Behavioral response of Bornean ungulates, including bearded pigs and sambar deer, to anthropogenic disturbance in Sabah, Malaysia

doi:10.21203/rs.3.rs-4279935/v1

Understanding wildlife behavioral responses is crucial for assessing the effects of anthropogenic disturbance. We used camera traps to investigate the behavioral responses of two ungulate species, bearded pigs (Sus barbatus) and sambar deer (Rusa unicolor), to anthropogenic disturbance in three protected areas in Sabah, Malaysia, that have varying levels of human activity. We found that human activities generally influence the activity patterns of both ungulates, albeit with variations among the sites. The temporal activity pattern of bearded pigs was affected by anthropogenic disturbance, especially in the area targeted by poachers. While the core activity pattern of sambar deer remained consistent across sites, poaching pressure appeared to impact their behavior within specific environments. Bearded pigs approached plantations at times of low human activity, presumably to forage, indicating that they adjust spatiotemporal activity patterns to minimize human contact. We observed a reduction in active times for both species at sites of high anthropogenic disturbance. Despite these challenges, both species demonstrated behavioral adaptability to anthropogenic disturbance by utilizing artificial environments such as roads and oil palm plantations as foraging places, thereby potentially compensating for reduced feeding times. Our study underscores the negative impact of human activities on the activity patterns of the two ungulate species. Nevertheless, it also highlights their behavioral plasticity in response to anthropogenic disturbance, suggesting their ability to efficiently utilize alternative food resources. Our methodology provides insights into wildlife management strategies. We recommend urgent long-term monitoring of wildlife population dynamics, including behavioral responses, especially in Southeast Asia.

Diel activity pattern

Human activity

Camera trapping

Spatiotemporal activity

Hunting

Since prehistoric times, human activities have transformed the environment, although the most marked environmental changes caused by humans, such as air pollution, climate change, forest decline, and loss of species diversity, have occurred during the past 300 years (Andermann et al. 2020; Goudie 2013). Loss of species diversity is a serious environmental issue, given the functional roles of every organism in each ecosystem (Ceballos et al. 2020; Srivastava, et al. 2012). Mammals are a useful indicator of global biodiversity loss because of their diverse ecological niches (Ceballos and Ehrlich 2002). Of 173 mammalian species examined, more than 70% lost their geographic ranges in the past 200 years, mainly due to anthropogenic disturbances (Ceballos and Ehrlich 2002). At least 85 mammalian species have gone extinct since 1500 AD (IUCN 2024). Human activities, including overhunting and habitat destruction, have immense negative impacts on wildlife (Andermann et al. 2020; Brodie et al. 2015; Tilman et al. 2017).

Among the various human activities, hunting directly affects wildlife abundance, and some mammals change their ecological characteristics in response to human activities (Stankowich 2008; Tilman et al. 2017). Unregulated hunting is a major cause of population decline among wildlife (Schipper et al. 2008; Tilman et al. 2017). Globally, most species at threat from hunting are in Africa and Southeast Asia (Ripple et al. 2016). The extinction risk of terrestrial mammals particularly high in Southeast Asia (Ceballos and Ehrlich 2002; Schipper et al. 2008; Tilman et al. 2017), where regional population decline of most large species has occurred within the past 50 to 100 years (Corlett 2007). On the island of Borneo, the remarkable level of wildlife habitat loss has mainly been caused by rapid deforestation and the conversion of forests to monocultural plantations (Gaveau et al. 2016; Ocampo-Peñuela et al. 2020). However, hunting rather than logging is the main threat to medium- to large-bodied mammals (Brodie et al. 2015). Megafauna (body mass >44 kg) inhabiting Borneo, including orangutans, bearded pigs, sambar deer, and banteng, are particularly vulnerable to human impact and habitat change (Corlett 2010). Given that hunting/poaching pressures are not always fully controlled, even within protected areas (Corlett 2007; Harrison et al. 2016), and because hunting has direct (abundance) and indirect (behavior) impacts on wildlife (Kays et al. 2016; Ripple et al. 2016), monitoring of wildlife populations and their behavior are necessary to assess and control hunting/poaching in a given area.

Wildlife often shows physiological and behavioral responses to anthropogenic stimuli; such responses can be broadly categorized as attraction, avoidance, and habituation (Bejder et al. 2009; Whittaker and Knight 1998). Even non-harmful encounters, such as during outdoor recreational activities, can cause behavioral changes in wildlife (Bejder et al. 2009; Tablado and Jenni 2017). These include changes in spatial distribution, temporal activity patterns, feeding behavior, reproduction, and social structure (Bejder et al. 2009). Adjusting spatial and temporal activity patterns is a common response of wildlife to avoid encountering humans (Ciuti et al. 2012; Gaynor et al. 2018; Nix et al. 2018). The fear response of some wildlife to humans exceeds that to natural predators (Ciuti et al. 2012; Visscher et al. 2023). Thus, to evaluate hunting pressure, it is important to understand wildlife behavioral responses that represent a trade-off between the risk of being hunted versus the requirement for resources such as food and water (Crosmary et al. 2012).

Ungulates often show apparent vigilance for hunting risks (Ciuti et al. 2012; Cromsigt et al. 2013; Stankowich 2008). They change temporal and spatial land-use patterns in response to human disturbance depending on various factors, including their sex, the season, the presence of offspring, the level of exposure to humans, and their physical condition (Stankowich 2008). An increase in nocturnality is a key behavioral response to avoid humans (Gaynor et al. 2018). Nocturnal activity patterns of some ungulate species are affected by the lunar phase (Brivioa et al. 2017; Colino-Rabanal et al. 2018). Ungulate responses to hunting pressure could be behavioral adaptations because they are hunted globally for food, materials, sport, and population management (Cromsigt et al. 2013; Pascual-Rico et al. 2021). In Southeast Asia, they have long been targeted by hunters, and many ungulate species are now threatened due to overhunting and habitat loss (Corlett 2007, 2010). In Malaysian Borneo, ungulates including bearded pigs (Sus barbatus) and sambar deer (Rusa unicolor) account for up to 80% of hunted animals (Bennett et al. 2000; Saikim et al. 2023; Yi and Mohd-Azlan 2020). Given that the behavioral responses of ungulates show intraspecific variation and differ according to the level of anthropogenic disturbance (Stankowich 2008), it is important to assess behavioral responses of different populations of ungulates to human stimuli, to evaluate anthropogenic disturbance and the management of protected areas. The increasing use of camera traps has enabled many studies of wildlife populations and their size in Southeast Asia. However, relatively little attention has been paid to assessing the effect of anthropogenic disturbance on animal behavior, despite knowledge of behavior being central to wildlife management, along with long-term monitoring of wildlife population dynamics (Kays et al. 2016; Crosmary et al. 2012).

Here, we investigated behavioral responses of two ungulate species, bearded pigs and sambar deer, to anthropogenic disturbance in three protected areas under different levels of disturbance in Sabah, Malaysia. We compared their diel activity patterns and the effect of artificial environments such as oil palm plantations and gravel roads on their spatiotemporal movements, using camera trap data obtained over a three-year period. We also assessed the effects of the lunar phase on their nocturnal activities in relation to their approach to artificial environments.

Study sites
We selected three protected areas in Sabah as study sites: Danum Valley Conservation Area (DVCA), Lower Kinabatangan Wildlife Sanctuary (LKWS), and Tabin Wildlife Reserve (TWR), where the poaching frequency and target species differed. All sites are protected areas and ecotourism is officially operated. Hereafter, we define any hunting activities in the study sites as poaching.

DVCA (4° 50'–5°05' N, 117°30'–117°48' E) is a forest reserve covering 438 km². Most of the area comprises mature lowland evergreen dipterocarp forest (Marsh and Greer 1992). The study area consists of old-growth forest surrounding a tourist lodging facility (5°01' N, 117°44'E). A gravel road leads to the facility (Figure 1).

LKWS (5° 10’–5° 50' N, 117° 40' –118° 30' E) is located along the Kinabatangan River, which reaches 560 km inland. LKWS comprises 10 forest blocks totaling an area of 270 km², including seasonal and tidal swamp forests, permanent freshwater swamps, mangrove forests, and lowland dipterocarp forests (Abram et al. 2014; Goossens et al. 2005). The southern area of the study site is extensively covered by secondary forest. The northern area has been deforested for oil palm plantations, except for a protected zone along the river (Figure 1).

TWR (5°05'–5°22'N, 118°30'–118°55'E) covers approximately 1,225 km². It is entirely surrounded by large oil palm plantations. Most of TWR was heavily logged in the 1970s and 1980s, leaving mainly regenerating mixed dipterocarp tropical rainforest (Mitchell 1994). The study area was located on the western boundary of TWR (5°11'N, 118° 30'E) (Figure 1). The main road on the western border of TWR is frequently used by vehicles, including the public transportations.

There was little to no poaching around the study area in DVCA (Hearn et al. 2017; Wong et al. 2004), due to various geographical and administrative factors, such as being 50 km from the nearest village and having just one access road with a government-controlled access gate. Conversely, poaching has been reported in LKWS (Hearn et al. 2017; Love et al. 2017); even within protected areas, ungulate population density, especially sambar deer, may occasionally be affected by hunting (Matsuda et al. 2015). However, as local communities are of Muslim faith, they do not consume bearded pigs (Kurz et al. 2021), so the population density of bearded pigs is generally high with possibly little impact on their behavior (Matsuda et al. 2015). TWR is surrounded by oil palm plantations and relatively close to the human settlements where hunting predominantly occurs (Saikim et al. 2023). In the TWR study area, there was sporadic poaching (Hearn et al. 2017). Sambar deer are a target species for poachers in both LKWS and TWR.

The minimum and maximum daily temperatures and annual precipitation did not differ significantly among the study sites (annual temperature: 22–33 ℃, annual precipitation 2,400–3,100 mm; Matsuda et al. 2019; Mitchell 1994; South East Asia Rainforest Research Partnership Unpublished data. https://www.searrp.org/), although no recent, precise climate data were available for TWR.

Data collection
We set up 15, 30, and 28 infrared-triggered sensor cameras (Bushnell, Trophy Cam^TM) in DVCA (July 2010–August 2011 and May 2014–December 2016), LKWS (July 2010–December 2014), and TWR (May 2010–June 2012), respectively. The cumulative number of camera operation days in DVCA, LKWS, and TWR were 14,134, 18,265, and 4,980, respectively, totaling 37,379 days. The camera operating days in DVCA, LKWS, and TWR were 942.2 ± 152.0 (mean ± SD, range = 682–1,229), 608.8 ± 531.4 (range = 28–1,315), and 177.9 ± 123.2 (range = 26–539), respectively.
We defined non-independent photo-capture events as consecutive photos of the same or different individuals of the same species taken within a 30-minute interval and removed these photos from our analysis. We followed Nakabayashi et al. (2021) for details of the camera setting and data collection methods.

Temporal activity analysis

We plotted the activity patterns of each species using the von Mises kernel estimation using the “activity” package (Rowcliffe 2023) in R version 4.2.2 (R Development Core Team 2023).

We divided a day into three periods: nighttime (19:00–04:59h local time (GMT + 8)); daytime (07:00–16:59h); and twilight (05:00–06:59h and 17:00–18:59h). During the study period, twilight hours essentially corresponded to 1 hour between sunset and sunrise, at 05:54–06:25 and 17:50–18:25 in DVCA, 05:51–06:23 and 17:47–18:25 in LKWS, and 05:50–06:21 and 17:46-18:22 in TWR (data from https://www.timeanddate.com). After converting the time data of each photo-capture event into radians, we fitted a circular kernel density distribution estimated by 1,000 bootstrap resampling to the radian time data. For species with fewer than 200 photos, we estimated errors by bootstrapping with sampling from the fitted probability density distribution. For all other species, we sampled from the data (Rowcliffe 2023).

First, we applied multinomial logistic regression analysis to assess the effect of the presence of young on the photo-recorded periods (night, daytime, and twilight periods) of the study species in each study site, using the R package “mlogit” (Croissant 2020). We set the presence of young as a fixed effect and the camera operating days as a weight. There were few photos of sambar deer with young; therefore, we did not conduct this analysis for this species.

We next categorized the species’ activity patterns by testing the selectivity of active periods, i.e., the proportion of periods in which a photo was captured to any of the three periods in a day (e.g., nighttime), using the R package “adehabitatHS” (Calenge 2023). The identification of specific individuals was not possible from the photos, so we used a design Ⅰ resource selection function, selecting at the population level. Once we had conducted multinomial logistic regression analysis and confirmed that the presence of young at each site significantly affected active periods, we separately analyzed the activity patterns of bearded pigs by the presence or absence of young. We divided the activity patterns of the animals into four categories: nocturnal (active at night); crepuscular (active during twilight periods); diurnal (active during daytime); and cathemeral (active during all periods). We defined the activity patterns of the species that showed a statistically higher proportion of photo-captures during the nighttime, daytime, and twilight periods than at other periods as nocturnal, diurnal, and crepuscular, respectively. When there were no differences in the photo-capture proportions among the three periods, we defined the activity pattern as cathemeral. To test for site differences in the activity level estimates during each active period (night, day, twilight), we used the Wald chi-squared test with Bonferroni post hoc correction for the estimated circular kernel density distribution of both species.

We estimated the effects of anthropogenic disturbance on the activity of the two species during each active period (daytime, twilight, and nighttime) using generalized linear mixed models (GLMMs) in the R package “lme4” (Bates et al. 2023). We set the number of independent photo-capture events for each period as the response variable, the distance from oil palm plantations or gravel roads as a fixed variable, year and month as random effects, and the number of camera working days as an offset term. For nocturnal activity, we included the ratio of the illuminated part of moon as an additional fixed variable. We used ArcGIS Pro (ESRI, Redland, CA) to measure the shortest linear distance from each camera to the oil palm plantations in LKWS and TWR and to the gravel road in DVCA, as an indicator of artificial disturbance. In LKWS and TWR, there were both oil palm plantations and gravel roads in the vicinity of the study areas (Figure 1). Once we had confirmed the multicollinearity between these two parameters, we selected oil palm plantations as an index of anthropogenic effect, given that the number of humans there was large and their effect would be greater. We checked the ratio of the shining part of the moon (https://www.arachne.jp/onlinecalendar/mangetsu/) and used this as an indicator of a night’s brightness. The significance of fixed effects was examined by likelihood-ratio tests using the “car” package in R (Fox et al. 2023).

Sampling effort

We recorded 4,825 and 713 photos of bearded pigs and sambar deer, respectively. Bearded pigs were present in 1,747, 2,090, and 988 and sambar deer were present in 411, 89, and 213 photos in DVCA, LKWS, and TWR, respectively. The number of photos of bearded pigs with young was 193, 126, and 33 in DVCA, LKWS, and TWR, respectively. Due to the small sample size, we sampled from the fitted probability density distribution of bearded pigs with young and sambar deer to estimate bootstrapping errors with sampling.

Diel activity patterns of bearded pigs

Table 1 shows the multinomial logistic regression analysis results, including the coefficients and p-values for the different active periods (daytime and twilight) in bearded pigs. The nighttime period was considered the base outcome. In LKWS, bearded pigs with young were more active during the day than during the night. In TWR and DVCA, bearded pigs with young were more active in both the daytime and twilight periods compared with nighttime. These results indicated that the activities of bearded pigs at all sites were affected by the presence of young. Therefore, we subsequently conducted separate analyses of bearded pig activity depending on the presence or absence of young.

The results of selectivity of active time in bearded pigs with young showed a similar tendency, as they were more active in the daytime than expected (all p<0.01) and they were less active in the nighttime than expected (all p<0.01). Bearded pigs in DVCA and TWR showed no selectivity for the twilight period (p = 0.47 in DVCA; p = 0.02 in TWR); however, in LKWS the twilight period was used significantly less than expected (p<0.01) (Figure 2). The activity pattern of bearded pigs with young was defined as diurnal at all three sites. Bearded pigs without young in DVCA and LKWS showed similar tendencies to each other, as they used daytime significantly more than expected (both p<0.01) and nighttime significantly less than expected (both p<0.01). The usage of the twilight period showed no statistical significance (p = 0.22 in DVCA; p = 0.76 in LKWS). In TWR, bearded pigs used the twilight period significantly more than expected (p<0.01) but the usage of both daytime and nighttime periods showed no statistical significance (p = 0.22 and p = 0.14, respectively) (Figure 3). We defined the activity patterns of bearded pigs without young in DVCA and LKWS as diurnal and those in TWR as crepuscular.

The among-site comparison of activity level estimates is shown in Table 2. In bearded pigs with young, there were no statistical differences in activity levels in any periods across the sites. For bearded pigs without young, significant differences in activity levels were found for all time periods. At night, the activity level of individuals in TWR differed from those in DVCA. The difference in activity levels between individuals in TWR and LKWS gave a marginal significance level. In the daytime, the activity levels of individuals in TWR differed from those in both DVCA and LKWS. The activity levels of individuals during the twilight period differed among the three sites, although the significance was marginal between DVCA and LKWS.

Diel activity patterns of sambar deer

Sambar deer were more active during twilight than expected across the sites (all p<0.01). In DVCA, sambar deer were less active during the daytime than expected (p<0.01), while their activity during the nighttime showed no statistical significance (p = 0.45). There were no significant differences in the time usage during daytime (p = 0.34) or nighttime (p = 0.093) by individuals in LKWS. Individuals in TWR were less active during the nighttime than expected (p = 0.012), while their activity during the daytime showed no statistical significance (p = 0.16) (Figure 4). We defined sambar deer activity patterns as crepuscular at all three sites.

For sambar deer, there was a marginally significant difference in daytime activity levels between LKWS and TWR (Table 2). There were no significant differences for the other combinations.

Effects of anthropogenic disturbance on diel activity patterns at the various sites

Twilight and daytime periods

The GLMM results of the effects of anthropogenic disturbance during twilight and daytime periods indicated that both species responded similarly to oil palm plantations. Due to the small sample size of bearded pigs with young in TWR, however, we could not conduct an analysis of this group. In LKWS and TWR, when bearded pigs without young were closer to plantations, they were significantly more active during twilight. For all other cases, there were no significant differences (Table 3). In contrast, in DVCA, when individuals without young were farther from the gravel road, they were significantly more active during the daytime. Similar to bearded pigs, sambar deer in LKWS and TWR were significantly more active during twilight the closer they were to plantations (Table 3). In TWR, they were more active during the daytime, as they were close to the plantation. There were no tendencies between the photo numbers and the distance from the road in the individuals in DVCA.

Nighttime period

The GLMM results showed that during the nighttime, there were differences in the effects of the brightness of the moon and the distance from oil palm plantations or a gravel road on nocturnal activity, depending on species and study site (Table 4). Due to the small sample size of bearded pigs with young, we could not analyze them for all sites. In LKWS, bearded pigs tended to be more active when more of the moon was illuminated. In LKWS and TWR, the nocturnal activity of bearded pigs also increased the closer they were to oil palm plantations. However, unlike bearded pigs in LKWS, sambar deer in TWR were more active on nights when the moon was less illuminated. Sambar deer in TWR tended to be more active at night when they were close to oil palm plantations, although the significant differences were marginal. On the other hand, in DVCA, the degree of illumination of the moon and the proximity to the gravel road had no effect on the levels of activity of either bearded pigs or sambar deer.

The diel activity of bearded pigs varied by site, while sambar deer showed relatively fixed diel activity patterns across the three sites. Responses to human-modified environments also varied across the sites and species, suggesting that such differences were related to the level of anthropogenic disturbance. The small number of photos of bearded pigs with young in TWR and sambar deer in LKWS suggest that in these sites these individuals may have moved away from the study areas, which were relatively close to anthropogenically disturbed environments. Inter-site differences in diel activities in each species and the impact of anthropogenic disturbance levels on these species are discussed in more detail below.

Bearded pigs

Overall activity patterns across different sites

We found that, when traveling with young, the temporal activity patterns of bearded pigs were generally diurnal. This not only supports the results of previous studies conducted in Sabah (Davison et al. 2019; Love et al. 2017; Ross et al. 2013) but is also consistent with observations in other ungulates (Stankowich 2008). This activity pattern appears to be an adaptation to reduce predation on vulnerable young by avoiding the active period of Sunda clouded leopards (Neofelis diardi), their primary predator in Borneo (Ross et al. 2013). In addition, given that the closely related wild boar (Sus scrofa) lacks the tapetum lucidum necessary to enhance vision in low light levels (Gordigiani et al. 2022; Ollivier et al. 2004), it is likely that bearded pigs share a similar vision impairment, suggesting that diurnal activity with their young may facilitate behaviors such as foraging, as well as predation avoidance.

Conversely, the temporal activity pattern of bearded pigs without young in TWR differed from the other two sites; it was predominantly crepuscular in TWR while in DVCA and LKWS it was consistently diurnal, irrespective of the presence or absence of young. As the active period of individuals in TWR overlaps with that of Sunda clouded leopards (Hearn et al. 2018; Ross et al. 2013), factors other than predation pressure may be shaping their activity patterns. Crepuscular activity patterns are, for example, often explained as helping to avoid high temperatures (Davison et al. 2019; Owen-Smith and Cain 2007; Peterson et al. 2021). Indeed, irrespective of the presence of young, individuals in TWR reduce their activity during the hottest part of the day (1200–1600h, Supplementary Table 1). However, considering that individuals in the logged forest at other sites, i.e., LKWS, exhibit diurnal activity patterns, the effect of air temperature on the distinct crepuscular activity pattern of individuals in TWR is likely to be low.

One plausible explanation for such a difference in TWR could be anthropogenic disturbance. Given that the regular operating hours of oil palm plantations in Sabah are from 0600h to 1800h (Davison et al. 2019), the highest probability of encountering humans is during the daytime. Thus, the activity pattern of bearded pigs, where oil palm plantations form part of their ranging area, has been reported to shift to nocturnal activity (Davison et al. 2019; Love et al. 2018). This is consistent with becoming nocturnal as a general response of wildlife to avoid encounters with humans (Gaynor et al. 2018). By contrast, poaching typically occurs at night in Sabah (Wong et al. 2004). For individuals in TWR, where poaching for this species likely occurs most frequently among the study sites (Hearn et al. 2017), the twilight period may be more suitable for their activities because they are less likely to encounter plantation workers or be hunted than during the day- or night-time.

Spatiotemporal activity pattern across the various sites

Based on the locations where cameras were installed at each study site, our analysis also revealed a tendency for human avoidance in bearded pigs. In particular, there are oil palm plantations in the vicinity of both LKWS and TWR, and the response of bearded pigs to the oil palm plantations was similar at each site. As they came nearer to the plantations, individuals without young became more active during the twilight and nighttime periods. This was presumably to forage for oil palm fruits (Kurz et al. 2021; Love et al. 2018) when there were fewer people around. In short, they avoid encounters with humans but take advantage of human-modified environments, especially those they can use for feeding. In DVCA, where there are no oil palm plantations in the vicinity, the population without young tends to avoid engaging in activity near the gravel road. In DVCA, where interactions with humans are generally less likely, bearded pigs may be more sensitive to the presence of human occurrences and more vigilant to anthropogenic stimuli.

Effects of lunar illumination on nocturnal activity

In LKWS, nocturnal activity patterns of bearded pigs were notably affected by the degree of lunar illumination; they were more active when the moonlight was brighter. Considering the poor nocturnal vision of closely related wild boar (Gordigiani et al. 2022; Ollivier et al. 2004), a plausible explanation of such behavioral changes observed in LKWS could be that they rely on moonlight for foraging and predator detection, as has been reported in European wild boar (Brivioa et al. 2017). It should be noted, however, that bearded pigs in LKWS are typically diurnal. Therefore, in this context, these are complementary foraging behaviors that benefit from a specific nocturnal environment that increases predator/hunter detectability and foraging efficiency, as well as the advantage of avoiding human contact in the vicinity of the plantation.

On the other hand, individuals without young in TWR that approached the plantation at night did not show a similar tendency to that seen in LKWS, suggesting that lunar luminosity is not an essential factor for their nocturnal activity in TWR. Given the greater poaching pressure on bearded pigs in TWR compared with LKWS (Hearn et al. 2017; Kurz et al. 2021), poachers would likely be more active on brightly moonlit evenings in TWR, and thus, unlike in LKWS, bearded pigs might avoid activity on such nights. Hence, it could be possible that the effects of lunar illumination in TWR were not as influential as in LKWS. Nighttime foraging in plantations should thus have some benefits for bearded pigs in TWR beyond these risks at night. It should be noted, however, that there are limitations to analyzing the impact of the degree of lunar illumination on activity patterns, as the actual level of brightness on the ground must be investigated to take into account the possibility of cloud cover. With respect to the 36 to 41% lower level of daytime activity among individuals without young in TWR compared with DVCA and LKWS (Supplementary Table 1), daytime activity, especially foraging behaviors, may be constrained by several factors. As discussed earlier, a plausible factor for the reduction in time spent being active during the day in bearded pigs in TWR, which reduces their interactions with plantation workers, may contribute to their limited hours spent foraging during the daytime. Hence, they may approach plantations where they can more efficiently acquire highly nutrient-rich oil palm fruits, despite increased hunting and predation pressure during the night. Furthermore, the percentage of time bearded pigs without young were active in TWR was never less than 2% (Supplementary Table 1), supporting the possibility that they maintain a certain amount of activity throughout the day to find food that meets their nutritional requirements.

Sambar deer

Overall activity patterns across different sites

Overall, sambar deer were predominantly crepuscular throughout the study sites, and their activity levels during each period (night, day, twilight) did not differ significantly among the three sites. Unlike bearded pigs, sambar deer have a tapetum lucidum (VerCauteren and Pipas 2003), so it is likely that crepuscular and also nocturnal activity patterns are a predominant mode for their behavior. The overall pattern of sambar deer activity tended to be similar at all sites, although there were slight differences at TWR. Those in TWR were less active at night, while those in DVCA were less diurnal. Given the considerable overlap of the activity patterns of clouded leopards with sambar deer, i.e., lower daytime activity levels (Supplementary Table 1), factors other than predation by clouded leopards may shape the differences in activity patterns of sambar deer in TWR (Ross et al. 2013). One explanation for this difference in TWR may be human disturbance. In TWR, where poaching has been reported (Hearn et al. 2017), sambar deer may reduce nighttime activities to avoid encounters with poachers. However, sambar deer in Sarawak, Sabah’s neighboring state are equally active day and night, irrespective of predation or poaching pressure, although smaller, closely related taxa, e.g., barking deer (Muntiacus spp.) and mouse deer (Tragulus spp.), flexibly adjust their activity patterns depending on the degree of logging and hunting in their environment (Bersacola et al. 2019). The larger sambar deer may be less sensitive to environmental changes but may respond to excessive hunting/predation pressure as found in TWR. Future work should quantify and compare these pressures across different sites and examine human influence on sambar deer activity patterns.

Spatiotemporal activity patterns across different sites

Considering the camera installation locations at each study site, the effect of oil palm plantations and the gravel road on sambar deer activity patterns showed differences between study sites. During the daytime, deer in DVCA stayed away from the gravel road, while those in TWR were frequently photographed near the plantation. This species is shy and elusive (Leslie 2011) and avoids villages (Singh et al. 2022), so in DVCA, it is not surprising that deer avoid the gravel road used by vehicles during the daytime. However, individuals in TWR approached plantations when plantation workers were active. If it is not lethal human contact, such as hunting, but non-lethal stimuli e.g., continuous noise and human disturbance, sambar deer may adjust to such stimuli (e.g., Stankowich 2008) and subsequently invest more in feeding than avoiding humans.

Nocturnal activity patterns and effects of lunar illumination on nocturnal activity

The nocturnal behavior of sambar deer differed across sites and was affected by oil palm plantations. In TWR, they showed a similar tendency to bearded pigs, being active near the plantation at night. Herbicides are commonly used in oil palm plantations (Dilipkumar et al. 2020), reducing food resources for typical herbivores such as sambar deer (Leslie 2011). Sambar deer often feed on understory foliage in logged forests and at forest edges (Davies 2001). In the TWR study area, a gravel road and the oil palm plantation are both located on the forest edge (Figure 1). In Borneo, the roadside understory includes grass, herbs, and shrubs (Padmanaba and Sheil 2014), i.e., food resources for sambar deer, which at night are often observed near the road in TWR (Nakabayashi et al. 2014), suggesting sambar deer in TWR approach the road/plantation to feed. This may be their foraging strategy to avoid times when humans are active in the plantation, although there may be an increased risk of poaching near the road/plantation at night. Conversely, in LKWS, sambar deer may feed on the riparian understory in riverine environments, avoiding the need to approach the plantation. The balance between the conflicting factors of poaching risk and the need to forage may cause sambar deer to adjust activity patterns in human-modified environments at night.

Our results showed sambar deer were typically active during twilight periods but also active at night. Additionally, in TWR they were often photographed when lunar illumination was low. Considering their good nocturnal vision (VerCauteren and Pipas 2003), they can exploit the dark to avoid poaching, especially in TWR, where poaching pressure is high. A night without moonlight is unsuitable for poaching, but suitable for sambar deer to actively forage.

Our findings suggest that activity patterns of bearded pigs and sambar deer are generally affected by human activities; these behavioral responses differed across study sites. In terms of temporal activity patterns, bearded pigs are affected by anthropogenic disturbance, especially poaching when they are the target species. Temporal activity of sambar deer may also be affected by poaching pressure in some environments, although their core activity patterns were generally consistent across all sites. Spatiotemporal activity patterns also indicated elements of human avoidance. Bearded pigs approached plantations during times of low human presence, presumably to forage, especially in study sites where poaching occurs. Where poaching was rare, they exhibited spatiotemporal activity patterns that avoided human contact, staying away from roads. Notably, our results suggest reduced active times in both species in TWR, possibly reducing foraging times and fitness and increasing energy expenditure (Kiffner et al. 2014). Thus, human activities may negatively affect the activity patterns of these two ungulate species. The impact of human activity, especially on diel activity in both ungulate species, may reduce the predation success of their natural predators, such as clouded leopards, which consequently may affect the ecological balance by altering predator–prey densities.

These ungulate species also, however, exhibit some behavioral plasticity to anthropogenic disturbance, which could potentially positively affect their feeding. Artificial environments such as oil palm plantations and roads provide food, and the two ungulate species may efficiently compensate for reduced feeding times by using such food resources. The different movement patterns in sambar deer between LKWS and TWR may also reflect their behavioral plasticity, despite being exposed to poaching pressure at both sites. A previous study suggested they can persist in disturbed habitats (Granados et al. 2016). Local people in Sabah observed that bearded pigs become highly sensitive to the presence of humans in hunting areas (Kurz et al. 2021). Our results show they can adapt to human-modified environments, with some individuals of both species in DVCA habituated and often observed near tourist accommodation. When regulated seasonal hunting is permitted, some ungulates that regularly experience non-lethal human activities do not show a fear response to humans (Stankowich 2008). In our study sites, poaching is irregular and ecotourism operates year-round. These factors might reduce the fear response to humans, enabling these ungulate species to use resources in plantations and the surroundings.

This study was conducted within protected areas where, although sporadic poaching occurs, the two ungulate species were relatively common. Hunting and poaching pressure outside the protected areas is likely much higher than in the study sites. In such areas, local wildlife populations would suffer greatly without appropriate hunting regulations.

Camera trapping is an efficient means of estimating wildlife population dynamics and behavior (Wearn and Glover-Kapfer 2019) and can facilitate regional comparisons of wildlife behavioral responses. We recommend urgent evaluations of current wildlife management plans and long-term monitoring of wildlife population dynamics, including studying behavioral and physiological responses that directly measure animals’ stress, especially in Southeast Asia.

Funding

This research was partially funded by the Japan Society for the Promotion of Science KAKENHI (# 22687002 to GH; #26711027 and #19H03308 to IM; #17K15285 and #20K15555 to MN), for Core-to-Core Program, Advanced Research Networks (#JPJSCCA20170005 to S. Kohshima), and for the Grant-in-Aid for JSPS Fellows (#25-597 and # 201608680), and the Sasagawa Scientific Research Grant from the Japan Science Society (#22-537 to MN).

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Author Contributions

MN, TK, IM and GH conceptualized the initial idea. MN, AM, TK and IM set up the cameras and

obtained the field data. JT, AT, TPM, HB and AHA arranged the sampling in the field. MN

performed and interpreted the statistical analyses. MN, TK, IM and GH drafted the manuscript. All

authors contributed to the final version of the manuscript.

Data Availability

Data are available on request from the authors.

Acknowledgements

We thank the Sabah Biodiversity Centre, the Sabah Forestry Department, the Sabah Wildlife Department, and the Danum Valley Management Committee for granting us permission for this research. We are grateful for the support from our research assistants in the fields. This research was partially funded by the Japan Society for the Promotion of Science KAKENHI (# 22687002 to GH; #26711027 and #19H03308 to IM; #17K15285 and #20K15555 to MN), for Core-to-Core Program, Advanced Research Networks (#JPJSCCA20170005 to S. Kohshima), and for the Grant-in-Aid for JSPS Fellows (#25-597 and # 201608680), and the Sasagawa Scientific Research Grant from the Japan Science Society (#22-537 to MN).

Abram NK, Xofis P, Tzanopoulos J, MacMillan DC, Ancrenaz M, Chung R, et al. (2014) Synergies for improving oil palm production and forest conservation in floodplain landscapes. PLoS One 9:e95388. https://doi.org/10.1371/journal.pone.0095388
Andermann T, Faurby S, Turvey ST, Antonelli A, Silvestro D (2020) The past and future human impact on mammalian diversity. Science Advances 6. https://doi.org/10.1126/sciadv.abb2313
Bates, D. et al. (2023) lme4: Linear Mixed-Effects Models using 'Eigen' and S4. R package version 1.1-35.1. https://CRAN.R-project.org/package=lme4 (Accessed 14 Jan 2024).
Bejder L, Samuels A, Whitehead H, Finn H, Allen S (2009) Impact assessment research: use and misuse of habituation, sensitisation and tolerance in describing wildlife responses to anthropogenic stimuli. Mar Ecol Prog Ser 395:177-185. https://doi.org/10.3354/meps07979
Bennett EL, Nyaoi AJ, Sompud J (2000) Saving Borneo's Bacon: The Sustainability of Hunting in Sarawak and Sabah. In: Robinson J, Bennett E (eds) Hunting for Sustainability in Tropical Forests. Columbia University Press, New York, pp 305-324
Bersacola E, Sastramidjaja W, Rayadin Y, Macdonald D, Cheyne SM (2019) Occupancy patterns of ungulates and pig-tailed macaques across regenerating and anthropogenic forests on Borneo. Hystrix, the Italian Journal of Mammalogy 30:126-133.
Brivioa F, Grignolioa S, Brogia R, Benazzib M, Bertoluccib C, Apollonio M (2017) An analysis of intrinsic and extrinsic factors affecting the activity of a nocturnal species: The wild boar. Mammalian Biology 84:73–81
Brodie JF, Giordano AJ, Zipkin EF, Bernard H, Mohd-Azlan J, Ambu L (2015) Correlation and persistence of hunting and logging impacts on tropical rainforest mammals. Conserv Biol 29:110-121. https://doi.org/10.1111/cobi.12389
Calenge, (2023) adehabitatHS: Analysis of habitat selection by animals. R package version 0.3.17. https://CRAN.R-project.org/package=adehabitatHS (Accessed 14 Jan 2024).
Ceballos G, Ehrlich PR (2002) Mammal population losses and the extinction crisis. Science 296:904-907.
Ceballos G, Ehrlich PR, Raven PH (2020) Vertebrates on the brink as indicators of biological annihilation and the sixth mass extinction. Proc Natl Acad Sci U S A 117:13596-13602.
Ciuti S, Northrup JM, Muhly TB, Simi S, Musiani M, Pitt JA, et al. (2012) Effects of humans on behaviour of wildlife exceed those of natural predators in a landscape of fear. PLoS One 7:e50611. https://doi.org/10.1371/journal.pone.0050611
Colino-Rabanal VJ, Peris SJ, Langen TA, Lizana M (2018) Ungulate: vehicle collision rates are associated with the phase of the moon. Biodiversity Conservation 27:681–694
Corlett RT (2007) The Impact of Hunting on the Mammalian Fauna of Tropical Asian Forests. Biotropica 39:292-303. https://doi.org/10.1111/j.1744-7429.2007.00271.x
Corlett RT (2010) Megafaunal extinctions and their consequences in the tropical Indo-Pacific. Terra Australis 32:117-131
Croissant Y (2020) mlogit. R package version 1.1-1. https://CRAN.R-project.org/package=mlogit (Accessed 14 Jan 2024).
Cromsigt JPGM, Kuijper DPJ, Adam M, Beschta RL, Churski M, Eycott A, et al. (2013) Hunting for fear: innovating management of human–wildlife conflicts. Journal of Applied Ecology 50:544-549. https://doi.org/10.1111/1365-2664.12076
Crosmary W-G, Makumbe P, Côté SD, Fritz H (2012) Vulnerability to predation and water constraints limit behavioural adjustments of ungulates in response to hunting risk. Animal Behaviour 83:1367-1376.
Davies G, Heydon MJ, Leader-Williams N, MacKinnon J, Newing H (2001) The effects of logging on tropical forest ungulates. In: Fimbel RA, Grajel A, Robinson JG (eds) The Cutting Edge: Conserving Wildlife in Logged Tropical Forests. Columbia University Press, New York, pp. 93–124.
Davison CW, Chapman PM, Wearn OR, Bernard H, Ewers RM (2019) Shifts in the demographics and behavior of bearded pigs (Sus barbatus) across a land‐use gradient. Biotropica 51:938-948. https://doi.org/10.1111/btp.12724
Dilipkumar M, Chuah TS, Goh SS, Sahid I (2020) Weed management issues, challenges, and opportunities in Malaysia. Crop Protection. 134: 104347. https://doi.org/10.1016/j.cropro.2017.08.027
Fox J et al. (2023) car: Companion to Applied Regression. R package version 3.1-2. https://CRAN.R-project.org/package=car (Accessed 14 Jan 2024).
Gaynor KM, Hojnowski CE, Carter NH, Brashares JS (2018) The influence of human disturbance on wildlife nocturnality. Science 360:1232-1235. https://doi.org/10.1126/science.aar7121
Gaveau DL et al. (2016) Rapid conversions and avoided deforestation: examining four decades of industrial plantation expansion in Borneo. Scientific Reports 6:32017. https://doi.org/10.1038/srep32017
Gaynor KM, Hojnowski CE, Carter NH, Brashares JS (2018) The influence of human disturbance on wildlife nocturnality. Science 360:1232-1235.
Goudie AS (2013) The human impact on the natural environment: Past, present, and future, seventh edition edn. Wiley-Blackwell.
Goossens B et al. (2005) Patterns of genetic diversity and migration in increasingly fragmented and declining orang-utan (Pongo pygmaeus) populations from Sabah, Malaysia. Molecular Ecology 14:441-456.
Gordigiani L, Viviano A, Brivio F, Grignolio S, Lazzeri L, Marcon A, et al. (2022) Carried away by a moonlight shadow: activity of wild boar in relation to nocturnal light intensity. Mammal Research 67:39-49.
Granados A, Crowther K, Brodie JF, Bernard H (2016) Persistence of mammals in a selectively logged forest in Malaysian Borneo. Mammalian Biology 81:268-273.
Harrison RD, Sreekar R, Brodie JF, Brook S, Luskin M, O'Kelly H, et al. (2016) Impacts of hunting on tropical forests in Southeast Asia. Conservation Biology 30:972-981. https://doi.org/10.1111/cobi.12785
Hearn AJ, Ross J, Bernard H, Bakar SA, Goossens B, Hunter LTB, et al. (2017) Responses of Sunda clouded leopard Neofelis diardi population density to anthropogenic disturbance: refining estimates of its conservation status in Sabah. Oryx 53:643-653. https://doi.org/10.1017/s0030605317001065
Hearn AJ, Cushman SA, Ross J, Goossens B, Hunter LT, Macdonald DW (2018) Spatio-temporal ecology of sympatric felids on Borneo. Evidence for resource partitioning? PloS one 13:e0200828. https://doi.org/10.1371/journal.pone.0200828
IUCN (2024) The IUCN Red List of Threatened Species. Version 2023-1. <https://www.iucnredlist.org>
Kays R, Parsons AW, Baker MC, Kalies EL, Forrester T, Costello R, et al. (2016) Does hunting or hiking affect wildlife communities in protected areas? Journal of Applied Ecology 54:242-252. https://doi.org/10.1111/1365-2664.12700
Kiffner C et al. (2014) Interspecific variation in large mammal responses to human observers along a conservation gradient with variable hunting pressure. Anim Conserv 17:603-612. https://doi.org/10.1111/acv.12131
Kurz DJ, Saikim FH, Justine VT, Bloem J, Libassi M, Luskin MS, et al. (2021) Transformation and endurance of Indigenous hunting: Kadazandusun‐Murut bearded pig hunting practices amidst oil palm expansion and urbanization in Sabah, Malaysia. People and Nature 3:1078-1092. https://doi.org/10.1002/pan3.10250
Leslie DM (2011) Rusa unicolor (Artiodactyla: Cervidae). Mammalian Species 43:1-30.
Love K, Kurz DJ, Vaughan IP, Ke A, Evans LJ, Goossens B (2017) Bearded pig (Sus barbatus) utilisation of a fragmented forest–oil palm landscape in Sabah, Malaysian Borneo. Wildlife Research 44. https://doi.org/10.1071/wr16189
Marsh CW, Greer AG (1992) Forest land-use in Sabah, Malaysia: an introduction to Danum Valley. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 335:331-339
Matsuda I, Ancrenaz M, Akiyama Y, Tuuga A, Majalap N, Bernard H (2015) Natural licks are required for large terrestrial mammals in a degraded riparian forest, Sabah, Borneo, Malaysia. Ecological Research 30:191-195. https://doi.org/10.1007/s11284-014-1219-1
Matsuda I, Nakabayashi M, Otani Y, Yap SW, Tuuga A, Wong A, et al (2019) Comparison of plant diversity and phenology of riverine and mangrove forests with those of the dryland forest in Sabah, Borneo, Malaysia. In Primates in Flooded Habitats: Ecology and Conservation (eds Nowak, K. et al.) pp 15–28, Cambridge University Press.
Mitchell AH (1994) Ecology of Hose’s langur (Presbytis Hosei) in logged and unlogged dipterocarp forest of northeast Borneo. PhD dissertation, Yale University.
Nakabayashi M, Nakashima Y, Bernard H, Kohshima S (2014) Utilisation of gravel roads and roadside forests by the common palm civet (Paradoxurus hermaphroditus) in Sabah, Malaysia. Raffles Bulletin of Zoology 62: 379-388
Nakabayashi M, Kanamori T, Matsukawa A, Tangah J, Tuuga A, Malim PT, et al. (2021) Temporal activity patterns suggesting niche partitioning of sympatric carnivores in Borneo, Malaysia. Scientific Reports 11: 19819. https://doi.org/10.1038/s41598-021-99341-6
Nix JH, Howell RG, Hall LK, McMillan BR (2018) The influence of periodic increases of human activity on crepuscular and nocturnal mammals: Testing the weekend effect. Behavioural Processes 146:16-21.
Ocampo-Peñuela N, Garcia-Ulloa J, Kornecki I, Philipson CD, Ghazoul J (2020) Impacts of four decades of forest loss on vertebrate functional habitat on Borneo. Frontiers in Forests and Global Change 3: 53. https://doi.org/10.3389/ffgc.2020.00053
Ollivier FJ, Samuelson DA, Brooks DE, Lewis PA, Kallberg ME, Komáromy AM (2004) Comparative morphology of the tapetum lucidum (among selected species). Veterinary ophthalmology 7:11–22.
Owen-Smith N, Cain JW (2007) Indicators of adaptive responses in home range utilization and movement patterns by a large mammalian herbivore. Israel Journal of Ecology and Evolution 53:423–438
Padmanaba M, Sheil D (2014) Spread of the Invasive Alien Species Piper Aduncum via Logging Roads in Borneo. Tropical Conservation Science 7: 35-44
Pascual-Rico R, Morales-Reyes Z, Aguilera-Alcala N, Olszanska A, Sebastian-Gonzalez E, Naidoo R, et al. (2021) Usually hated, sometimes loved: A review of wild ungulates' contributions to people. Science of the Total Environment 801:149652. https://doi.org/10.1016/j.scitotenv.2021.149652
Peterson M, Jorge MLSP, Jain A, Keuroghlian A, Oshima JEF, Richard-Hansen C, et al. (2021) Temperature induces activity reduction in a Neotropical ungulate. Journal of Mammalogy 102:1514-1524.
R Development Core Team (2022) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. version 4.2.1. https:// www.R- project.org/.
Ripple WJ, Abernethy K, Betts MG, Chapron G, Dirzo R, Galetti M, et al. (2016) Bushmeat hunting and extinction risk to the world's mammals. Royal Society Open Science 3:160498. https://doi.org/10.1098/rsos.160498
Ross J, Hearn A, Johnson P, Macdonald D (2013) Activity patterns and temporal avoidance by prey in response to Sunda clouded leopard predation risk. Journal of Zoology 290:96-106
Rowcliffe M (2023) Activity: Animal activity statistics. R package version 1.3.4. https://CRAN.R-project.org/package=activity (Accessed 14 Jan 2024).
Saikim FH, Mohd YNS, Mahyudin A, Nathan SKSS, Muin H, Hamid A, et al. (2023) Empty Forest Syndrome: Are we there yet? A preliminary Analysis of the Hunting and Poaching Activities in Selected Areas in Sabah, Malaysia. Journal of Tropical Biology and Conservation 20: 219-244.
Schipper J, Chanson JS, Chiozza F, Cox NA, Hoffmann M, Katariya V, et al. (2008) The status of the world's land and marine mammals: diversity, threat, and knowledge. Science 322:225-230. https://doi.org/10.1126/science.1165115
Singh H, Sharief A, Joshi BD, Kumar V, Mukherjee T, Chandra K, et al. (2022) Multi-species occupancy modeling suggests interspecific interaction among the three ungulate species. Scientific Reports 12:17602. https://doi.org/10.1038/s41598-022-20953-7
Srivastava DS, Cadotte MW, MacDonald AAM, Marushia RG, Mirotchnick N (2012) Phylogenetic diversity and the functioning of ecosystems. Ecology Letters 15: 637–648.
Stankowich T (2008) Ungulate flight responses to human disturbance: A review and meta-analysis. Biological Conservation 141:2159-2173.
Tablado Z, Jenni L (2017) Determinants of uncertainty in wildlife responses to human disturbance. Biological reviews of the Cambridge Philosophical Society 92:216-233.
Tilman D, Clark M, Williams DR, Kimmel K, Polasky S, Packer C (2017) Future threats to biodiversity and pathways to their prevention. Nature 546:73-81. https://doi.org/10.1038/nature22900
VerCauteren KC, Pipas MJ (2003) A review of color vision in white-tailed deer. Wildlife Society Bulletin 31: 684-691.
Visscher DR, Walker PD, Flowers M, Kemna C, Pattison J, Kushnerick B (2023) Human impact on deer use is greater than predators and competitors in a multiuse recreation area. Animal Behaviour 197:61-69.
Wearn OR, Glover-Kapfer P (2019) Snap happy: camera traps are an effective sampling tool when compared with alternative methods. Royal Society Open Science 6:181748. https://doi.org/10.1098/rsos.181748
Whittaker D, Knight RL (1998) Understanding wildlife responses to humans. Wildlife Society Bulletin 26:312-317.
Wong ST, Servheen CW, Ambu L (2004) Home range, movement and activity patterns, and bedding sites of Malayan sun bears Helarctos malayanus in the rainforest of Borneo. Biological Conservation 119:169-181
Yi MCK, Mohd-Azlan J (2020) Wildlife Hunting and Utilization in Ulu Baleh, Sarawak, Malaysian Borneo. Ethnobiology Letters 11:76-84. https://doi.org/10.14237/ebl.11.1.2020.1647

Table 1. Results of multinomial logistic regression analysis for the effect of the presence of young in bearded pigs on periods of activity. The nighttime period was the base outcome. Significant values are shown in bold.

site	variables	daytime		twilight
site	variables	coefficient	p-value	coefficient	p-value
DVCA	Intercept	1.09	< 0.01	0.18	0.02
DVCA	presence of young	2.11	< 0.01	1.06	0.02
LKWS	Intercept	1.42	< 0.01	0.46	< 0.01
LKWS	presence of young	1.21	< 0.01	0.37	0.41
TWR	Intercept	-0.06	0.44	-0.04	0.63
TWR	presence of young	1.99	< 0.01	1.44	0.03

Table 2. Results of Wald-test pairwise comparisons of activity levels during each period among the study sites. Significant values are shown in bold.

species		period	site comparison	Differences between estimates	SE	Wald statistic	p-value
species		period	site comparison	Differences between estimates	SE	Wald statistic	p-value
Bearded pig	with young	Night	DVCA vs LKWS	0.01	0.08	0.01	0.94
			DVCA vs TWR	-0.13	0.10	1.65	0.20
			LKWS vs TWR	-0.14	0.10	2.14	0.14
		Day	DVCA vs LKWS	-0.01	0.04	0.08	0.78
			DVCA vs TWR	0.07	0.07	1.01	0.31
			LKWS vs TWR	0.09	0.08	1.27	0.26
		Twilight	DVCA vs LKWS	0.03	0.03	0.98	0.32
			DVCA vs TWR	0.01	0.03	0.13	0.72
			LKWS vs TWR	-0.01	0.03	0.26	0.61

	without young	Night	DVCA vs LKWS	-0.04	0.04	1.19	0.28
			DVCA vs TWR	-0.12	0.03	12.57	< 0.01
			LKWS vs TWR	0.08	0.04	4.46	0.03*
		Day	DVCA vs LKWS	-0.03	0.02	1.77	0.18
			DVCA vs TWR	-0.10	0.03	14.76	< 0.01
			LKWS vs TWR	0.07	0.03	7.67	< 0.01
		Twilight	DVCA vs LKWS	-0.02	0.01	4.87	0.03*
			DVCA vs TWR	-0.05	0.01	24.24	< 0.01
			LKWS vs TWR	0.03	0.01	6.17	0.01

Sambar deer		Night	DVCA vs LKWS	0.02	0.06	0.08	0.78
			DVCA vs TWR	0.01	0.06	0.03	0.86
			LKWS vs TWR	-0.01	0.07	0.01	0.92
		Day	DVCA vs LKWS	0.08	0.07	1.65	0.20
			DVCA vs TWR	-0.06	0.07	0.74	0.39
			LKWS vs TWR	-0.14	0.06	5.44	0.02*
		Twilight	DVCA vs LKWS	-0.03	0.02	1.82	0.18
			DVCA vs TWR	-0.01	0.02	0.15	0.70
			LKWS vs TWR	0.02	0.02	1.01	0.31

Table 3. Results of the GLMM for the effects of anthropogenic disturbance on activities during twilight and daytime periods. Significant values are shown in bold.

species	site			daytime			twilight
species	site		variables	coefficient	χ²	p-value	coefficient	χ²	p-value
Bearded pig	LKWS	with young	Intercept	-2.98			-3.18
		with young	distance from the oil palm plantation	-0.00	2.66	0.09	-0.00	2.75	0.10
		without young	Intercept	-3.46			-3.58
		without young	distance from the oil palm plantation	-0.00	0.01	0.91	-0.00	8.70	< 0.01
	TWR	with young	Intercept	NA			NA
		with young	distance from the oil palm plantation	NA			NA
		without young	Intercept	-1.08			1.65
		without young	distance from the oil palm plantation	-0.00	0.59	0.44	-0.00	5.74	0.017
	DVCA	with young	Intercept	-4.91			-5.82
		with young	distance from the gravel road	0.00	0.75	0.39	0.00	2.07	0.15
		without young	Intercept	-6.97			-5.32
		without young	distance from the gravel road	0.00	35.93	< 0.01	0.00	1.48	0.22

Sambar deer	LKWS		Intercept	-5.11			-3.62
	LKWS		distance from the oil palm plantation	-0.00	0.39	0.53	-0.00	19.46	< 0.01
	TWR		Intercept	-1.57			-0.43
	TWR		distance from the oil palm plantation	-0.00	9.03	< 0.01	-0.00	3.87	0.049
	DVCA		Intercept	-5.54			-4.63
	DVCA		distance from the gravel road	0.00	0.20	0.65	-0.00	1.53	0.22

Table 4. Results of the GLMM for the effects of anthropogenic disturbance and lunar illumination on nocturnal activity. Significant values are shown in bold.

species	site	Variables	coefficient	χ²	p-value
Bearded pig	LKWS	Intercept	-4.00
		moon brightness rate	0.00	3.89	0.05
		distance from the oil palm plantation	-0.00	11.35	< 0.01
	TWR	Intercept	-3.53
		moon brightness rate	0.00	2.41	0.12
		distance from the oil palm plantation	-0.00	6.03	0.01
	DVCA	Intercept	-5.46
		moon brightness rate	-0.00	0.42	0.52
		distance from the gravel road	0.00	0.02	0.89

Sambar deer	LKWS	Intercept	-5.79
		moon brightness rate	0.00	0.49	0.48
		distance from the oil palm plantation	0.00	0.00	0.98
	TWR	Intercept	-2.08
		moon brightness rate	-0.02	14.36	< 0.01
		distance from the oil palm plantation	-0.00*	3.54	0.06
	DVCA	Intercept	-5.55
		moon brightness rate	0.00	0.15	0.70
		distance from the gravel road	-0.00	0.05	0.82

No competing interests reported.

SupplementaryTable1.docx

Behavioral response of Bornean ungulates, including bearded pigs and sambar deer, to anthropogenic disturbance in Sabah, Malaysia

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusions and outlook

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1