Assessment of medium and large‐sized mammals and their behavioral response toward anthropogenic activities in Jorgo‐Wato Protected Forest, Western Ethiopia

Abstract Medium and large‐sized mammals of Jorgo‐Wato Protected Forest have not yet been documented though the forest established before four decades. Hence, this study aims to document medium and large mammals and the behavioral responses of selected mammals toward anthropogenic activities in the study area. The study was conducted from February 2015 to June 2016, encompassing the wet and dry seasons. Data were collected mainly through camera traps, indirect and direct evidence. The study revealed about 23 medium and large‐sized mammals that belong to seven orders namely Bovidae, Carnivora, Primates, Rodentia, Tubulidentata, Lagomorpha, and Hyracoidea. Papio anubis, C. guereza, and C. aethiops were the most abundant large mammals in JWPF. Because of high anthropogenic activities, African buffalo shifted its activity period from diurnal into crepuscular and nocturnal. African buffalo traveled longer distances during the wet season (mean = 14.33 km, SD = 1.25 km) than during the dry season (mean = 9.00 km, SD = 2.16 km). This could be due to the fact that the local people were less likely to go to the forest for resource exploitation during the wet season as they are fully engaged in agricultural activities. However, low agricultural activities during the dry season allow the local people to extract resources and involve in bushmeat hunting which could limit the movement of mammals to their refugia. African buffalo preferred to rest on and adjacent to a gravel road (22.1%) in the forest, followed by on open rocky hilltops (14.7%) at night time, but rest in the bottomland thicket vegetation during the dry daytime. Regardless of high human pressure in the area, this study has revealed a good number of medium and large‐sized mammals that could be used as baseline information to design a sound conservation and management action plan of large mammals and their habitat in Jorgo‐Wato Protected Forest.


| INTRODUC TI ON
Mammals are important for the proper functioning of an ecosystem.
They play a fundamental role in nutrient cycling (Doughty et al., 2016), recruitment of plants (Snyder et al., 2006), pollination, monitoring the structure and composition of vegetation, and seed dispersal (Jordano et al., 2007). They are also important in fulfilling the needs of humans such as cloth, food, and spiritual values (Boesch et al., 2017). However, mammals are severely affected by habitat loss, overexploitation, invasive species, and climate change (Pacifici et al., 2017). As a result, the extinction of animals in protected areas may affect ecosystem processes in ways that we do not yet comprehend (Boddicker et al., 2002). Therefore, it is imperative to document and monitor mammalian species in and around protected areas to plan on their future conservation and management activities (Nichols & Williams, 2006). The presence or absence of mammals, their distribution, and abundance in different areas can be assessed by various methods (Wilson et al., 1996). Most of the methods developed to survey larger mammals inhabiting open savanna or grassland habitats have been easily applied. In contrast to open habitats, investigating medium and large mammals inhibited in tropical forest habitats is difficult (Andreka et al., 1999). Hence, it is particularly challenging to locate, count, and monitor mammals in tropical forests. This could force researchers to use flexible methods to assess and monitor mammal communities in and around protected areas (Boddicker et al., 2002).
Observations of mammals in the tropical rainforest are difficult because of the thick forest, high canopies, and nocturnal activities of most animals. For such animals, indirect evidence such as footmarks, droppings, sound, and feeding remains is used to verify presence (Burton et al., 2015;Wilson et al., 1996) and used to develop indices of presence and abundance of mammals. A camera trap is a costeffective technique used to monitor secretive mammalian species in forested ecosystems (Burton et al., 2015). It is less invasive, timeconsuming, and cheaper than other methods to survey and estimate mammalian species in inaccessible areas (Burton et al., 2015;Cutler & Swann, 1999). Camera traps have been used to quantify the presence and relative abundance of rainforest mammals (Burton et al., 2015;Gast & Stevenson, 2020;Martins et al., 2007). Cameras also have been used by researchers to estimate population density and relative abundance of wildlife species (Burton et al., 2015;Nielsen & McCollough, 2009). It is also used to assess the feeding ecology and activity patterns of mammals (Frey et al., 2017;Springer et al., 2011).
Wildlife species respond to anthropogenic activities that range from behavioral to distributional changes. These changes may depend on the type, intensity, and frequency of anthropogenic activities (Gaynor et al., 2018). Anthropogenic activities are the main cause of the disturbance of large mammals (Darimont et al., 2015).
As result, many species of mammals showed a strong fear response to anthropogenetic activities (Clinchy et al., 2016;Smith et al., 2017).
Fear created due to anthropogenic activities is known to affect the behavioral and activity patterns of mammals (Suraci et al., 2019).
The intensity of anthropogenic activities varies based on season, weekdays (Monday to Friday), and weekends (Saturday and Sunday).
Human recreation can negatively affect wildlife, particularly on weekends when human activity is highest (González et al., 2006;Nix et al., 2018). Moreover, the local people commonly collect resources from protected areas during the dry season due to poor agricultural activity, which may confine mammalian activity to their local refugia (González et al., 2006;Perona et al., 2019). On the other hand, the local people participate in agricultural and other livelihood activities throughout the rainy season and on weekdays. This could reduce the impact of human-induced stresses on mammalian activity patterns (Erena et al., 2020). Such temporal change in human-animal interactions can cause changes in wildlife behavior, such as increased stress, missing foraging opportunities, lower reproductive success, avoidance of certain areas, and higher mortality (Longshore et al., 2013;Martin & Réale, 2008;Simmonds & Keay, 1997).
Diurnal species are sensitive to increased human activities on weekends and weekdays due to greater temporal variation and activity overlap with humans (Longshore et al., 2013;Roy et al., 2014).
This may alter the behaviors of mammals to adapt to increased human activities. Continuous human-wildlife interaction may lead to increased stress in wildlife, decrease foraging opportunities, reduce reproductive success, increased mortality, and avoidance of certain habitats by animals (Longshore et al., 2013;Martin & Réale, 2008). It also reduces the activity of individuals, increased avoidance of areas used by humans, altered behaviors, and reduced fitness (Tadesse & Kotler, 2012;Tarjuelo et al., 2015). Recently, increased anthropogenic activities in protected areas have become a global conservation concern as it poses a severe impact mostly on large vertebrate taxa (Monti et al., 2018;Spaul & Heath, 2017). Mitigating the negative impacts of anthropogenic activities on mammals becomes a challenging task to conserve wildlife in protected areas (Krausman et al., 2008).
Understanding wildlife movements and habitat use are critical for species conservation and management on a landscape scale (Allen & Singh, 2016). Movement data can be used to evaluate the response of animals toward anthropogenically disturbed habitats as means of behavioral adaptions and survival (Carbone et al., 2005). Distance traveled by animals can be linked to energy use, risk avoidance, resource use, and the extent of anthropogenic disturbances (Fahrig, 2007). Tracking distance traveled by mammals is very important to address the various issues related to anthropogenic disturbances.
Distance traveled by mammals in protected areas is correlated with the different weekdays, weekends, and seasons because humans' access to protected areas for recreation and resource harvest varied temporally (Noonan et al., 2019). Therefore, this study also aims to examine activity periods of common mammalian species captured by camera traps, and the effects of weekends, weekdays, and seasons on movements of African buffalo in the area (Figure 1).

| Methods
This study has focused on the medium and large-sized mammal community composition of the study area. Medium mammals are those whose body weight is between 2 and 5 kg, and large-sized mammals are those over 5 kg body mass (Njoroge et al., 2009). Indirect and direct evidence and camera traps were employed to assess medium and large-sized mammal species, their activity patterns, and responses of African buffalo (the largest mammal in the area) to anthropogenetic activities in JWPF. During January, April, and May 2017, 8 HC500 HYPERFIRE camera traps were installed and stationed along common animal trails, salt licks, water holes, and other clear areas where mammal pathways were evidenced. Camera stations were changed and rotated weekly for a total of 56 days for each camera. All camera traps placed at different stations were about 5 km far from each other to make an independent sampling location. At each station, a camera trap was set at the height of about 1 m above the ground and positioned slightly downward ( Figure 3). All cameras were adjusted to take three photographs per trigger with an interval of one second between pictures. Only one out of the three photographs captured during each trigger was considered for the estimation of relative abundance and activity patterns of mammals. If the same mammal was assumed to be captured more than three times at a time, the remaining photographs were rejected.

| Data analysis
Relative abundance for each species was calculated by summing up the number of individuals recorded in all transects using Microsoft excel 2016. The abundance of medium and large-sized mammals was computed as the total number of individual species/total number of species observed in the sampled habitat multiplied by a hundred.
The frequency of occurrence of medium and large-sized mammals was described as common (C), frequent (F), occasional (O), and rare (R) (Zerihun et al., (2012). Moreover, the conservation status of mammals identified from JWPF was described as vulnerable and least concerned as per the IUCN Red List of 2008. Differences in the distribution of signs attributed to mammal observations in different habitats and African buffalo resting preferences were tested by chisquare test of association.

| Species composition of mammals
Twenty-three medium and large-sized mammals were recorded in Jorgo-Wato Protected Forest using a camera trap, direct and indirect evidence (Table 1). Among these, 6 mammal species were evidenced by a total of 464 photographs captured by a camera trap during the study periods. The recorded mammals were categorized into seven orders which include the following: the order Bovidae, Carnivora, Primates, Rodentia, Tubulidentata, Lagomorpha, and Hyracoidea.
Among all orders, the maximum species were recorded for the order Carnivora (8 species), followed by the order Bovidae (6 species) and Primates (5 species). The order Hyracoidea was represented by two species, whereas one species was recorded each for the order Rodentia, Tubulidentata, and Lagomorpha. Most of the recorded mammals were detected by indirect evidence such as droppings and footprints. Except for nonhuman primates, all the detected mammals were rarely seen during the daytime in both seasons.

| Relative abundance and conservation status of mammals
During this study period, a total of 877 individual mammals were recorded in JWPF. Among these, Olive baboons were the most abundant comprising 21.55% of the recorded individual mammals followed by Colobus monkey (18.81%) and Grivet monkey (6.42%), F I G U R E 2 Map of the study area F I G U R E 3 A camera trap fixed at 1 m height during the study period respectively. Besides, rocky hyrax (0.34%) and caracal (0.34%) were the least abundant mammalian species followed by a leopard (0.46%) ( Table 2).
Similarly, among the recorded mammalian species, about 91.30% were categorized as least concern and the remaining 8.70% species were categorized as vulnerable (Table 3).

| Activity periods of selected mammals
Activity periods of six mammals captured by camera trap are summarized in  During both seasons, the dense forest was less frequently utilized by mammals and dense riparian forest (10.80%) during the wet season (Table 5). The distribution of signs attributed to mammal observations in different habitat types during the wet and dry seasons showed significant variation (χ 2 = 22.78, df = 3, p < .05).

| Behavioral response of African buffalo to anthropogenic activities
African buffalo traveled a longer distance during the wet season (mean = 14.33 km, SD = 1.25 km) than during the dry season (mean 9.00 km, SD = 2.16 km) ( Figure 5). However, there was significant difference in the distance traveled by African buffalo during the wet and the dry seasons (χ 2 = 11.29, df = 1, p < .05). The mean total distance traveled by African buffalo during weekdays and weekends of the wet season was mean = 14.14 km, SD = 2.10 km, and mean = 13.57 km, SD, 1.2 km, respectively. The total distance

| Resting preferences of African buffalo
During the present study, a total of 402 typical resting sites (wet

| Species composition of mammals
A survey of mammals in protected areas is very important to plan and determine the future conservation plan of species and their habitats. As revealed by Jackson et al. (2006), documenting the presence or absence of species of conservation concern is of great interest for ecologists to design future conservation strategies. In this study, an assessment of medium and large-sized mammals con- Traglaphus scriptus Common Bushbuck Occasional Least concern Sylvicapra grimmia Common duiker Occasional Least concern Potamochoerus larvatus Bushpig Frequent Least concern Hylochoerus meinertzhageni Giant forest hog Occasional Least concern Phacochoerus africanus Warthog Occasional Least concern Crocuta crocuta Spotted hyaena Frequent Least concern Canis aureus Common jackal Occasional Least concern Civettictis civetta African civet Occasional Least concern Panthera pardus Leopard Rare Vulnerable

Felis caracal
Caracal Rare Least concern Felis sylvestris African wild cat Occasional Least concern Felis serval Serval cat Rare Least concern Helogale parvula Common dwarf mongooses

Occasional Least concern
Papio anubis Olive baboon Common Least concern Colobus guereza Colobus monkey Common Least concern Chlorocebus aethiops Grivet monkey Common Least concern Cercopithecus mitis Blue monkey Frequent Vulnerable

Histrix cristata Crested porcupine Occasional Least concern
Orycteropus afer Aardvark Occasional Least concern Lepus starcki Stark's hare Occasional Least concern Procavia capensis Rock hyrax Rare Least concern Hetrohyrax brucei Bush hyrax Frequent Least concern TA B L E

| Relative abundance and conservation status of mammals
Papio anubis, C. guereza, and C. aethiops were the most abundant mammalian species in JWPF. These species are diurnal and have adapted to feed on diverse sources of food items. As reported by Johnson et al. (2012), the adaptation of these species helped them to widely distribute across different habitat types in Africa. The families of Cercopithecoidae and Colobidae are known to inhabit the forested ecosystem of long and dense trees. The distribution of these species mostly ranges from savanna grassland habitats to montane forests of Africa (Kingdon, 2003). Among the identified mammals in the study area, blue monkey and leopard are categorized under the vulnerable status of IUCN Red Data list, whereas the remaining species were categorized as least concern (IUCN, 2008). Blue monkeys are very sensitive to habitat destruction and fragmentation, and their population and range are dwindling from time to time.
This could be attributed to the unlimited access of local people into protected areas for the collection of forest and nonforest products.
Leopard was categorized as vulnerable probably due to reduced home range, prey, and other anthropogenic activities in the area.

| Activity periods of selected mammals
Higher camera trap capture frequency was recorded for larger mammals such as African buffalo during the nighttime and for nonhuman primates (Olive baboon) during the daytime. This could be linked to the increased activity and home range size of large mammals. As described by Stirrat (2003), capture frequency increased as the activity, and home range size of mammals increased. Though nocturnal activity is a strategy used against predation and the advantage of utilizing underutilized food niches (Gómez et al., 2005), African buffalo restricted its activity to nighttime because of increased human activities during the daytime in JWPF. However, the high capture frequency recorded for Olive baboons could be linked to their extended group living style and large troop size. In this study, only six out of 23 mammal species were captured by a camera trap in the JWPF. In addition, only 11 out of 23 mammal species were recorded both during the wet and dry seasons. This could be ascribed to the low density of each mammal and the thick forest of JWPF (Jiménez et al., 2010). Moreover, relatively small-sized mammals were less detected and have low capture frequency as compared to large mammals (Kelly & Holub, 2008). This could be due to the limited movement and small range use of medium mammals compared to larger mammals. Incidences of mammal records were more during the dry than the wet seasons. This could be attributed to the reduction of resources during the dry season has increased the movement of mammals in order to fulfill their nutritional requirements (Jiménez et al., 2010).
The activity period of S. caffer, C. crocuta, C. civetta, and Helogale parvula were confirmed to be nocturnal. However, the activity period of P. anubis was diurnal as the photographs were entirely taken during the daytime. Though photographs of all mammals were not captured by camera traps, most of the captured mammals seem to be nocturnal and crepuscular except nonhuman primates. This could be ascribed to unrestricted human access into the forest during the daytime, and poaching for bushmeat and other resource extractions.
In JWPF, it seems that extensive resource extraction from the forest has forced most mammals to shift their activity periods to nocturnal and or crepuscular. For instance, the activity period of S. caffer was mainly diurnal where there is no human disturbance, but it completely shifted into nocturnal and crepuscular in JWPF.

| Habitat association of mammals
All habitat types do not possess balanced resources and have not been equally used by ungulates throughout the year (Bjørneraas et al., 2011). Consequently, habitat utilization by ungulates varies on a seasonal and circadian basis (Demarchi & Bunnell, 1995;Dussault et al., 2004) and is governed by trade-offs between associated costs and benefits (Rettie & Messier, 2000 (Prins, 1996). However, F I G U R E 6 Distance travelled by African buffalo during weekdays and weekends in response to anthropogenic activities in Jorgo-Wato Protected Forest

F I G U R E 7
Recorded resting sites of African buffalo during the wet and dry seasons in JWPF large mammals such as African buffalo have shifted their foraging period into nocturnal and or crepuscular mainly due to high human disturbances in JWPF as reported by Di Bitetti et al. (2008). Di Bitetti et al. (2008) also stated that increased poaching pressure can alter activity patterns of hunted species into nocturnal. As reported by Skinner and Smithers (1990), African buffalo graze closer to rivers and take shelter in the thick riverine vegetation during the dry season. Hence, the difference in habitat association of mammals in the present study area might be mainly a functional response to anthropogenic activities in the forest. An increased habitat association of mammals to riparian habitats could be linked to risk avoidance in the open forest. To optimize cost-benefit relationships, ungulates may use habitats with good cover during the daytime (Demarchi & Bunnell, 1995;Dussault et al., 2004) and visit open forage-rich habitats during the night as they are less visible to humans (Godvik et al., 2009;Lykkja et al., 2009). In the present study, mammals did not prefer dense forest due to reduced forages as reported by Perrin and Brereton-Stiles (1999).

| Behavioral response of African buffalo to anthropogenic activities
African buffalo traveled a longer distance during the wet season compared to the dry season. This could be due to the fact that the local community is less likely to go to the forest for resource extraction during the wet season as they are always busy with agricultural activities. However, the dry season is the resting period for the local people due to low agricultural activities. Consequently, the local people frequently visit the forest for resources extraction and bushmeat hunting thereby limit the movement of large and shy mammals such as African buffalo to their refugia. Moreover, African buffalo relatively traveled longer distances during weekdays and weekends of the wet season but low during the same periods of the dry season.
The long resting time during the dry weekdays and weekend provides the local people an opportunity to frequently visit the forest.
This could limit the activity of mammals to their refugia during the daytime. However, during the wet season, the local people are fully engaged in agricultural activities and are less likely to visit the forest for resource extraction. As described by Nix et al. (2018), the movement of animals is very slow in areas where human activities are high. In such areas, animals use contrasting activity periods with humans to avoid risks. Animals travel longer distances and spent more time away from their refugia in areas where human disturbance is low or absent (Perona et al., 2019).

| Resting preferences of African buffalo
The present study has revealed that buffalo preferred to rest on gravel roads, open rocky hilltops, and clearings during the wet season. These resting sites were used during the night as these sites  (Fonkwo et al., 2011;Obioha et al., 2012). Contrary to Jorgo-Wato buffalo, forest buffalo rest in the forest at night and in clearings during the day (Melletti et al., 2007). Consequently, they might have been more susceptible to poachers. However, Jorgo-Wato buffalo preferred to rest more in the thicket vegetation and at the base of montane forests and hilly terrain during the daytime. This could be attributed to the avoidance of risks encountered in clearings and open forests.
Buffalo's preference for mountainous terrain and rocky hilltops during the wet season could be linked to thermoregulation as revealed by Harris et al. (2002). As stated by Harris et al. (2002), areas of high ground are warmer than bottomland at night and hence preferred for nocturnal resting by herbivores.

| CON CLUS ION
JWPF is a potential wildlife habitat for diverse arrays of medium and large mammals in the western parts of Ethiopia. However, studies on JWPF in general and mammals, in particular, have not yet been well documented because of the inaccessibility and remoteness of the area. In this study, a combination of field survey techniques such as camera traps, direct observation, and indirect evidence was employed to document the medium and large mammal species composition of JWPF. A total of 23 species of medium and large mammals were recorded in the study area during the wet and dry seasons.
Most of these mammals were recorded by camera traps and indirect evidence because they were rarely observed due to human distur- Enterprise and my field assistants for their support during the study periods.

CO N FLI C T O F I NTE R E S T
There are no conflicting interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data used in this work are archived in Dryad Digital Repository, https://doi.org/10.5061/dryad.6hdr7 sr2g.