indeed, chimpanzee attacks on humans, especially children, have occurred widely in the region, including in Bulindi (Reynolds, 2005; McLennan, 2008; McLennan and Hockings, 2016; unpubl. data). Even so, many residents acknowledge that chimpanzees exhibit human-like intelligence and are interesting to watch (McLennan and Hill, 2012).

At the time of this study, the combined area of forest fragments within the chimpanzees’ 5 km 2 core area (i.e., the most commonly used portion of their home range) was approximately 0.4 km 2. Remaining forest patches in Bulindi are interspersed with houses and agricultural gardens on either side of a 10.5-meter-wide road (Figure 1). This major road connects the two expanding urban centers of Hoima (141,442 inhabitants in 2024; Uganda Bureau of Statistics, 2024) and Masindi, and is part of a 238 km road corridor in western Uganda, linking traffic from Rwanda, Burundi, Democratic Republic of Congo, and Sudan (African Development Bank, 2012). The road carries heavy vehicle traffic including motorbikes, cars, buses, trucks, agricultural vehicles, as well as cyclists and pedestrians. It is therefore used by a wide variety of road users, including local residents and long-distance travelers. Notably, traffic volume has increased recently (McLennan and Asiimwe, 2016), and the road was widened and paved in 2018 to support the oil and gas industry (Mawejje, 2018; McLennan et al., 2021). During our study, average daytime traffic volume per minute along the section of road crossed by the chimpanzees (see below) was 0.8 pedestrians, 0.2 cyclists, and 2.4 motor vehicles, of which the majority were motorbikes (Tellier et al., 2025). Despite the presence of pedestrians and traffic signs, the asphalt surface allows vehicles to reach speeds of up to 120 km/h (Tellier et al., 2025). Nevertheless, Bulindi’s chimpanzees cross this road frequently (i.e., at least once in a week during this study; Tellier et al., 2025), and a fatal collision wherein an adult female was knocked dead by a speeding taxi occurred in 2015 before upgrades (McLennan and Asiimwe, 2016). In response to the risks faced by both chimpanzees and road users, two speed humps were installed at key chimpanzee crossing points (Figure 2). Their construction began in late-2018, at the start of our study, and was completed by February 2020.

Data collection

Interactions between road users and chimpanzees were documented along a one-kilometer stretch of the Hoima-Masindi road, where the chimpanzees frequently cross (McLennan and Asiimwe, 2016; Tellier et al., 2025). For analysis, this portion of the road was split into two sections with differing levels of danger for both chimpanzees and road users (Figure 2):

Zone A (200 m): Between two speed humps, considered safer due to traffic slowing, but dense vegetation on both sides limits driver visibility. Chimpanzees cross here most often (Tellier et al., 2025).

Zone B (730 m): Higher risk with faster vehicles and limited visibility due to bends and an absence of humps; vegetation cover at roadsides is lower, mostly bushes or open croplands.

The Bulindi chimpanzees are habituated to researchers, with most individuals tolerating observations within 10 meters (Satsias et al., 2022). Habituation was achieved gradually during 2012–2015 through repeated, neutral, and predictable researcher presence in the chimpanzees’ habitat (McLennan et al., 2019). During the 38 months of our study, from November 2018 to December 2021, local field assistants JMB and TS monitored the activities of Bulindi chimpanzees for 5 to 7 days each week, between 0700 and 1900 h. During this period, chimpanzee community size averaged 17 individuals, excluding infants (0 to 3 years old). All observations made for this study were based on opportunistic video recordings made by JMB, with some alternate recordings filmed by MRM, using handheld digital cameras. When a road crossing seemed imminent, as indicated by the direction of chimpanzees’ travel towards the road, where possible they moved ahead of the chimpanzees to position themselves at likely crossing points to film the crossing. Some recordings were incomplete because researchers did not always arrive at the road before the crossing began or because chimpanzees crossed at different locations (Tellier et al., 2025). Additionally, because the chimpanzees frequently split into subgroups (Satsias et al., 2022), the field assistants often followed only a portion of the community at a time. For these reasons, many crossings were certainly missed. Therefore, our dataset represents a sample of road crossings that occurred during the study period. In some videos, field assistants intervened by signaling motorists to slow down or asking pedestrians to wait so that the chimpanzees could cross (Figure 3). For analysis, we retained only videos in which at least one road user was within 50 meters of chimpanzees (i.e. close enough to be clearly visible to chimpanzees), in which a reaction from the road user was likely to occur.

A “road-crossing event” was defined as the observation of at least one chimpanzee crossing the road (Cibot et al., 2015; Tellier et al., 2025), usually after waiting at the roadside. It was considered a single crossing event when two or more chimpanzees crossed the road in the same direction, on the same stretch of road, within 30 minutes. The “number of chimpanzees visible to road users” was assimilated to the number of individuals in the crossing group, excluding infants, whether they were waiting at the roadside or crossing the road, when the road user was closest to them. Infants were excluded from this count since they are carried by their mothers during crossings and are therefore less visible to road users. For each crossing event where proximity between chimpanzees and at least one road user could be estimated, we noted the date, hour, zone, number of chimpanzees, and whether field assistants actively interacted with road users during the event. Interactions were defined as any observable or audible action directed at road users, such as conversing with them, vocalizing or gesturing in their direction (e.g., to request that they wait while chimpanzees cross), or adopting a guarding posture on the road in order to be visible to approaching road users (Figure 3). Events without such behaviors were classified as having no interaction. The crossing time and composition of chimpanzee groups were corroborated by the field assistants’ written notes. We also recorded the total amount of motor vehicles (one road user per motor vehicle, i.e. the driver) and non-motorized road users (i.e. cyclists and pedestrians) visible during the video and their closest respective distance from the chimpanzees [close proximity–within 50 m; or distant proximity–greater than 50 m; this distance threshold follows studies of driver reactions to pedestrians (Katz et al., 1975; Várhelyi, 1998)]. Traffic counts included both visible road users and those inferred from audio cues. Specifically, individuals audible to researchers but outside the camera’s field of view were included to ensure a more comprehensive assessment of road user presence. For these audio-identified users, we estimated the number of vehicles or people based on the number of distinct voices or engine sounds; when in doubt, a default value of one was assigned.

We noted how each road user reacted to the presence of chimpanzees on or by the side of the road, classifying each response along two axes: caution (cautious vs. reckless, with respect to collision risk or perceived risk of aggression from the chimpanzees) and interest (curious vs. indifferent). This approach enabled us to develop a novel ethogram of human behavior during such interactions (Table 1). In addition, we noted the size of the chimpanzee group involved in each encounter and whether they were at the roadside or crossing. Finally, we categorized vehicles as motorbikes, mid-sized vehicles (including cars, taxis and vans) or larger trucks (including mini-buses, buses, trucks and agricultural vehicles). For vulnerable road users (i.e. cyclists and pedestrians) in close proximity to chimpanzees, we identified their age-sex category (women, men, or children–including young teenagers). Individuals traveling in groups (≥2 persons; Pellegrini et al., 2010; Qiu and Hu, 2010) were treated as a single behavioral unit (Bose et al., 2023).

Statistical analysis

First, we used χ 2 tests to compare the observed versus expected proportions of each road user category in close proximity with chimpanzees. This analysis aimed to determine whether the category of road user significantly influences the likelihood of chimpanzees crossing roads or approaching traffic.

Second, road-user responses were classified according to the specific behaviors described in Table 1. For each encounter, the observed behavior was coded using two binary variables: a caution score (cautious = 1; reckless = 0) and an interest score (interest displayed=1; no apparent interest = 0). Binomial generalized linear mixed model (binomial GLMMs) were used to assess associations between reckless or interest-driven behavior with intrinsic (road user category, age-sex category for pedestrians and cyclists, vehicle category for motorists) and extrinsic (road zone, chimpanzees location, number of chimpanzees visible to road users, interaction with field assistant) variables while taking into account the potential effect of the recorded video. Indeed, as multiple crossing events may be present for the same video, the video was treated as a random factor. We used the lmer function in the R package lme4 (Bates et al., 2015). All models presented here converged successfully and were checked for overdispersion using the R package DHARMa; only models without significant overdispersion were retained for analysis.

Our predictions concerned Road user vulnerability (1), Chimpanzee visibility (2) and Contextual influences (3) and how these different aspects can influence reckless and interest-driven behaviors. We specifically tested these predictions with the following variables:

We tested the effect of the road user category (pedestrians and cyclists versus motorists). Also, among pedestrians, we examined the effect of age and sex by dividing them into three categories: women (alone or in group, with or without children), men (same definition), and children traveling alone. This classification assumes that children follow the behavior of accompanying adults (Gergely et al., 2002; Rakoczy et al., 2010; Li et al., 2013). The number of mixed pedestrian groups composed of men and women (n = 9) was too small for meaningful statistical analysis. Third, we examined the effect of being in a group or not, distinguishing individuals traveling alone from those traveling in groups (i.e. ≥2 persons), regardless of age-sex category. Fourth, we tested the influence of the vehicle category for motorists (motorbike, mid-size vehicle such as a car or taxi or larger truck). In this analysis, mid-size vehicles and trucks were grouped since their drivers are physically protected and they potentially cover greater distances than motorcyclists (i.e. they are more likely to be non-locals; Yannis et al., 2007; Hu et al., 2018).

We tested the effect of chimpanzees’ location (roadside or road surface). We also evaluated the effect of the number of chimpanzees visible to road users (small group from 1 to 6 individuals; medium group from 7 to 12; large group from 13 to 18).

We tested the effect of observable or audible interactions with field assistants (absence or presence). We also tested the effect of the road zone (A or B).

Data analysis was carried out using R (version 4.0.3; R Core Team, 2022). The significance threshold was set at P< 0.05.

Dataset and distribution of road user types

Among all road users recorded in the full set of 209 videos, regardless of their proximity to chimpanzees, motorists represented 78% of observations (= 936/1198; Figure 4), including 63% motorcyclists (= 5 92/936). Among vulnerable road users (n = 262), pedestrians accounted for 77% of observations (= 202/262), while cyclists represented 23% (= 60/262).

For subsequent analyses, we retained 147 videos containing at least one instance of chimpanzees in close proximity to motorists, pedestrians or cyclists. These videos spanned 108 different days and documented 119 distinct road crossing events. The sample represents 5 hours and 57 minutes of footage, with video durations ranging from 14 seconds to 16 minutes (mean ± SD: 2.43 ± 2.60 minutes).

Within this subset, we identified 376 instances of close proximity between a chimpanzee and a road user (Figure 4). Of these, 63% (= 238/376) occurred during active chimpanzee road crossings (Figure 5), while the remaining 37% involved chimpanzees at the roadside, varying in visibility to road users. Motorists were involved in 69% of close-proximity interactions (= 261/376), a proportion significantly lower than their representation in the full video dataset (χ 2 1 test for given probabilities: χ 2 1 = 16.15, P < 0.0001). Among motorists, motorcyclists accounted for 71% of interactions (= 186/261), which was significantly higher than expected based on their overall representation (χ 2 1 = 7.65, P = 0.006). Among vulnerable road users, pedestrians represented 85% of cases (= 98/115), whereas cyclists accounted for 15% (= 17/115). This distribution also differed significantly from the proportions observed in the full dataset (χ 2 1 = 4.38, P = 0.04). All recorded cyclists were male. Regarding pedestrians, we documented 43 instances involving men or groups of men, 34 instances involving women or groups of women, 9 instances featuring mixed groups comprising at least one man and one woman, and 12 instances involving children only. Overall, 36% of interactions involving vulnerable road users (= 41/115) concerned groups of people rather than lone individuals.

Road user vulnerability

Vulnerable road users were significantly more cautious than motorists when in close proximity to chimpanzees (Z value = 6.28, P < 0.0001; Figure 6a). The predicted probability of being cautious is 84% for vulnerable road users and 44% for motorists. Vulnerable road users also displayed significantly greater interest in chimpanzees than motorists (predicted probabilities = 79% vs. 47%, respectively; Z value = 5.08, P < 0.0001; Figure 6a). However, in 88% of cases where individuals were observed filming chimpanzees (= 7/8 instances), they were either passengers or drivers of vehicles which had stopped to observe the apes. In one instance, we observed a truck driver offering bottled water to a departing chimpanzee. Among vulnerable road users, responses did not differ significantly between men, women, and children, with predicted probabilities ranging from 75–88% for cautious behaviour and from 72–75% for interest (Figure 6b). However, vulnerable road users were significantly more cautious when traveling in groups rather than alone (predicted probabilities = 93% vs. 76%, respectively; Z value = 2.13, P = 0.03; Figure 6c), but we found no difference in interest between groups and lone individuals (predicted probabilities = 80% vs. 72%, respectively; Z value = 1.04, P = 0.3). Among motorists, we found no significant difference in responses between motorcyclists and drivers of mid-size vehicles and trucks (caution: Z value = -1.16, P = 0.25; interest: Z value = -0.76, P = 0.45; Figure 6d).

Chimpanzee visibility

Overall, road users were significantly more cautious and showed more interest in chimpanzees that were crossing the road versus waiting at the roadside (caution: predicted probabilities = 73% vs. 25%, respectively; Z value = -7.28, P < 0.0001; interest: 77% vs. 17%; Z values = -8.08, P < 0.0001; Figure 6e). However, road users were significantly less cautious when faced with large crossing groups of chimpanzees compared to smaller groups [predicted probabilities: 29% for large groups, 55% for small groups (Z value = 2.29, P = 0.022), and 66% for medium-sized groups (Z value = 2.96, P = 0.003); Figure 6f]. By contrast, the level of interest shown by road users did not differ significantly according to chimpanzee group size, with predicted probabilities of 41%, 58%, and 59% for large, medium-sized, and small groups, respectively.

Contextual influences

Of all instances of close proximity between a road user and at least one chimpanzee, 35 included interactions between a field assistant and road users (Figure 3). Road users were more likely to exhibit caution and interest in the chimpanzees when there was interaction between the road user and a field assistant (caution: predicted probabilities are 76% when there is an interaction and 55% when there is no interaction; Z value = 2.29, P = 0.02; interest: predicted probabilities = 90% vs. 54%, respectively; Z value = 3.66, P = 0.0003; Figure 6g). However, the road zone (A or B) did not influence the responses of motorists (predicted probabilities are 40-51%; caution: Z value = -1.13, P = 0.26, interest: Z value = -1.31, P = 0.19) or road users in general (predicted probabilities are 52-61%; caution: Z value = -1.21, P = 0.23, interest: Z value = -1.26, P = 0.21; Figure 6h).

In this novel study we revealed broad behavioral patterns of road users in response to wild chimpanzee road crossings in Uganda.

Which road users do chimpanzees encounter?

Among all road users captured in the 209 videos, vulnerable road users (pedestrians and cyclists) were more likely to be in close proximity to the chimpanzees compared to their overall representation in the dataset. This may indicate that chimpanzees are more inclined to cross near vulnerable road users than in front of approaching vehicles. Our findings also suggest that chimpanzees are most likely to avoid larger motor vehicles on the road, especially cars and trucks, compared to motorcyclists, presumably reflecting their perception of the risk posed by large motor traffic (Tellier et al., 2025). The recklessness of some motorcyclists is therefore a significant risk factor for collisions, since chimpanzees are often willing to cross in proximity to motorcycles.

Does vulnerability influence road user behavior?

As predicted, vulnerable road users were more cautious than motorists when encountering chimpanzees. This likely reflects their greater physical exposure and concerns about personal safety compared to drivers. Vulnerable road users may also be more familiar with the risks posed by chimpanzees than drivers of larger vehicles. The Hoima–Masindi road is part of a 238 km international corridor linking Uganda to neighboring countries (African Development Bank, 2012), and it therefore carries substantial through-traffic, including long-distance and cross-border drivers who are less likely to be local residents and may have limited prior exposure to chimpanzees. In contrast, pedestrians, cyclists, and motorcyclists are more likely to be residents of the local area. However, we did not identify any significant behavioral differences between motorcyclists and drivers of mid-size vehicles and trucks. Local road users are generally acquainted with chimpanzees due to their well-known presence in the area (McLennan and Hill, 2010, McLennan and Hill, 2012) and associated community conservation initiatives (Bersacola et al., 2023). Cases of chimpanzee attacks on humans in western Uganda–including in Bulindi and nearby areas (McLennan and Hockings, 2016)–have been well-documented by local and national media (e.g., Monitor, 2021), which may explain why locals exhibit more caution. Alternatively, some local road users may show less caution due to frequent exposure to chimpanzees, which can lead to over-familiarity and a diminished perception of risk (Kaltenborn et al., 2006; Ramp et al., 2016; Pyrialakou et al., 2020). We might also expect them to be less curious about the chimpanzees than non-local individuals, owing to their greater familiarity, and indeed we observed that people stopping to film the chimpanzees on the road were mainly motorists. Overall, however, pedestrians and cyclists showed more interest in the chimpanzees compared to motorists, perhaps because their slower pace on the road afforded them more time to observe the animals.

Contrary to prediction, women or children were not more cautious around chimpanzees than men. While vehicle drivers showed least caution, vulnerable road users may not be sufficiently careful when encountering chimpanzees. The presence of field assistants at the roadside may make people feel safer (Golkar et al., 2013; Qi et al., 2020), potentially masking differences in caution among women, children and men. Again, this could also be because most pedestrians and cyclists have likely encountered chimpanzees on roads before, and this familiarity may have reduced their perception of risk. It is also possible that women and children maintained greater distances from chimpanzees (within the 50 m threshold) than men did. However, we could not assess this possibility because the video recordings lacked fixed spatial reference points or measurement markers needed to estimate short distances accurately. Obtaining precise measurements for each instance of proximity would have required additional resources and personnel in the field. Future research should investigate how local people perceive encounters with the chimpanzees on roads (e.g., whether or not they feel at risk), to better understand pedestrian behavior in our study. It is also worth noting that sample size for children was relatively small, which may have limited statistical power to detect behavioral differences among pedestrian categories.

Besides, contrary to our expectations, vulnerable road users exhibited more cautious behavior when traveling in groups rather than alone. We speculate that individuals traveling alone may suppress visible reactions such as caution because they are aware of being observed and filmed by the field assistants, and therefore continue on their way. Previous research has shown that surveillance can influence human behavior (Risko and Kingstone, 2011; Jansen et al., 2018; see below).

Does chimpanzee visibility influence road user behavior?

As predicted, the location of the chimpanzees had a significant impact on road users’ behavior. People exhibited more caution and interest towards them when they were crossing the road as opposed to waiting at the roadside–presumably because they were most visible (and active) when moving on the road surface. However, chimpanzees that were concealed at the roadside behind bushes or trees, could appear abruptly on the road, running or walking quickly, increasing the risk of sudden close encounters with vulnerable road users or collisions with motorists.

Contrary to expectations, road users were less cautious when faced with a crossing of a larger group of chimpanzees; motorists tended to continue without slowing down, while vulnerable road users at times continued walking past the chimpanzees or directly towards them. We suggest this reflects a risk-avoidance strategy prioritising speed to limit exposure: road users may want to avoid large chimpanzee groups as quickly as possible. However, this approach paradoxically heightens collision and attack risks by curtailing reaction time and safety margins.

Although no collisions occurred during the study, we recorded seven near-miss incidents involving chimpanzees (four with motorcycles, three with mid-sized vehicles). In each of these cases, we qualitatively noted road bends, slopes, or dense vegetation that may have reduced drivers’ visibility. Although not sufficient to infer causality, these anecdotal observations are broadly consistent with our statistical findings identifying poor visibility as a key risk factor (see also McLennan and Asiimwe, 2016). Most near-collisions involved ‘reckless’ driving (= 6/7), with drivers failing to slow down despite curves, visible chimpanzees, or clear warnings from field assistants.

Do field assistants and road features influence road user behavior?

As expected, road users exhibited more caution and more interest in the chimpanzees when they engaged with a field assistant, whether for a conversation about the chimpanzees or when advised by an assistant to slow down, wait or proceed. In our study, field assistants JMB and TS are a familiar sight on the road, and many local residents know them personally. Interactions with familiar field assistants may increase compliance with road safety guidance, heighten attention to risk (McGuire, 1985), and encourage greater interest in the chimpanzees. Interactions with road users were potentially more frequent in situations perceived by the field assistants as risky, prompting them to intervene. Consequently, some behavioral responses observed may reflect the circumstances that prompted an intervention rather than the interaction itself. However, we did not examine the context of interactions in this study specifically. Moreover, it remains unclear how the mere presence of field assistants influences road users’ behavior. Additional studies, such as self-report questionnaires or using remote cameras, could help assess this effect. In practice, however, deploying remote cameras would be challenging, as chimpanzee crossings are unpredictable and roadside equipment would be vulnerable to theft and damage. Again, we should consider how being filmed affects road users’ reactions. It is likely that some road users, especially pedestrians and cyclists who are not concealed within a vehicle, would behave differently when field assistants are not present. For instance, awareness of being filmed probably encourages some individuals to hide their fear of chimpanzees and to act with less caution or show less interest (Risko and Kingstone, 2011; Jansen et al., 2018). Nevertheless, our findings demonstrate that road users are responsive to the expertise and interventions of external agents.

Against our expectations, motorists’ reactions did not differ between the two crossing zones, despite the presence of speed humps in zone A that necessitated slower vehicle speeds. We suggest that dense vegetation coverage in zone A likely reduced visibility (Figure 2), concealing chimpanzees near the road and affecting drivers’ ability to anticipate crossings, potentially biasing our analysis.

Conservation and management insights

Our findings identify several factors influencing road-user behavior during chimpanzee road crossings. While we did not evaluate mitigation measures in the present study, the patterns observed suggest a number of management interventions that could be tested in future studies.

First, our results show that motorists’ responses did not differ between the two crossing zones, despite the presence of speed humps in one section. This suggests that traffic-calming infrastructure alone may be insufficient to consistently influence driver behavior in chimpanzee crossing areas. More targeted or reinforced speed management strategies may therefore be required, particularly in low-visibility locations to reduce the risk of collisions (McLennan and Asiimwe, 2016).

Road design and visibility also appear to play an important role in shaping encounter risk. Road users were more cautious when chimpanzees were actively crossing than when they were at the roadside, indicating that detectability strongly influences responses. Near-miss events further supports the idea that delayed detection of chimpanzees contributes to risky situations. Improving early visibility, for example by clearing roadsides of dense vegetation cover, could allow drivers to better anticipate wildlife crossings (Rea, 2003; Milton et al., 2015). However, such interventions must be balanced against potential ecological consequences, as reduced cover might inhibit some individuals from crossing the road (Tellier et al., 2025). Applied to female chimpanzees especially, this measure might limit ranging and migration during adolescence. Another option would be to build underground passages or large vegetated overpasses (Huijser and McGowen, 2010; Plumptre et al., 2010; Sawaya et al., 2013; Smith et al., 2015; Huijser et al., 2016) where chimpanzees cross roads most frequently, although costs are high ($0.5–15 million per installation; Sijtsma et al., 2020; Brennan et al., 2022) and use by the animals is uncertain (but see Chan et al., 2020; Maria et al., 2022). Pilot trials in frequently used crossing areas such as Bulindi would be needed to evaluate their feasibility.

Our results also highlight the influence of direct interactions with locally employed field assistants on road-user behavior. This effect may partly reflect the trust they have established within the local community. As familiar and respected figures, they may be more successful at influencing road-user behavior than impersonal traffic signs or unfamiliar authorities. Installing traffic controls along the road in Bulindi, and on road segments crossed by chimpanzees elsewhere, may therefore be most effective when combined with the involvement of trusted local personnel.

Moreover, our findings have implications for targeted awareness strategies. Chimpanzees were more likely to cross near pedestrians, cyclists, and motorcyclists than larger vehicles, indicating that different road-user groups experience different levels of exposure risk. Among pedestrians, we found no behavioral differences between women, children, and men, suggesting broadly similar responses within this group. Vulnerable road users were overall more cautious and showed greater interest in chimpanzees than motorists. These patterns suggest that awareness campaigns should be tailored to specific road-user groups, particularly targeting non-local drivers who may have limited prior exposure to chimpanzees, as well as motorcyclists who are frequently involved in close encounters.

Broadcasting radio messages and installing signs and traffic controls would be beneficial (Jägerbrand and Antonson, 2016; Jägerbrand et al., 2018), especially to inform non-local road users of the presence of chimpanzees. However, signage alone is unlikely to effectively reduce vehicle speeds or wildlife-vehicle collisions (Dumalakas et al., 2025). In addition, signs highlighting chimpanzee presence along roads used by tour companies, such as the Hoima-Masindi road, may inadvertently encourage tourists to stop along the road, potentially discouraging chimpanzees from crossing or prompting tourists to trespass on private land in an attempt to ‘track’ chimpanzees, as has occurred in Bulindi (pers. obs.).

In human–wildlife coexistence areas such as Bulindi, education campaigns for local people about chimpanzee crossing behavior might also contribute to safer interactions. For example, residents could be informed that chimpanzees tend to cross in groups, one after the other, but with unpredictable intervals between individuals (Cibot et al., 2015; Tellier et al., 2025). Therefore, road users should anticipate that additional individuals may follow after the first chimpanzee(s) seen crossing, and therefore it is advisable to wait before proceeding. Children should be discouraged from traveling alone on roads crossed frequently by chimpanzees, given their potential vulnerability to aggression from these great apes (Hockings et al., 2010; McLennan and Hockings, 2016).

Additionally, as with some other primates (e.g., Maille et al., 2017; Nabuule et al., 2024), road users should be warned about the risks of feeding chimpanzees. Although we observed only one occurrence of this kind of behavior in our study, we recognise it as a growing risk in areas frequented by tourists and in human-wildlife cohabitation zones (Pragatheesh, 2011; Ilham, 2024). Given their behavioral flexibility and propensity to eat cultivated foods (Hockings and McLennan, 2012; McLennan et al., 2020), such interactions could potentially occur with chimpanzees and pose numerous risks such as aggression, pathogen transmission, or traffic accidents.

Lastly, since chimpanzees are considered a flagship and umbrella species for conservation (Plumptre et al., 2010; Albert et al., 2018), protective measures for chimpanzees on roads would likely benefit many other species. In Bulindi, roadkill of colobus monkeys (Colobus guereza), vervets (Chlorocebus tantalus), various snakes and a pangolin (Manis sp.) have been recorded along the same road (pers. obs.). At Sebitoli in Kibale National Park, the road that traverses the chimpanzee home range is also used by numerous species of reptiles, amphibians, birds, and mammals (Cibot et al., 2015).

This study examines how road users respond to wild chimpanzee crossings in an human-dominated landscape in Uganda. Our findings show that road-user behavior is shaped by physical vulnerability, visibility conditions, traffic context, and direct interactions with local field assistants on the road, with clear implications for road planning in wildlife habitats. They suggest that traffic-calming measures alone are insufficient; effective mitigation should combine infrastructure that improves early detection of wildlife with measures that reduce sudden encounters. Our results also underscore the importance of integrating locally embedded personnel into road safety strategies. More broadly, differences among road-user types indicate that mitigation and awareness efforts should be tailored to exposure risk. Overall, the study highlights the need for traffic calming, improved visibility, socially grounded interventions, and community education to enhance safety and conservation in shared human-wildlife environments.