Research

INTRODUCTION

The African lion (Panthera leo) is the largest terrestrial carnivore in Africa and an iconic species representing the essence of wildness (Funston et al. 2017). Lion populations are declining fast, with illegal killing, habitat loss and fragmentation, and bush meat poaching of their prey resulting in a massive decline of their numbers over the past 20 years (Beattie et al. 2020). As ecosystems become more fragmented, most protected areas are becoming too small to support wide-ranging carnivores like lions, and such species are forced to use adjacent areas for supplementary food and other needs (Laizer et al. 2014). According to Tumenta et al. (2013), lions have vanished from more than 90% of their historic range, and their home range sizes and space use patterns have changed due to such pressures.

The home range size of carnivores such as lions is generally as large as is necessary but as small as possible to satisfy energetic needs (Laizer et al. 2014). Various factors may influence lion home range sizes. For instance, in Hwange National Park, Loveridge et al. (2009) found that prey availability and distribution significantly affected the home range size of lions. The authors further asserted that home range size varies with sex, with home ranges of lionesses mainly influenced by pride size and dispersion, whereas for male lions it is determined by prey biomass and female pride density. In the Serengeti ecosystem, Kittle et al. (2016) indicated that the availability and vulnerability of prey to predation both motivate lion movement decisions, with their relative importance mediated by overall prey abundance. This suggests that although top carnivores such as lions are consistently cognizant of how landscape features influence individual prey vulnerability, they also adopt a flexible approach to space use by adjusting spatial behavior according to fluctuations in local prey abundance (Kittle et al. 2016).

Nevertheless, seasonality is another important factor influencing both the home range sizes and space use patterns of lions, especially in savanna ecosystem (Briers-Louw 2017). In Tarangire National Park, Tanzania, Laizer et al. (2014) determined the home range and movement patterns for six prides using data obtained from the radio and GPS collars and established that home ranges ranged from 52 km² to 616 km² and extensively overlapped. Home ranges were larger in the wet season and smaller in the dry season. Both the movement and space use of prides was largely influenced by the migration of prey species particularly zebra and wildebeest, governed by the seasonal availability of food (Laizer et al. 2014).

The fact that lion home range size and space use differ with areas as governed by prey abundance and availability calls for site-specific studies (Briers-Louw 2017). Surveys in Chizarira National Park (CNP) have shown a drastic decline of prey abundance (Dunham et al. 2015, Youldon and Abell 2015). Based on the second prediction of the resource distribution hypothesis (RDH), which states that home range size increases with an increase in resource dispersion (Mbizah et al. 2019), it is expected that the home range sizes and the space use by resident lions in CNP would increase and expand beyond the park. Consequently, this would result in conservation challenges faced by the iconic species, considering that the park is surrounded by communities. One possible challenge is human-lion conflicts as a result of increased livestock depredation by lions (Cushman et al. 2018).

It has come to the attention of many studies that the conservation of smaller, isolated lion populations is essential in the quest to curb the current decline (Snyman et al. 2018). Understanding what factors drive lion space use underlying landscape-level distribution is essential for their conservation efforts (Kittle et al. 2016). Changes in the local environment can either disturb or enhance the habitat for lions and their prey (Laizer et al. 2014). The key environmental drivers of the population need to be understood before one can understand how their change can affect the home range and space use of a species. Indeed, understanding the spatial ecology of lions is fundamental to making informed decisions about lion conservation (Visser et al. 2009). Again, to mitigate human-lion conflicts and for effective conservation planning, an improved understanding of the space use and home range sizes of lions is crucial to ensure the continued survival of their populations. We therefore seek to determine the seasonal home range sizes and analyze the intrinsic (environmental) and extrinsic (anthropogenic) factors influencing lion space use in CNP.

METHODOLOGY

Study area

Chizarira National Park, located in the northwestern part of Zimbabwe, covers a spatial extent of 1910 km² and forms the core of the Sebungwe ecological region. The park is part of the Kavango Zambezi Transfrontier Conservation Area (KAZA TFCA), which is shared between Angola, Botswana, Namibia, Zambia, and Zimbabwe. Chizarira National Park shares a boundary with Chirisa Safari Area to the south, which stretches for about 18 km, joining the Simuchembu Communal area (Fig. 1). The park is unfenced and the communal area is found right up against the boundary. In addition to a growing human population and the spread of agricultural practices that impinge on corridors, the hard-edge effect due to lack of fences increases the occurrence of human-wildlife conflicts, such as human-lion conflict (Loveridge et al. 2020). Ninety percent (90%) of Chizarira National Park is surrounded by communities (Fig. 1).

Three main rivers drain through the park, namely: the Busi River in south Chizarira and the Mucheni and Kaswiswi Rivers in the north center region of the park. Water is found in isolated pools that last throughout the year in a normal rainy season. There are perennial springs dotted across the park that provide water for wildlife species, especially during the dry season. Chizarira National Park is typically mountainous, with elevations ranging from 400 m to 2285 m. It consists of a plateau, heavily dissected by drainage lines, sloping to the Busi/Sengwa Valley, in which the Chirisa Safari Area lies (Chizarira General Management Plan, unpublished draft, 2020–2030).

Climate

The park receives annual precipitation of approximately 700 mm, which falls during a single rainy season extending from December through to May. The temperature range varies from an average maximum temperature of 29 °C in November to an average minimum of 16 °C in July. Overall annual temperature is 23 °C. The area experiences two separate seasonal conditions characterized as: dry (June to November) and wet (December to May). The effect of the wet season is noticeable one month after the initial rains, when vegetation starts to become green and prey species begin to change their movement.

Vegetation

The vegetation consists of woodland and open grasslands (Youldon and Abell 2015), with vegetation dominated by miombo and mnondo trees, generally Brachystegia and Julbernardia, respectively, with mixed Combretum and Colophospermum mopane woodlands also common. Fire is a threat to some grassland portions of the park.

Wildlife species

The park supports different kinds of mammal species albeit at low densities (Youldon and Abell 2015) that are preyed upon by lions, including: African buffalo (Syncerus caffer), bushbuck (Tragelaphus scriptus), common duiker (Sylvicapra grimmia), eland (Tragelaphus oryx), elephant (Loxodonta africana), impala (Aepyceros melampus), greater kudu (Tragelaphus strepsiceros), common reedbuck (Redunca arundinum), sable (Hippotragus niger), warthog (Phacochoerus africanus), common waterbuck (Kobus ellipsiprymnus), and plains zebra (Equus quagga; Dunham et al. 2015). Lions in the park compete for the few prey with spotted hyenas (Crocuta crocuta) and leopards (Panthera pardus; Mann et al. 2020).

Study individuals and collaring

Between August 2017 and July 2019, wild free-ranging resident lions in Chizarira National Park were captured and fitted with global positioning system (GPS) satellite collars. This was done under the Chizarira/African Lion and Environmental Research Trust (ALERT) Lion Project (Research Permit 23(1) (C) (II) 14/2019). Strict ethical standards for the capture and handling of lions were adhered to, following guidelines from the Zimparks Manual for Live Sale, Capture and Translocation of Wildlife in Zimbabwe (2017). Before this lion collaring exercise, a survey was carried out to estimate the number of resident lions in the park. A total of 36 lions were estimated based on the biomass of buffalo alone (Monks et al. 2015).

With the help of a certified wildlife veterinarian doctor, four lions from different prides were captured and fitted with GPS collars. This was done during the night with a combination of call-up surveys and baited cage traps. Upon capture, the lions were immobilized with a combination of 0.03-0.05 mg/kg medetomidine 20 mg/ml and 0.5-1.0 mg/kg Zoletil 100 (tiletamine-zolazepam) based on estimated weight by dart injection, following the protocols by Snyman et al. (2018). Medetomidine was reversed with 0.2 mg/kg atipamezole 5 mg/ml administered intramuscularly. Captured lions were sexed and classified as yearlings, sub-adults, and adults based on tooth eruption and wear patterns.

The GPS collars were programmed to take four fixes, covering lion daytime (13:00 hrs), midnight time (02:00 hrs), the morning time (07:00 hrs), and the evening hunting time (19:00 hrs). Apart from the GPS locations, the collars also recorded date, distance traveled (m), and locational altitude (m).

All GPS collars had a very high frequency (VHF) backup mechanism. This was useful for tracking individual lions manually to retrieve the collars (in the case of mortality), or to find the lion (in case of collar failure, or a kill site). A complete set of GPS data for the study was downloaded from the Internet on the African Wildlife Tracking, South Africa site (www.awt.co.za). It was then cleaned and all irregular outlying points filtered out. These irregular outlying points are sometimes caused by satellite or recording inaccuracies. The resultant raw data collection for each pride is displayed in Figure 2.

Home range estimation

Location data from collared lion GPS fixes were used to estimate home range sizes between the dry (June-November) and wet season (December-May) for each pride using the kernel density estimator (KDE). The KDE estimates a kernel or probability density over each location point and removes outlying location points (Oriol-Cotterill et al. 2015). In this study, the KDE was chosen because it creates isopleths of intensity of utilization by calculating the mean influence of data points at grid intersections, using fixed kernel density home range estimation (Kittle et al. 2016). To clearly understand the seasonal home range utilization, two levels of territoriality were considered, i.e., the 95% range and the 50% core utilization distributions. The 95% range (excluding 5% of the outermost locational data as outliers) is the overall area used and defended to some extent by the lions. The 50% core area is the space in which the lions spend the majority of their time, which usually includes key resources such as water sources and productive hunting grounds, and is the area actively defended from any and all competitors (Mancinelli et al. 2018). The home range sizes were computed in R (R Core Team 2019) using functions from the adehabitatHR package (Calenge 2006).

The seasonal home range sizes for each pride were treated independently because the lions were collared on different dates and had different collar life spans (Table 1). For comparison purposes, the same number of months between the wet and dry season for each pride were filtered and used in the seasonal home range computations.

Species occurrence and environmental variables

Five environmental variables were purposely selected on the basis that they were landscape features and/or factors that hypothesized a priori good predictors of lion space use. These factors were analyzed to establish factors that significantly influence lion space use patterns following protocols by Snyman et al. (2018). The factors were prey species, water points, human settlements, elevation, and habitat type.

Prey species occurrence data were obtained from the Spatial Monitoring Application Reporting Tool (SMART) database maintained at CNP. For this study, prey occurrence data for buffalo, waterbuck, kudu, warthog, and zebra sightings were selected and merged into a single shapefile. The sighting data were selected and filtered to correspond with the collaring period of the four prides. Physical lion tracking was carried out using a 4 x 4 vehicle (backed with telemetry equipment) to record lion kill sites to validate lion movement and space use. The selected wildlife species were found to be preferred prey species for lions in typical savanna ecosystems (Loveridge et al. 2009, Verschueren 2017).

Data on water points (springs, isolated pools, and natural water pans) were obtained from the SMART database. Each time rangers are deployed for patrols, they mark encountered waterpoints with GPS receivers and cybertackers. These locations were exported into ArcGIS to create merged shapefile for water points. Water points were considered as an important environmental factor in this study because during the dry season, most ungulates spend time near water points, and this will in turn influence lion space use. In Hwange National Park, Loveridge et al. (2009) reported that lions spend a significant amount of time within two kilometers of a water hole, especially in dry season.

Chizarira National Park is not fenced, making the reasons for lion movement less obvious than in protected areas with predator-proof fences. For this reason, the distance from human settlements was considered to be an important factor influencing lion space use. In this study, settlements within a buffer of 20 km from the park boundary were considered. This range was selected because most incidences of livestock depredation received at CNP rarely exceed a radius of 20 km. Human settlement data were collected from the eight surrounding villages. With the help of trained lion guardians, locations of cattle kraals and villages were marked with a handheld GPS receiver (Garmin Trex 10). These locations were exported into ArcGIS to create a merged shapefile for settlements.

Using the Euclidean distance function in ArcGIS, these environmental variables used to understand the distance to a feature, i.e., distance from prey species, distance from water point, and distance from settlement (buffer of 20 km from Chizarira National Park) had their distances calculated before being converted into distance raster layers as in Chinoitezvi et al. (2024). Figure 3 shows the environmental variables used in the modeling.

A study by Zehnder et al. (2018) found that different habitat types influence prey availability and hunting success, which in turn shapes the lion’s space use. Data on habitat types were downloaded from Landsat 8 Operational Land Imager (OLI) and thermal infrared sensor (TIRS) images (bands 2, 3, 4), with a spatial resolution of 30 m, for April 2020. April was selected for analysis because cloud cover was less than 10%. Classification of habitat types was done using supervised classification by maximum likelihood in ArcMap 10.3. Supervised classification entails selecting sample pixels that represent specific classes and then directing the image processing software to use these as training sites for the classification of similar pixels in the image. The training sites were selected based on the researcher’s knowledge of the sites. Google Earth imagery was also used to aid with the accuracy of training samples.

The created training sample areas were then selected and used to identify each habitat type class, complemented by the base map of the area, control points, and the researchers’ knowledge of Chizarira National Park. Maximum likelihood classification was then performed. It was assumed that training pixels were normally distributed and the probability that the selected pixels belonged to a specific habitat class would be calculated. For this study, five habitat type classes were produced (Fig. 3b).

Elevation data were downloaded free from the United States Geological Survey portal on a spatial resolution of 30 m-by-30 m digital elevation model (DEM) layer obtained from the Shuttle Radar Topography Mission (SRTM; Chinoitezvi et al. 2024). The terrain for CNP is characterized by deep valleys and escarpments and is rugged and mountainous therefore elevation was considered an important environmental factor influencing lion movement and space use.

Multicollinearity between environmental variables

In this study, raster files representing various environmental variables, including water points, human settlements, habitat type, prey species, and elevation, were obtained and loaded into R for analysis. The raster layers were stacked to facilitate the extraction of values, which were then converted into a data frame. To ensure clarity, the columns of the data frame were appropriately renamed. The dataset was examined for missing values, which were to be addressed either by removal or imputation, ensuring a complete dataset for subsequent analysis. A correlation matrix was computed to assess multicollinearity among the environmental variables with particular attention paid to correlation coefficients above 0.7 or below -0.7, indicating potential multicollinearity. To visualize the relationships among variables, a heatmap of the correlation matrix was generated using ggplot2. species distribution modeling (Fig. 4).

The maximum entropy (MaxEnt) modeling technique was used to analyze the selected environmental variables. First, environmental variables were checked for conflating or confounding using the variance inflation factor (VIF) in ArcMap 10.3 following protocols by Muposhi et al. (2016). Second, because species distribution-based modeling has a common problem of spatial autocorrelation, Moran’s I under the spatial statistics tool in ArcGIS was computed to address this. All environmental layers were converted to raster format with an identical cell size. Environmental variables used to understand the distance to a feature (distance from water points, distance from settlements, and distance from prey species) were converted into distance raster layers. Elevation and habitat type were the continuous and categorical variables used in this study, respectively.

All the data sets for the environmental variables were exported as American Standard Code for Information Interchange (ASCII) files before being run in MaxEnt software. Further, before analysis, the coordinates of the selected datasets were converted from WGS84 to the projected spatial reference system of Chizarira/UTM zone 35S, using the data management and projection tool in ArcMap 10.3. Validation performance of the MaxEnt model was done using the area under the receiver operating curve (AUC) concept. The AUC values range from 0 to 1, in which AUC values less than 0.65 indicate random to poor models, whereas values greater than 0.65 indicate good to high performance models (Phillips et al. 2006).

Statistical analysis

To test if there were significant differences between the wet and dry seasonal home range sizes (95% and 50% contours) for the study prides, the prop.test in R-Studio was run at a 95% confidence interval. The prop.test is used for testing the null hypothesis that proportions between groups are the same or that they equal certain given values.

RESULTS

Multicollinearity test between environmental variables

The heatmap displays correlation coefficients among environmental variables, revealing no perfect correlations among the variables, with the highest being 0.17 between habitat type and prey species (Fig. 4). Most correlations are weak, indicating minimal relationships among variables, such as water points and human settlements (0.05), and water points and habitat type (0.03). The absence of coefficients exceeding 0.7 suggests no significant multicollinearity, hence all variables were considered suitable for our study, allowing for the inclusion of these variables in further analyses without concern for inflated standard errors. Overall, the results indicate that the environmental variables measure distinct aspects of the ecosystem.

Seasonal variation in the home range size of collared lions

The home range sizes varied from 95.99 km² to 316.53 km² and from 50.53 km² to 706.89 km² (95 % KDE) in dry and wet season, respectively. Similarly, the core home ranges (50% KDE) varied from 16.69 km² to 78.02 km² in the dry season and from 4.08 km² to 130.65 km² in the wet season. All prides had their home range sizes increase in the wet season, except for SAT2169, whose wet season home range shrank to almost six times its dry season home range (Table 2).

Although there was no collared male lion with the same temporal extent as with SAT2168, this male sub-adult lion (Table 2) had the largest home range size when compared to other individual collared prides. His seasonal home range size (95% KDE) was highly significant (prop.test, p < 0.000). Despite an increase in his core home range area (50% KDE) in the wet season, statistically this increase was insignificant (prop.test, p > 0.05). Additionally, zero is contained in the 95% confidence interval range, reflecting an insignificant increase (Table 2).

Considering the influence of the sample size and the time span over which fixes have been recorded, seasonal home range size also differs between sexes. Males (SAT2168 and SAT2169) seem to have established a larger home range size (95% KDE) compared to females (SAT2170 and SAT2171) whose core areas (50% KDE) doubled the size of their dry season core areas in wet season. For example, the core home range of SAT2170 rose sharply from 44.76 km² in the dry season to 130.65 km² in the wet season. This increase was highly significant (prop.test, p < 0.000). Similarly, the 95% confidence interval (for the proportions) does not contain zero in the range, i.e., from the lower to the upper value (Table 2).

The findings in this study showed that the home ranges (95% KDE) were largely within the park boundaries, except for SAT2168 (Fig. 5a), whose wet season home range extended outside the western park boundary. The home range for the four prides expanded in the wet season and shifted toward the northern and western edges of the park. However, none of the prides established home ranges in the northeastern and southern parts of the park (Fig. 5). When overlaid with the environmental variables used in this study (Fig. 6), clearly indicate that the core home ranges of all the studied prides were established in areas of high species prey occurrence and where water resources were abundant.

The observed home ranges extensively overlapped between seasons (n = 2) and across individual lion prides (n = 4; Fig. 5). Some prides (SAT2168 and SAT2170) had established more than one core home range in different areas within the park. However, only lion SAT2168 had established its core home range close to the park edges (Fig. 5).

The established seasonal home ranges varied considerably across the four collared prides in Chizarira National Park (n = 4). Larger prides seemed to establish small home ranges (95% KDE) compared to smaller prides (Table 2, Fig. 5). For example, SAT2168, which comprised 3 coalition sub-adult males, had established a wet home range size (95% KDE) that was almost six times bigger than that of SAT2171 (15 individual lions). In contrast, smaller prides seemed to establish very small core home ranges relative to their range use. Overall, SAT269 and SAT2171 had established the smallest core area (50% KDE) and range use (95% KDE) sizes of all the studied prides, respectively (Table 2).

Model performance

The calculated receiver operating characteristics (ROC) confirmed that the model fitted well to test data of distance from water points, distance from settlements, distance from prey species, habitat types, and elevation (Test AUC = 0.76; Fig. 6). From the graph, the test data line is not five close to the random prediction line, indicating that the model was successful enough to explain how six of the selected five factors influence lion space use in Chizarira National Park (Fig. 6).

Factors influencing lion space use

The Jackknife results are useful in checking the importance of individual variables to the model. The dark blue and light blue bars represent the change in gain when the model is run with the corresponding variables only or without them, respectively. The red bar is the total gain with all 16 variables tested. Results from the Jackknife of AUC for each variable contribution indicated that prey species, elevation and water points, respectively, contain the most important information to influence the space use patterns of lions in CNP. Conversely, distance from settlements and habitat type contain the information (Fig. 7).

Additionally, response curves derived from the MaxEnt model showed a decreased probability of lion presence with increase in distance from the prey species (Fig. 8a). Areas with settlements closer to the park boundary experienced low probability of lion presence (Fig. 8b). Results showed higher probability of lion presence in altitude of 1000 m to zero probability in either side of the peak (Fig. 8c). There was higher probability of lion presence in areas close to water points (< 3000 m), with the probability decreasing with increase in distance from water points (Fig. 8d). Of the five habitat types, results showed a higher probability of lion presence in escarpment areas, open to closed habitats (Jesse) and riverine areas. There was lower probability of lion presence in deep valleys (Fig. 8e).

Similarly, distance from prey species (75.9%), elevation (13.3%), and water points (6.5%) contain much information while the influence of settlements (3.4%) and habitat type (0.9%) explained the least variation in the model. The permutation importance of each environmental variable also varied accordingly (Table 3).

Results showed higher probability of lion presence in altitude of 1000 m to zero probability on either side of the peak (Fig. 6c). There was a higher probability of lion presence in areas close to water points (< 3000 m), with the probability decreasing with an increase in distance from water points (Fig. 6d). Of the five habitat types, results showed a higher probability of lion presence in escarpment areas, open to closed habitats (Jesse) and riverine areas. There was a lower probability of lion presence in deep valleys (Fig. 6e).

DISCUSSION

Spatial and temporal heterogeneity in resource availability partially underlies seasonal shifts in herbivore distribution patterns (Muposhi et al. 2016) and home range use (Visser et al. 2009). Home range sizes of the studied prides were hypothesized to show no seasonal differences between wet and dry seasons. Results in this study showed a variation in home range sizes between the dry and wet seasons of individual collared prides in Chizarira National Park. The home range sizes varied from 50.53 km² to 706.89 km² (95% KDE), with the smallest and largest home range sizes recorded in dry and wet seasons, respectively.

The results of the present study concur with other studies in savanna ecosystems, where dry-season home range sizes were smaller than wet-season home range sizes (Laizer et al. 2014, Snyman et al. 2018). A possible explanation for this could be seasonal variation in the occurrence and distribution of prey species. For instance, in the dry season, sedentary prey species usually congregate around surface water points, and lions subsequently decrease their ranges by concentrating around these water points, therefore reducing home range sizes (de Boer et al. 2010). Conversely, due to a wider abundance and distribution of water sources during the wet season, prey tends to be widely distributed. As a result, prey density is often less dense in high-risk areas such as water points during the wet season, and this would result in larger home ranges in the wet season (Lehmann et al. 2008, Laizer et al. 2014). However, in Waza National Park, Cameroon, Visser et al. (2009) reported that lions spent much of the dry period outside the park boundaries and their home ranges increased during this season.

The core home ranges (both dry and wet seasons) of other resident prides (SAT2170 and SAT2171) recorded in this study overlapped extensively with other predators recorded recently in the park (Mann, 2020). This was also confirmed during the lion collaring exercise, in which a combination of call-up surveys and baiting was used; several other carnivores responded (Fig. 5). These results suggest some degree of spatial co-existence between guild members. Although these results may reflect an ecological paradox that lion movement might be unconstrained by the spatial patterns of interspecific competitors, there is a need for studies on intraguild competition in Chizarira National Park.

In Kenya, Ogutu et al. (2016) found out that the size of the home range and distance traveled by lions negatively correlated with the density of native prey and available surface water. However, this varied with sex of individuals because male lions tend to have larger home ranges than females (Visser et al. 2009, Tumenta et al. 2013). In this study, despite the fact that the sample size was small (n = 4), the home range sizes seem to differ between sexes as well. The reasons behind this scenario are not immediately obvious, but it is possible that male lions are more likely to depend on both food resources and the need to defend and access female prides (Elliot et al. 2014), whereas female ranges are configured around access to resources (Schaller 2009). The differences could be caused by both behavioral and physical characteristics of males and females because males are much bigger than females, which results in higher energetic needs (Schaller 2009). With regard to a bigger home range of young male lions compared to the adult lions, this could be due to a combination of factors. As they mature, especially during the age of 2 to 3, young male lions are forced to leave their natal pride, leading to a period of exploration and dispersal as they search for a new pride or establish their own territory (Packer and Pusey 1982). During this time, they cover larger distances in pursuit of potential mates, reproductive opportunities, and resources, while also avoiding areas with high concentrations of dominant males to minimize competition and conflict (Grinnell and McComb 1996). As a result, young males exhibit a more extensive and dynamic home range behavior, which is shaped by their unique life-history stage and the challenges associated with establishing dominance and securing mating opportunities.

The study hypothesized that selected environmental predictors (prey species, elevation, settlements, water points and habitat type) invariably influence the space use patterns of lions in Chizarira National Park. We provided evidence that prey species occurrence, elevation, and water points significantly influence lion space use in Chizarira National Park. The MaxEnt model results indicated that distance from prey species was the most important environmental variable with the highest gain when used in isolation to explain space use by lions in CNP. These results are similar to a recent study in Hwange National Park by Mbizah et al. (2019) who stated that the distribution of large carnivores, such as lions, is determined by the availability and occurrence of suitable prey. In that particular study, the probability of lions selecting a particular area decreased with increasing distance from prey species. However, Hayward and Kerley (2005) argued that the occurrence and distribution of prey species alone cannot clearly explain the choice of habitat by lions due to the complexity of decision mechanisms related to resource selection functions. Generally, some of the prey species have features that reduce predation either morphologically (e.g., sable horns), ecologically (e.g., roan and sable occurring at low density), or behaviorally (e.g., the large herd size and increased vigilance of eland). Nonetheless, in CNP, these species occur at very low densities and were not recorded from the kill sites that were identified during the period of our study.

During the study period, it was observed that lions were sometimes snared in pursuit of prey species such as buffalo and kudus. According to Loveridge et al. (2020), lions appear to be highly vulnerable to being snared as accidental by-catch because of their similar size to target animals (large-medium sized herbivores) and due to their tendency to be attracted to the carcasses of animals caught in other nearby snares.

In a multi-predator landscape like CNP, the abundance and distribution of inter-specific competitors also influence decisions on space use by lions (Kittle et al. 2016). For example, in many African savanna ecosystems, lions and hyenas are the most important predators and are potentially strong direct competitors given that their diet and ecological range extensively overlap (Funston et al. 2001, Kittle et al. 2016). These two carnivores typically exhibit negative interactions in the form of direct aggression and kleptoparasitism as they compete for the same suite of prey resources. Surveys that were carried out in Chizarira National Park revealed that hyenas are more abundant than resident lions in the park (Monks et al. 2015). Recent sightings by law enforcement rangers confirm the presence of cheetahs in the park, which potentially compete for the same prey species. The influence of interspecific competitors on lion space use was however beyond the scope of this study.

Protected areas such as national parks (NPs) often provide refuge for lions from anthropogenic disturbances while conserving high-quality habitat. However, the ability of NPs to do so is hampered by increasing human pressures (e.g., poaching, habitat fragmentation, and livestock encroachment; Mills et al. 2020). Areas far away from the park boundary were surprisingly found to be more utilised by lions in Chizarira National Park (Fig. 7b). This could have resulted from pride SAT2168, which moved from Chizarira to Hwange National Park via many settlements along its way (Fig. 3). Furthermore, the effect of human settlement encroachment, particularly on the northeastern (in and close to the park boundary) and southern (close to the park boundary) cannot be ruled out with absolute certainty in influencing lion space use patterns in CNP. Due to limited time and differences in the collaring dates of the studied prides, the comparison of individual pride space use was not investigated.

The result that lions prefer to occupy and use communal areas where settlements are scattered across the landscape is not unique to CNP. Earlier studies by Mills et al. (2020) in West Africa also found lions exhibited no avoidance of hunting concessions despite higher human occupancy than in neighboring national parks. A possible explanation for this could be the lions’ heavy reliance on depleted prey populations and preferred habitat characteristics in communal areas. Indeed, this may call for potential management interventions to indirectly influence lion space use in national parks through the enhancement of water availability (Muposhi et al. 2016) and by improving corridor connectivity with other surrounding land uses (Cushman et al. 2018), thus reducing human-lion conflicts.

One other key reason why settlements close to CNP have a lower probability of lion presence could be good animal husbandry practices in these villages. Those villages close to the park boundary, e.g., Sinasengwe and Sinampande (Fig. 5) have received training on how to deal with and reduce livestock depredation from NGOs such as ALERT, in their respective wards. A total of four mobile cattle bomas were constructed in the wards, providing strong predator proof areas from livestock depredation. The recent creation of the Mucheni Conservancy in the Sinasengwe area has seen many villagers benefiting from various human-wildlife conflicts (HWC) mitigation measures training, including the erection of mobile cattle bomas (Sinasengwe Ward 4 councilor, personal communication). Elsewhere, mobile cattle bomas were found to be effective methods of deterring and reducing livestock depredation by lions in many human-dominated landscapes, e.g., in Waza National Park, Cameroon (Visser et al. 2009), Zambezi Region, Namibia (Moeller 2014), Nairobi National Park, Kenya (Verschueren 2017), and Northern Tanzania (Beattie et al. 2020).

Results showed a high probability of lion presence in areas close to water points, particularly within a 3 km radius of water sources. These results are consistent with the findings by Valeix et al. (2010) who suggested that lions adopt area-restricted searching in the vicinity of waterholes and reduce their search effort to minimize the time spent far from a waterhole in Hwange National Park. Surface water availability offers greater opportunities for encountering prey seeking water and suitable food resources, particularly during the dry season (Lehmann et al. 2008). Similarly, the prides under study established their core home ranges in areas close to rivers and water points. It was also evident from the response curves that the riverine habitat type explained a higher probability of lion presence in CNP.

The establishment of home range sizes by lions, particularly core areas, can be related and inferred to by lion habitat preferences, which management can use and optimize to regulate herbivore (lion prey species) distribution or water provisioning to retain the occurrence of heterogeneous wildlife habitat. An understanding of lion spatial ecology is therefore imperative for lion conservation to (1) inform management on the key drivers that determine lion space use, (2) help in drawing up robust human-lion conflict mitigation plans based on movement and space use data, and (3) identify of wildlife dispersal routes.

CONCLUSION

This study was based on two hypotheses, (1) the home range size of resident lions would be larger in the wet season compared to the dry seasons, and (2) environmental variables and human factors (i.e., prey species, water points, human settlements, elevation, and habitat type) invariably influence the space use patterns of wild free-ranging satellite collared lions in CNP.

Results of this study showed that home ranges varied across the four studied prides. The home range sizes varied significantly because small areas were traversed in the dry season compared to the wet season. Therefore, it is concluded that the home range sizes of resident lions in Chizarira National Park is greatly influenced by seasons, and that they vary with pride composition and sex.

Results also indicated that prey species occurrence (75.9%), elevation (13.3%), and water points (6.5%) contained the most important information to explain the lion space use patterns in CNP. It can be concluded that in heterogeneous habitats, lion space use is variably influenced by environmental factors, with lion space use motivated by prey species presence and occurrence. These results suggest that although top carnivores are consistently cognizant of how landscape features influence individual prey species occurrence, they adopt a flexible approach to space use by adjusting spatial behavior according to fluctuations in local prey abundance and distribution.

LITERATURE CITED

Beattie, K., E. Olson, B. Kissui, A. Kirschbaum, and C. Kiffner. 2020. Predicting livestock depredation risk by African lions (Panthera leo) in a multi-use area of northern Tanzania. European Journal of Wildlife Research 66:11. https://doi.org/10.1007/s10344-019-1348-5

Briers-Louw, W. D. 2017. Ecology of apex predators in Majete Wildlife Reserve, Malawi. Thesis. Stellenbosch University, Stellenbosch, South Africa. https://scholar.sun.ac.za/server/api/core/bitstreams/1cee7c64-4b5b-4adb-b5f0-e0e50441a271/content

Calenge, C. 2006. The package “adehabitat” for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling 197:516–519. https://doi.org/10.1016/j.ecolmodel.2006.03.017

Chinoitezvi, H., S. Kusangaya, C. P. Muzamba, M. Ndlovu, and C. Hungwe. 2024. Modelling poaching risk zones in Sengwa Wildlife Research Area: a progressive step towards poaching management. Open Access Library Journal 11:e11509. https://doi.org/10.4236/oalib.1111509

Cushman, S. A., N. B. Elliot, D. Bauer, K. Kesch, L. Bahaa-el-din, H. Bothwell, M. Flyman, G. Mtare, D. W. Macdonald, and A. J. Loveridge. 2018. Prioritizing core areas, corridors and conflict hotspots for lion conservation in southern Africa. PLoS One 13:e0196213. https://doi.org/10.1371/journal.pone.0196213

de Boer, W. F., M. J. Vis, H. J. de Knegt, C. Rowles, E. M. Kohi, F. van Langevelde, M. Peel, Y. Pretorius, A. K. Skidmore, R. Slotow, S. E. van Wieren, and H. H. T. Prins. 2010. Spatial distribution of lion kills determined by the water dependency of prey species. Journal of Mammalogy 91:1280–1286. https://doi.org/10.1644/09-MAMM-A-392.1

Dunham, K. M., C. S. Mackie, G. Nyaguse and C. Zhuwau. 2015. Aerial survey of elephants and other large herbivores in the Sebungwe, Zimbabwe. Great Elephant Census, Seattle, Washington, USA and Zimbabwe Parks and Wildlife Management Authority, Harare, Zimbabwe. https://static1.squarespace.com/static/6450cf25e943031b610c4022/t/64afafce05260857ad5972d3/1689235426490/Oct+2015+Great%2BElephant%2BCensus%2BSebungwe%2BSurvey%2BReport%2B2015.pdf

Elliot, N. B., S. A. Cushman, A. J. Loveridge, G. Mtare, and D. W. Macdonald. 2014. Movements vary according to dispersal stage, group size, and rainfall: the case of the African lion. Ecology 95:2860–2869. https://doi.org/10.1890/13-1793.1

Funston, P. J., P. Henschel, L. Petracca, S. Maclennan, C. Whitesell, E. Fabiano, and I. Castro. 2017. The distribution and status of lions and other large carnivores in Luengue Luiana and Mavinga National Parks, Angola. KAZA TFCA Secretariat, Kasane, Botswana.

Funston, P. J., M. G. L. Mills, and H. C. Biggs. 2001. Factors affecting the hunting success of male and female lions in the Kruger National Park. Journal of Zoology 253:419–431. https://doi.org/10.1017/S0952836901000395

Grinnell, J., and K. McComb. 1996. Maternal grouping as a defense against infanticide by males: evidence from field playback experiments on African lions. Behavioral Ecology 7(1):55–59. https://doi.org/10.1093/beheco/7.1.55

Hayward, M. W., and G. I. H. Kerley. 2005. Prey preferences of the lion (Panthera leo). Journal of Zoology 267:309–322. https://doi.org/10.1017/S0952836905007508

Kittle, A. M., J. K. Bukombe, A. R. E. Sinclair, S. A. R. Mduma, and J. M. Fryxell. 2016. Landscape-level movement patterns by lions in western Serengeti: comparing the influence of inter-specific competitors, habitat attributes and prey availability. Movement Ecology 4:17. https://doi.org/10.1186/s40462-016-0082-9

Laizer, H., T. M. C. Tarimo, and B. Kisui. 2014. Home range and movement patterns of lions (Panthera leo) in Tarangire National Park, Tanzania. American Journal of Research Communication 21:44–58. https://www.usa-journals.com/wp-content/uploads/2014/01/Laizer_Vol21.pdf

Lehmann, M. B., P. J. Funston, C. R. Owen, and R. Slotow. 2008. Home range utilisation and territorial behaviour of lions (Panthera leo) on Karongwe Game Reserve, South Africa. PLoS One 3:e3998. https://doi.org/10.1371/journal.pone.0003998

Loveridge, A. J., L. L. Sousa, J. Seymour-Smith, J. Hunt, P. Coals, H. O’Donnell, P. A. Lindsey, R. Mandisodza-Chikerema, and D. Macdonald. 2020. Evaluating the spatial intensity and demographic impacts of wire snare bush meat poaching on large carnivores. Biological Conservation 244:108504. https://doi.org/10.1016/j.biocon.2020.108504

Loveridge, A. J., M. Valeix, Z. Davidson, F. Murindagomo, H. Fritz, and D. W. Macdonald. 2009. Changes in home range size of African lions in relation to pride size and prey biomass in a semi-arid savanna. Ecography 32:953–962. https://doi.org/10.1111/j.1600-0587.2009.05745.x

Mancinelli, S., L. Boitani, and P. Ciucci. 2018. Determinants of home range size and space use patterns in a protected wolf (Canis lupus) population in the central Apennines, Italy. Canadian Journal of Zoology 96:828–836. https://doi.org/10.1139/cjz-2017-0210

Mbizah, M. M., M. Valeix, D. W. Macdonald, and A. J. Loveridge. 2019. Applying the resource dispersion hypothesis to a fission-fusion society: a case study of the African lion (Panthera leo). Ecology and Evolution 9:9111–9119. https://doi.org/10.1002/ece3.5456

Mills, K. L., Y. Harissou, I. T. Gnoumou, Y. I. Abdel-Nasser, B. Doamba, and N. C. Harris. 2020. Comparable space use by lions between hunting concessions and national parks in West Africa. Journal of Applied Ecology 57:975–984. https://doi.org/10.1111/1365-2664.13601

Moeller, M. L. 2014. Factors affecting lion (Panthera leo) spatial occurrence in the Zambezi region, Namibia. Thesis. Stellenbosch University, Stellenbosch, South Africa. https://core.ac.uk/download/pdf/37436829.pdf

Monks, N., D. A. Youldon, and B. Mandinyenya. 2015. Call up survey of Chizarira National Park and Chirisa/Sengwa Safari Area, Zimbabwe. Zimbabwe Parks and Wildlife Management Authority, Harare, Zimbabwe.

Muposhi, V. K., E. Gandiwa, A. Chemura, P. Bartels, S. M. Makuza, and T. H. Madiri. 2016. Habitat heterogeneity variably influences habitat selection by wild herbivores in a semi-arid tropical savanna ecosystem. PLoS One 11:e0163084. https://doi.org/10.1371/journal.pone.0163084

Ogutu, J. O., H.-P. Piepho, M. Y. Said, G. O. Ojwang, L. W. Njino, S. C. Kifugo, and P. W. Wargute. 2016. Extreme wildlife declines and concurrent increase in livestock numbers in Kenya: what are the causes? PLoS One 11:e0163249. https://doi.org/10.1371/journal.pone.0163249

Oriol-Cotterill, A., D. W. Macdonald, M. Valeix, S. Ekwanga, and L. G. Frank. 2015. Spatiotemporal patterns of lion space use in a human dominated landscape. Animal Behaviour 101:27–39. https://doi.org/10.1016/j.anbehav.2014.11.020

Packer, C., and A. E. Pusey. 1982. Cooperation and competition within coalitions of male lions: kin selection or game theory?. Nature 296(5859):740–742. https://doi.org/10.1038/296740a0

Phillips, S. J., R. P. Anderson and, R. E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190:231–259. https://doi.org/10.1016/j.ecolmodel.2005.03.026

R Core Team. 2019. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.r-project.org/

Schaller, G. B. 2009. The Serengeti lion: a study of predator prey relations. University of Chicago Press, Chicago, Illinois, USA. https://doi.org/10.7208/chicago/9780226736600.001.0001

Snyman, A., E. Raynor, C. Chizinski, L. Powell, and J. Carroll. 2018. African lion (Panthera leo) space use in the Greater Mapungubwe Transfrontier Conservation Area. African Journal of Wildlife Research 48:1–12. https://doi.org/10.3957/056.048.023001

Tumenta, P. N., M. van’t Zelfde, B. M. Croes, R. Buij, P. J. Funston, H. A. Udo de Haes, and H. H. De Iongh. 2013. Changes in lion (Panthera leo) home range size in Waza National Park, Cameroon. Mammalian Biology 78:461–469. https://doi.org/10.1016/j.mambio.2013.08.006

Valeix, M., A. J. Loveridge, Z. Davidson, H. Madzikanda, H. Fritz, and D. W. Macdonald. 2010. How key habitat features influence large terrestrial carnivore movements: waterholes and African lions in a semi-arid savanna of north-western Zimbabwe. Landscape Ecology 25(3):337–351. https://doi.org/10.1007/s10980-009-9425-x

Verschueren, S. 2017. Social structure, movements and home ranges of lions (Panthera leo) in relation to lion livestock conflicts around Nairobi National Park, Kenya. Thesis. Universiteit Antwerpen, Antwerpen, Belgium.

Visser, H., L. Müller, P. Tumenta, R. Buij, and H. De Iongh. 2009. Factors affecting lion (Panthera leo) home range, movement and diet in Waza National Park, Cameroon. Journal of Zoology 300:131–142.

Youldon, D. A., and J. Abell. 2015. Aerial census of Chizarira National Park and Chirisa Safari Area, Zimbabwe. Zimbabwe Parks and Wildlife Management Authority, Harare, Zimbabwe.

Zehnder, A., S. Henley, and R. Weibel. 2018. Home ranges of lions in the Kalahari, Botswana exhibit vast sizes and high temporal variability. Zoology 128:46–54. https://doi.org/10.1016/j.zool.2018.04.001