Page 1/32

Space use and simultaneous movement analyses oflions and spotted hyenasNancy A. Barker  ( nancybarkerzoologist@gmail.com )

University of KwaZulu-NatalFrancois G. Joubert Ludovica L. Vissat 

University of CaliforniaVincent Stowbunenko 

San José State UniversityKathleen A. Alexander 

Virginia TechRob Slotow 

University of KwaZulu-NatalWayne M. Getz 

University of California

Research Article

Keywords: Carnivores, Competition, Co-existence, Spatiotemporal behaviour, T-LoCoH, Hull association,Intersecting hulls, Overlap, Relative movement, Interference competition

Posted Date: May 3rd, 2022

DOI: https://doi.org/10.21203/rs.3.rs-1605101/v1

License: This work is licensed under a Creative Commons Attribution 4.0 International License.   Read FullLicense

Page 2/32

Abstract

BackgroundInterference competition among sympatric carnivores can potentially in�uence species viability throughpopulation reduction or extinction, and has important implications for the structure and function of largecarnivore communities. Carnivores may mitigate the risk of competition through �ne-scaled spatial or temporalseparation that subsequently affects predator interactions and community dynamics. Simultaneous telemetryrecords of sympatric carnivores allows for an in-depth analysis of individual space-use patterns andassociation metrics. Together with the application of a novel method for analyzing �ne-scaled movementinteractions, quantifying spatiotemporal separation between predator species provides insight intounderstanding the dynamic interactions among co-existing carnivores.

MethodsWe analyzed GPS satellite relocation data from 17 lions (Panthera leo) and 14 spotted hyenas (Crocutacrocuta) over four-years, across two distinct ecosystems. Data were used to investigate lion and hyena space-use patterns and �ne-scale movement of intra- and interspeci�c dyads. We assessed home range overlapamong lion and spotted hyena, and applied a novel T-LoCoH (Time-Local Convex Hull) method to study speciesrange overlaps from an interaction perspective. We also present the �rst application of our dyadic movementmethodology to characterize the relative movement of lion and spotted hyena dyads.

ResultsHeterospeci�c home-range overlap between lion prides and hyena clans was considerably greater thanconspeci�c pride-pride and clan-clan overlap. Spatial overlaps among heterospeci�c competitors occurredmainly at the edge of home ranges, and were indicative of joint, contemporaneous space-use. Close encountersbetween individual lion and hyena provided opportunities to assess and compare interactions within homeranges and, even, core areas across seasons and ecosystems. At distances < 5 km, hyenas mostly avoided lions,whereas lions either moved in the direction of hyenas, or showed random movements. Among intraspeci�cpairs, lion dyads demonstrated a high occurrence of coordinated movements, whereas hyenas mostly avoidedeach other at close distances, with some individuals tracking others (or moving in a shared direction) at largedistances.

ConclusionsDynamic interactions among sympatric lions and spotted hyenas illuminates how hyenas effectively reducetheir potential of interactions with lions by utilizing spatiotemporal partitioning strategies, and local reactiveavoidance behaviours within shared space-use areas. Insight into the drivers shaping the spatiotemporalpatterns enabling species co-existence among apex predators is essential to understanding the behaviouralchoices made by members of a guild that subsequently affects population dynamics and community structure

Page 3/32

of multi-species food webs. Knowledge of such processes can contribute towards informed managementstrategies for the effective conservation of multi-predator systems.

IntroductionSuccessful preservation of charismatic species in wildland areas under global climate change andanthropogenic landscape transformation is a challenge that requires an increasingly deep understanding ofecological processes at all spatiotemporal scales. This includes intra- and interspeci�c interactions amongindividuals as they move across landscapes in search of resources. Interactions among apex mammalianpredators, for example, are known to have profound implications in shaping community structure andpopulation dynamics [1–3], thus potentially exacerbating the ecological impacts of discrete carnivore species[4, 5]. The interactions themselves impact �tness through various mechanisms and ecological pathways. Theseinteractions can result in interference competition that has features of harassment [6], direct killing or intraguildpredation [3], shifts in habitat use [7], exclusion of one or the other species from certain habitats or regions [8],or by adjusting their use of space and time in response to the perceived presence of competitors [9–13].

Shifts in space-use patterns over time are possible to observe if data at the appropriate spatiotemporalresolution are available [14]. At much �ner scales behavioural modi�cations, such as increased vigilance ordecreased rest, that are employed to decrease the risk of competition and injury [15, 16], can be observed fromhigh-frequency relocation data [17–22]. Analyses are needed, however, to extract from such high-frequency datathe behavioral responses of individuals on the move to encounters with other individuals. This requires thatrelocation data be available for all the individuals involved in such encounters around the time of suchencounters.

Recently, there has been an increasing focus on spatiotemporal partitioning among sympatric predators [12,23–31], yet few have quanti�ed the effects of such species interactions on the behavioural responses (i.e.,movement characteristics) of sympatrically competing predator species [32]. Thus, much more is known aboutthe space-use decisions of either predator, whereas there is very little known about the behavioural ecology oftheir joint space use.

The African lion (Panthera leo) and the spotted hyena (Crocuta crocuta), two of Africa’s largest predators, co-exist within protected areas exhibiting a high degree of overlap in their habitat use [33], temporal activitybudgets [10], and diet [34]. Despite highly coordinated movements among lions within the same pride, thespatiotemporal patterns of lion ranges from different prides reveals small to minimal overlaps among femalelion territories, with negligible to zero overlap among male lion territories [35–38]. This indicates avoidancebehaviour among different prides.

Spotted hyena clan territories are largely exclusive, and clans avoid each other temporally in areas of interclanoverlap [39, 40]. Territoriality in hyenas results in very little overlap between neighbouring clans [41, 42] becauseof interclan competition [43]. Thus, it appears that intraspeci�c competition within lions alone, and withinspotted hyenas alone, results in a strong spatial separation of any two prides or any two clans, but not asstrong between some prides and some clans.

Page 4/32

The question then arises regarding the extent to which individual lion or hyenas are spatiotemporally separatedeven though the spatial extent of their respective pride and clan show considerable overlap. It is this gap in ourknowledge of the spatiotemporal dynamics of lion and spotted hyena interactions that motivates the analysespresented here.

Using relocation telemetry data that we obtained on the movement of lions and spotted hyenas in EtoshaNational Park, Namibia, and wildland areas in northern Botswana, including the Chobe National Park, we carriedout the following: 1.) an analysis, using well-established methods for evaluating home-range overlap [44–46], ofthe degree to which individual lions from different prides overlap in their space use during the courses of aseason, compared with the amount of overlap in space use of hyenas and lions occupying the same generalareas; 2.) an analysis, using a novel application of the T-LoCoH methodology [47], of the degree to whichspatiotemporal separation occurs between individual lion and hyenas whose home ranges exhibit strongoverlap; and 3.) the �rst application of the relative motion methods developed by Luisa Vissat et al [48] toanalysing the spatiotemporal behaviour of dyads in the context of interspeci�c competition.

The approach we lay out here, although limited by the number of individuals involved in our study, is a �rst stepto assessing how interspeci�c interactions among two species of predators ultimately affects the viability ofthe populations involved, with implications for assessing risks of extinction under global change. Knowing howsuch interactions change with respect to the abundance of prey in our study areas is also critical to evaluatingthe impacts of management strategies that are routinely implemented in National Parks throughout Africa.

Of particular relevance to wildlife management in Etosha and parts of northern Botswana is the practice ofcarcass burning during outbreaks of anthrax in large mammalian herbivore species (e.g., zebra Equus quaggaand elephant Loxodonta africana in Etosha, buffalo Syncerus caffer in northern Botswana). Since suchcarcasses provide a resource bonanza for lions and hyenas, as well as local scavenger guilds (e.g., jackalsCanis mesomelas, vultures Gypinae, corvids Corvus spp. and other meso facultative scavengers) [32, 49, 50],controlling anthrax or other diseases in mammalian herbivore populations [51, 52] may actually provedetrimental to the long-term resilience of either hyenas or lions, and the scavenger guilds involved.

Methods

Study AreasThis research was conducted within two areas of southern Africa (Fig. 1): the Etosha National Park, a fencedprotected area in northern Namibia; and the Chobe National Park, an unfenced protected area in northeasternBotswana, with its adjacent Linyanti Conservancy, a community managed wildlife area within the ChobeEnclave. The study area represented a subset of each of the parks, encompassing a region of 17,500km2 andfocused on the lion and spotted hyena populations in the Okaukeujo and Halali regions of Etosha, and alongthe riparian zone of the Chobe and Linyanti rivers, hereafter Chobe. See Supplementary information (AdditionalFile 1) for additional details.

Data CollectionGlobal positioning system (GPS) satellite telemetry collars with dual-axis accelerometers (IridiumTrackM, LotekWireless Inc., Newmarket, Ontario, Canada) were �tted to 17 adult lions (13 females and 4 males) and 14 adult

Page 5/32

spotted hyenas (10 females and 4 males), and set to record �xes every 30 min from 18h00–6h00 or 17h00–8h00, and every 5 min for two hours after sunset and before sunrise, with two �xes during the day at 10h00 and14h00 respectively [10]. See Supplementary information (Additional File 1) for details, and also seeSupplementary Table 1 (Additional File 2) for a summary of the data collected for all collared individuals, andthe number of tracking days over the tracking period for each individual. Collared animals were located usingsatellite uploads of GPS locations, and, when necessary, radio-tracked from a vehicle using a handheld 3-element Yagi antenna (Lotek), and a custom built vehicular-mounted 5-element antenna, with a SRX 600telemetry receiver. Satellite data were monitored on a daily basis to ensure individuals were alive and active, andindividuals with locations accessible by the road network prior to sundown were selected for continuous followsduring the night for the ground-truthing of accelerometer data. Animals chosen for collaring were identi�ed tobelong to separate groups (see Supplementary information, Additional File 1 for details), with females favoredas they represented natal breeding group members, and because males have a higher likelihood than femalesof dispersal. Thus, all known groups within the study area included a collared individual, and the ranges ofthese selected species groups overlapped extensively with the study area. All statistical analyses wereconducted in R version 3.5.1 (R Core Team, 2018), and all geographic information system (GIS) applicationswere conducted in ArcGIS (ESRI ArcMap v.10.0, Redlands, CA, USA).

Spatially and temporally overlapping hullsWe constructed seasonal utilization distributions (UDs) for each individual using a-LoCoH (adaptive LocalConvex Hull) method in the R package, “T-LoCoH” [44]. The T-LoCoH method itself facilitates the identi�cationof regularly revisited sites [46]. We used the 95% and 50% UDs to respectively represent the home ranges andcore use areas of individuals. We used the intersection function from the “rgeos” R package to compute theareas of overlap between neighbouring individuals’ ranges. We calculated the proportions of overlaps betweenlion-lion, hyena-hyena and lion-hyena individuals and used the Welch Two-Sample t-test (“stats” R package) tocompare the proportion of overlaps in the home ranges and core use areas between competitors acrossecosystems, seasons, and diel cycles. We used t-tests to compare the home range and core area overlaps ofeach lion-hyena pair from both seasons, and then from each season in each ecosystem. We then compared thehome range and core area overlaps of each lion-hyena pair to each lion-lion pair, and to each hyena-hyena pairacross seasons for both ecosystems. Finally, we compared the home range and core area overlaps of each lion-lion pair, and each hyena-hyena pair across ecosystems.

For each of the overlapping ranges between a lion and spotted hyena dyad (i.e., pair of individuals), wecomputed the hull metrics for spatially, and then temporally, overlapping hulls for each dyad. Spatiallyoverlapping hulls (different from the calculated overlap proportions above) are those that intersect irrespectiveof time, and the number of intersecting hulls provides a coarse measure of shared used space without regard totime [47]. Temporally overlapping hulls require that the overlap of hulls occur contemporaneously, or nearly so.Since hulls are composed of the hull parent point and nearest neighbours identi�ed by the time-scaled distance(TSD), and are unique to each individual, a measure of the temporal lag time between joint space use can bederived from the minimum time difference of the hull parent points for each individual [47]. Hulls that have notemporally overlapping counterpart are denoted by “NA” [47]. Thus, it is possible for spatially overlapping hulls(the number of intersecting hulls) and temporally overlapping hulls (minimum time difference between hullparent points) to differ in that the hulls overlap spatially but not temporally, and to differ between individualswithin a dyad.

Page 6/32

As we were interested in lion-hyena interactions on a �ne scale, we focused our hull metric analyses during theperiods of dusk and dawn, when our relocation data frequency was 5 min, and when lions and spotted hyenaswere most active. Using the “T-locoh.dev” R package, we extracted the frequency of intersecting hulls, and theminimum time lag between spatially overlapping hulls, for each dyad. We assessed the frequency of spatialoverlaps within dyads on the map, and determined whether high intensity spatial overlaps occurred within theouter edge of either individual’s home range, or within the inner area or the center of their home ranges. We �rstdesignated the 95% a-LoCoH isopleth as the de�ning line between the outer edge (remaining 5%) and the innerarea (95%) of the home range, using 5% increments until the 50% isopleth (with equal parts for the outer edgeand the center: i.e., 50% each). For each increment of the home range edge/inner area region, we counted thepercent frequency of high intensity spatial overlaps that occurred within the outer edge and the inner area foreach individual in a dyad. We also repeated this procedure using the home range edge (outside 95% isopleth) tocompare against the center of home ranges (i.e., inside 50% isopleth core use area). We then used the t-test oneach species group and from each ecosystem to determine whether increased spatial overlaps between dyadsoccurred more often at the outer edge of ranges (outside the 95% isopleth) or inside the inner area (inside the95% isopleth) or the center (within the 50% isopleth) of home ranges.

For temporally overlapping hulls, we used a maximum time interval of 2.5 minutes to generate hulls from parentpoints and their nearest neighbours, from which we could compute the mean centroid distances of time-matched hulls (Supplementary Appendix 1, Additional File 3). We also computed the centroid distances for anequal number of randomly paired hulls, and tested for equal means between time-matched and randomly pairedcentroid distances. We again used a t-test to evaluate whether the distribution of time-matched centroiddistances were signi�cantly different from a null model of no interaction, using an equal number of randomlypaired hulls. In addition, we used a two-sample Kolmogorov-Smirnov test (“stats” R package) to evaluatewhether time-matched and randomly paired centroid distances came from the same distribution.

Movement responses of individualsFor our movement analyses, we used high frequency (5 min sampling interval) data obtained during an earlymorning (4h00-6h00) and late evening (19h00-21h00) sampling periods from 13 lions (7 females and 3 malesfrom Namibia, with 3 females from Botswana) and 8 spotted hyenas (6 females and 1 male from Namibia, with1 female from Botswana). We analyzed the movement behaviour of pairs of individuals, using approach andretreat behaviours as a means of identifying the type of interaction occurring between individuals. This methodis described in detail elsewhere [48]. Brie�y, dyads were identi�ed from the 5 min sampling intervals as lion-lion,hyena-hyena, and lion-hyena pairs of individuals coming within a distance of 5 km of each other. For each ofthese pairs, we �rst computed the absolute heading of each individual, and then the relative heading of eachindividual with respect to the location of the other individual in the dyad. More speci�cally, the absolute headingof individuals at time t was taken to be the angle between the vector pointing North and the vector joining theposition of the individual at time t and at time t + 1 (angles are shown in blue in Fig. 2a). The relative heading ofeach individual at time t was computed as shown in green in Fig. 2a. In addition, for each individual in a dyad,we evaluated the difference between the absolute and relative headings at time t for each individual, as shownin red in Fig. 2a. We used the bearing function from the “geosphere” R package to calculate these angles.

For purposes of categorization, we divided a circle into eight sections and de�ned individual A as “movingtowards” individual B if the absolute value of the difference were either < 67.5 or > 292.5 degrees. However, if the

Page 7/32

absolute value was between 112.5 and 247.5 degrees, we de�ned individual A as “moving away from”individual B (Fig. 2b). Finally, we labelled any value falling between these two sections as “orthogonal” andlabelled the movement of A with respect to B as “orthogonal”. Individuals were coded as moving towards, away,or orthogonal relative to each other.

From these dusk and dawn periods, we counted the number of movements towards (blue) and away (red) fromthe other individual. We designated nine distance intervals of 0 < 50 m, 50 < 100 m, 100 < 200 m, 200 < 500 m,500 m < 1 km, 1 < 2 km, 2 < 3 km, 3 < 5 km, and 5 < 10 km, and grouped the counts for each individual when at aparticular distance in a given interval. We computed the 95% con�dence interval of approach (towards) andretreat (away) movements from the other individual in the dyad, for each of the distance intervals available forthat dyad. Using the “binom” R package [53], we calculated the con�dence interval for each distance bin, takingthe proportion of movement towards the other individual over the total number of towards/away movementscalculated for that dyad. We expected this value to be equal to 0.5 for random movement. If the calculated 95%con�dence interval did not include this value, we conclude that there was a statistically signi�cant differencebetween individual movements from random movements. We then calculated the ratio between the towardsmovements and the sum of towards and away movements. Values close to 0 implied movement mostly awayfrom the other individual, while values close to 1 implied movement mostly towards the other individual. Valuesclose to 0.5 represented a similar number of movements both towards and away from the other individual,which we de�ned as random movements. We also checked that this type of analysis was not sensitive to theinclusion of orthogonal movements. In one analysis we considered only the towards and away segments. Wethen checked if the results changed when using all the segments including the segments of orthogonalmovements, and the results did not change signi�cantly.

Finally, we plotted the distribution of towards, away, and orthogonal movements for each individual in the pair,and grouped them accordingly to the distances among the individuals in the pair. For certain pairs, whereindividuals spent some time at close distances to each other (i.e. < 15 m and < 100 m), and we had su�cientdata, we assessed, using the same statistical analysis, whether individuals exhibited different behaviours atthese two close distances (effectively adjacent (0–15 m), or nearby (15–100 m)).

Movement responses of dyadsFor given dyads (A,B), we repeated the analysis to evaluate simultaneous movement behaviours in which twoindividuals moved at the same time to come closer (approach) or further apart (retreat), or one approaches asthe other moves away. For each time point during that period of 5 min sampling intervals, we assigned one of�ve different types of simultaneous movements (Fig. 2b). We repeated the statistical analysis as above,excluding the orthogonal movements, and we evaluated whether or not the proportion of the four differentmovement behaviour types was signi�cantly different from random (Fig. 2b). Thus, when there is randommovement, we expected this proportion to be equal to 0.25 for each movement type. Again, we plotted thedistribution of the different movement types, and grouped them according to the distances between theindividuals in the pair. Finally, we presented the simultaneous behaviour over time for the dyad at distancesbelow 10 km, with each point coloured to the corresponding simultaneous movement type. We indicate caseswhere individual movements or the pair’s movement types differed signi�cantly from random movements withan asterisk. For each dyad, we illustrated, with the ratio and distribution plots, which individual wasdemonstrating movements towards or away from the other, for a given distance interval. In addition, distance

Page 8/32

over time plots revealed incidences of simultaneous approach/retreat behaviours exhibited by the pairthroughout the study period.

ResultsData Collection

We retrieved 57% of all possible relocations while collars were deployed, due to the loss of collars fromindividuals that were killed, or for which we were unable to retrieve the collar (n = 3 lions and 6 hyenas).  In theend, our data (n = 349,163) from the two study areas comprised of 55% lion and 45% spotted hyena relocations(Supplementary Table 1, Additional File 2).  Collared lions were sometimes observed together in both Etosha(51% of �xes within ≤50 m to the other) and Chobe (6% of �xes within ≤50 m to the other), while only femalecollared hyenas occurred together in Chobe (1% and 2% of �xes within ≤50 m to the other, respectively).  Of thecollared lions observed together within ≤50 m in Etosha, 65% were of female-female pairs, and 35% were ofmale-female pairs.  However, in Chobe nearly all of the pairs were male-female, with only 1% female-female.  

Overlaps among home ranges and core use areas

We found a high degree of spatial overlap between lions and spotted hyenas in their home ranges and coreareas across the two ecosystems.  The total area of overlap in the home ranges and core areas of the twopredators, both the other’s main competitor, are also presented for each pair in Supplementary Table 2(Additional File 4).  We calculated the proportion of lion home ranges and their core areas overlapped by spottedhyena home ranges and core areas, as well as the proportion of spotted hyena home ranges and their coreareas overlapped by lion ranges and core areas (Supplementary Table 3, Additional File 5).  We furthercalculated the proportion of overlap between lion conspeci�cs and between spotted hyena conspeci�cs(Supplementary Table 4, Additional File 6).  Lions and hyenas shared more of their home ranges than they didtheir core areas, with core area overlaps being at most a quarter of the entire overlap in their home ranges. Lions in Etosha shared up to (mean ± SE) 20 ± 4% of their home ranges and 8 ± 2% of their core use areas withhyenas.  Meanwhile, hyenas shared up to 17 ± 3% of their home range and 8 ± 2% of their core use areas withlions in Etosha.  By contrast, lion and hyena conspeci�cs shared less of their home ranges and core areas witheach other in Etosha (mean ± SE: lion home ranges 10 ± 2%, core areas 6 ± 2%; hyena home ranges 13 ± 4%,core areas 5 ± 3%).  Similarly in Chobe, lions shared up to (mean ± SE) 23 ± 5% of their home ranges and 3 ± 1%of their core use areas with spotted hyenas, while hyenas shared up to 23 ± 4% of their home ranges, and lessthan 1 ± 0.5% of their core use areas with lions.  However, lion and hyena conspeci�cs in Chobe shared more oftheir home ranges and core areas with each other (mean ± SE: lion home ranges 26 ± 7%, core areas 13 ± 5%;hyena home ranges 36 ± 16%, core areas 30 ± 15%).

The proportion of overlap in the core use areas between lions and hyenas were larger in Etosha than they werein Chobe (mean ± SE: Etosha 8 ± 2%, Chobe 3 ± 1%, t = 2.19, df = 81.0, p < 0.05; Supplementary Table 3,Additional File 5).  In addition, the proportion of overlap in the core use areas between lions and hyenas inEtosha were larger during the wet season than they were for the dry season (mean ± SE: wet season 14 ± 4%,dry season 3 ± 2%, t = -2.31, df = 34.7, p < 0.05; Supplementary Table 3, Additional File 5).  Although lions didnot differ in the sizes of overlapped areas between conspeci�cs or competitors, a larger proportion of Etoshalion home ranges overlapped with hyenas than they did with lions (mean ± SE: heterospeci�cs 20 ± 4%,

Page 9/32

conspeci�cs 10 ± 2%, t = -2.35, df = 103.7, p < 0.05; Supplementary Table 3, Additional File 5).  By contrast, lionsin Chobe shared signi�cantly more of their home ranges with other lions than they did with other lions in Etosha(mean ± SE: Etosha 10 ± 2%, Chobe 26 ± 7%, t = -2.19, df = 28.9, p < 0.05; Supplementary Table 4, Additional File6).   

Spatially and temporally overlapping hulls

The frequencies of overlapping hulls within an individual’s home range indicate the areas of spatial overlaps orshared space-use with a competitor (Figs. 3a-b).  Analyses of dyads’ hull association metrics are presented inSupplementary Appendix 2 (Additional File 7).  Areas of high spatial overlaps typically occurred at the outeredge of individual ranges (Table 1), although this varied between species and across ecosystems.  Spottedhyenas had more high intensity spatial overlaps in the outer 25% region of their home ranges (75% isopleth),while lions had more high intensity spatial overlaps in the outer 35% region of their home ranges (65% isopleth)(Table 1a).  However, when comparing to core use areas, both lions and spotted hyenas had more high intensityspatial overlaps in the outer 25% and 10% regions (75% and 90% isopleths, respectively) of their home ranges(Table 1b).  

With the minimum time lag separated into different periods of 0, >0-15 mins, >15 mins-1 hr, >1-2, >2-12, >12-24,and >24 hr, the majority of the minimum time between overlapping hulls for nearly all dyads occurred at >12-24and >24 hr (Figs. 3c-d).  Hulls with a minimum time lag of 0 min occurred for several lion-hyena dyads fromeach of the dry and wet seasons in Etosha and Chobe, and are indicative of joint space-use at the same times(see Supplementary Appendix 3, Additional File 8 for more details).  

Interactions between lion and spotted hyena dyads were identi�ed from the temporally overlapping hulls withinthese joint space-use areas (Figs. 3e-f).  There was signi�cant interaction (Welch Two Sample t-test p < 0.05) inthe temporally overlapping hulls among the home ranges (95% range) for most lion-hyena dyads in Etosha, andfor at least half of the lion-hyena dyads in Chobe, for each season (Table 2, and see Supplementary Appendix 4,Additional File 9).  Similarly, there was signi�cant interaction (p < 0.05) in the temporally overlapping hullswithin core areas (50% core use area) for the majority of the lion-hyena dyads in each season for both Etosha,and for at least half of the lion-hyena dyads in Chobe (Table 2).  Overall and during the wet season, Etoshadyads had higher occurrences of interactions between lions and spotted hyenas, both in home ranges and coreareas, than did Chobe dyads (Chi-square tests p < 0.05; Table 2).  

Additionally, there were signi�cant differences in the distributions of time-matched versus randomly pairedcentroid distances for all dyads (two-sample Kolmogorov-Smirnov test p < 0.05).  Histograms of the signi�canttemporal overlapping dyads (n = 74) revealed similar occurrences of right-skewed (44.6%) and normal (43.2%)distributions, with few left-skewed (12.2%) distributions.  Although normally distributed dyads indicate neitheran attraction or repulsion effect, right-skewed dyads (e.g., NU-33865 & GO-33869; SU-33868 & SA-33872 inSupplementary Appendix 4, Additional File 9) indicate that these individuals remained closer together duringtime-matched observations, whereas left-skewed dyads (e.g., OJ-33867 & SA-33872; KB-36717 & KW-33871 inSupplementary Appendix 4, Additional File 9) were further apart than random pairing.

Table 1.  Intensity of spatially overlapping hulls between individuals of interspeci�c dyads (see SupplementaryAppendix 2, Additional File 7 for all dyads) from the Etosha National Park, Namibia (ENP), and the Chobe

Page 10/32

National Park and Linyanti Conservancy, Botswana (CNP).  Spatially overlapping hulls occurred either (a) withinthe outer edge (outside of 95% isopleth), or inner area (inside of 95% isopleth) of the individual’s home range;and (b) within the outer edge (outside of 95% isopleth), or within the center (i.e., the core use area within the 50%isopleth) of the individual’s home range.  The direction of the Greater-than and Less-than signs indicate theregion with the greater extent of spatially overlapping hulls, and the t-test signi�cance at p < 0.05 is denoted bythe emboldened signs.   

(a)

  Percent area of home range (Outer edge/Inner area)

  5/95 10/90 15/85 20/80 25/75 30/70 35/65 40/60 45/55 50/50

Lion <  <  <  <  <  >  >  >  >  > 

Spottedhyena

<  <  <  >  >  >  >  >  >  > 

ENP Lion <  <  <  <  <  <  <  >  >  > 

Hyena <  <  <  >  >  >  >  >  >  > 

CNP Lion <  <  <  <  >  >  >  >  >  > 

Hyena <  >  >  >  >  >  >  >  >  > 

(b)

  Percent area of home range (Outer edge/Center)

  5/50 10/50 15/50 20/50 25/50 30/50 35/50 40/50 45/50 50/50

Lion <  >  >  >  >  >  >  >  >  > 

Spottedhyena

<  >  >  >  >  >  >  >  >  > 

ENP Lion <  <  <  <  >  >  >  >  >  > 

Hyena <  >  >  >  >  >  >  >  >  > 

CNP Lion >  >  >  >  >  >  >  >  >  > 

Hyena >  >  >  >  >  >  >  >  >  > 

  Table 2.  Percent frequencies of signi�cant interactions observed in temporally overlapping hulls among homeranges (95%) and core use areas (50%) of lion and spotted hyena dyads (n = 50) for each season from theEtosha National Park, Namibia and the Chobe National Park and Linyanti Conservancy, Botswana (seeSupplementary Appendix 4, Additional File 9 for all dyads).  (b) Chi-square results on the frequencies ofsigni�cant interactions from temporally overlapping hulls in (a).  

Page 11/32

(a)

  Region

Overall (%) Dry season (%) Wet season (%)

Home ranges Etosha (n = 8 dry, 22 wet) 87 63 96

Chobe (n = 8 dry, 12 wet) 60 50 67

Core use areas Etosha (n = 8 dry, 22 wet) 80 75 82

Chobe (n = 8 dry, 12 wet) 60 50 67

 (b)

   Region

Season Frequency Signi�cance

Home ranges ENP overall 26 < 0.001

dry 5 0.48

wet 21 < 0.001

CNP overall 12 0.37

dry 4 1

wet 8 0.25

Core use areas ENP overall 24 < 0.005

dry 6 0.16

wet 18 < 0.005

CNP overall 12 0.37

dry 4 1

wet 8 0.25

Movement responses of individuals and pairs from interspeci�c dyads

Among interspeci�c dyads (Supplementary Table 5.1, Additional File 10), spotted hyenas exhibited mostlyretreat movements in response to lions, while lions had more approach movements towards hyenas (Table 3a). During simultaneous pair movements, lions mostly tracked hyenas, and hyenas tracked lions more often thanthey either approached or moved away from each other simultaneously (Table 3b).  Although the movementdistribution ratios indicated variation among whether individuals in the dyad utilized approach or retreatbehaviours at various distance intervals (Fig. 4; and see Supplementary Appendix 5, Additional File 11 foradditional pairs), hyenas exhibited signi�cantly more retreat than they did approach or random movementstowards the lion counterpart (Table 3c).  By contrast, lions did not signi�cantly differ in their movementstowards their hyena counterpart.  At larger distance intervals of up to 5 and 10 km, more than half of the lion-

Page 12/32

hyena dyads presented a ratio close to 0.5 (67% and 70%, respectively), indicating mostly random movements. Conversely at distances of up to 5 km, 22% of the hyenas moved mostly away from lions. 

Table 3. Chi-square statistical results for differences among movement types for each focal species (lion orspotted hyena) within a dyad (lion-hyena, lion-lion, or hyena-hyena), as obtained from the (a) individual or (b)simultaneous movements (Supplementary Table 5.1-4, Additional File 10), and (c) movement distribution ratios(Supplementary Appendix 5-6, Additional Files 11-12). 

(a)

Dyad Focal Species Movement types Percentage of cases Signi�cance

Lion-Hyena Lion towards 38% x2 = 77.44, df = 3, p < 0.001

away 8%

Hyena towards 0

away 54%

Lion-Lion Lion towards 74% x2 = 23.04, df = 1, p < 0.001

away 26%

Hyena-Hyena Hyena towards 38% x2 = 5.76, df = 1, p < 0.05

away 62%

(b)

Dyad Movement types Percentage of cases Signi�cance

Lion-Hyena Lion → ←Hyena 12% x2 = 53.20, df = 3, p < 0.001

← Lion Hyena → 6%

Lion → Hyena → 53%

Hyena → Lion → 29%

Lion-Lion Lion → ← Lion 7% x2 = 47.54, df = 2, p < 0.001

← Lion Lion → 30%

Lion → Lion → 63%

Hyena-Hyena Hyena → ←Hyena 22% x2 = 23.12, df = 2, p < 0.001

← Hyena Hyena → 22%

Hyena → Hyena → 56%

 (c)

Page 13/32

 

 Dyad

Focal Species Percentage of cases

Movement types

Signi�cance

approach retreat random

Lion-Hyena Lion 35% 25% 40% x2 = 3.50, df = 2, p > 0.05

Hyena 14% 57% 29% x2 = 28.58, df = 2, p < 0.001

Lion-Lion Lion 34% 32% 34% x2 = 0.08, df = 2, p > 0.05

Hyena-Hyena Hyena 32% 32% 36% x2 = 0.32, df = 2, p > 0.05

Movement responses of individuals and pairs from intraspeci�c dyads

Individual movements from intraspeci�c dyads differed between species (Supplementary Table 5.2-3, AdditionalFile 10), in that lions exhibited mainly approach movements towards conspeci�cs, with spotted hyenasdemonstrating mostly retreat movements (Table 3a).  However, both lions and spotted hyenas tracked theirconspeci�c counterpart (i.e., following synchronously as the other individual moves) more often than theyapproached or moved away from each other simultaneously (Table 3b).  At distances of 3-5 km, lions eithertracked or moved away from their counterpart lion, whereas at larger distances of 5-10 km, Etosha lionscontinued to track their counterpart, with Chobe lions exhibiting mostly avoidance movements.  Conversely,spotted hyenas demonstrated mostly avoidance movements towards their counterpart hyena at distances of 1-3 km, despite tracking them at distances of 2-10 km.  

Furthermore, at close distances of 0-0.5 km, Etosha lions display highly coordinated movements in which theyeither moved in the same direction or tracked one another for extended periods.  Since three lion dyads wereobserved to move together for a signi�cant amount of time, we repeated the analysis while looking at theheading difference for both individuals at distances below 100 m and distances below 15 m (SupplementaryTable 5.4, Additional File 10).  This analysis indicated a ratio above 0.5 at distances below 100 m, suggestingthat these dyads tended to move in the same direction when close to each other.  In dyads that consisted ofmating pairs, we found that the females tended to move more towards the males when at closer distances (upto 15 m) when she would present herself to be receptive for mating.  At further distances (between 15 m and100 m), the female moved mostly away from the male, whereas the males moved mostly towards the females,presumably in anticipation of the next mating opportunity. 

The movement distribution ratios revealed no occurrences of hyena dyads occurring at distances below 2 km,while lion dyads were observed to occur at distances as low as 15 m (Supplementary Appendix 6, AdditionalFile 12).  Both lion and hyena movements were nearly equally distributed among approach, retreat, and randommovements towards their conspeci�c counterpart in the dyad (Table 3c).  However, most of the lion and spottedhyena individuals exhibited random movements (89% and 63%, respectively) at larger distance intervals up to10 km, although retreat movements were demonstrated in 25% of the hyena individuals.  In addition, hyenasapproached conspeci�cs more (33%) than they retreated (17%), with half of their movements random (50%; x2 =16.34, df = 2, p < 0.001) at shorter distances of up to 5 km.   Contrarily, lions did not differ between random andretreat movements towards conspeci�cs at distances up to 5 km. 

Page 14/32

DiscussionEcological niches are partitioned both in space and time [54, 55], and sympatric guild members have beendocumented to utilize spatiotemporal separation to facilitate co-existence in the same space [2, 56–61]. WithinAfrica’s large carnivore guild, carnivores mitigate the risk of dangerous interactions with dominant predatorsthrough spatial or temporal avoidance [10, 13, 62–64]. Our results demonstrate that lions and spotted hyenasbehaviourally mediate the potential for interference competition by adjusting their space-use and activitypatterns, and movement behaviours in different ways that have implications for co-existence.

Our �ndings indicate that lions and spotted hyenas utilize either minimal overlaps among core areas, or a largelag time of 12–24 and > 24 hr within shared space-use, as a spatiotemporal strategy to allow the two species toco-exist within the same areas without encountering one another. However, spatially �xed resources such aspermanent water points or large carcasses that cannot be carried away by the predator (e.g. elephantsLoxodonta africana and giraffes Giraffa camelopardalis), requires that carnivores use a different mechanism toavoid directly encountering one another at these locations. Similar to Atwood et al. [65] who recorded dominantcarnivore species at nearly all of the arti�cial water points, lions and spotted hyenas were observed at all of thepermanent water points in Etosha, providing limited opportunity for spatial resource partitioning to occur. Thus,the degree of partitioning may be associated with the differences in the densities of spatially �xed resources.Edwards et al. [63] found higher levels of temporal partitioning between free-ranging carnivores, including brownhyenas (Hyaena brunnea) and leopards (Panthera pardus) on farms with lower densities of water points.

Notwithstanding the spatiotemporal partitioning strategies observed in this study, lions and spotted hyenaspersisted in tracking one another, and sometimes for extended periods. Although the type of interaction(whether remaining closer together or further apart than random) differed among pairs, the results of this studyprovides empirical evidence for the presence and occurrence of important interactions among lions and spottedhyenas. Other studies have documented competitive interactions between lions and spotted hyenas [6, 22, 66–68], including induced changes in one species’ behaviour according to the density and distribution of the other[20, 69–72]; however, data, such as we present, on direct interactions between lions and hyenas are rare. In areview of lion-hyena interactions, Périquet et al. [33] identi�ed costs and bene�ts to both species throughexploitation and interference competition, scavenging opportunities, and differences in prey selection, whileacknowledging a lack of studies using telemetry data on both species simultaneously to examinespatiotemporal avoidance. A recent study in the Serengeti which deployed tracking collars simultaneously onboth lions and spotted hyenas, found lions to be strongly positively associated with areas of high hyenautilization [35]. To our knowledge, our study is the �rst of its kind to use �ne scale GPS relocation data on thetwo species simultaneously to allow for an analysis of the direct effects of individuals from the two species onone another.

In addition to interspeci�c interactions, our results indicated a prevalence of intraspeci�c interactions amonglions. Lion prides in both Etosha and Chobe exhibited a �ssion-fusion social system in which different liongroups would frequently join together for short periods of time (i.e., a few hours or days) before disbandingagain into their respective smaller groups [73, 74]. Although Etosha lion pairs were at closer distances morefrequently than Chobe lions [75], some of our collared lions demonstrated concurrent movements in the samedirection for extended periods, which corresponds with Benhamou et al. [76] who found a high degree of jointmovement between lion dyads. Lions derive bene�ts from associating with conspeci�cs to attain access to high

Page 15/32

quality habitats and for territorial defense [77, 78], cooperative group hunting [79], and to defend kills againstkleptoparasitism from spotted hyenas [66]. Our analysis revealed coordinated movements between conspeci�cpairs, and, while both individuals of the pair tended to move towards each other, we were able to determinewhich individual of the pair was tracking the other. This may prove advantageous in studies where socialdynamics among members of the group are important to discern in determining hierarchical ranks, or incompetitive systems in which we need to determine whether certain species are dependent on other species (i.e.,tracking them) in search of resources. In addition to knowing who was tracking whom, our analysis was alsoable to illustrate the movement behaviours synonymous with mating between lion pairs, which may provideresearchers with a means of con�rming the occurrence of mating bouts among species of economic orconservation interest.

In contrast, collared hyenas in Chobe were observed to exhibit tolerant behaviours towards one another whenthey came together, more so than was seen with the collared hyenas in Etosha [75]. Hyenas from different clansin Chobe were sometimes observed to share large carcasses, at times feeding together or within minutes ofeach other. Etosha clans occupied nearly exclusive territories that �uctuated seasonally with the migratorymovements of prey species, and encounters among members of neighbouring clans were rare. These �ndingscoincide with previous studies in Etosha [80, 81], and are similar to hyena clan territories in the Serengeti, whichremain largely exclusive with frequent foraging trips outside territories to feed on migratory herbivores [82]. Onthe rare occasions during our study when two individuals of different clans encountered one another, weobserved a sudden halt in movement, increased vigilance and acute sni�ng of the air in the direction of theother. Additionally, either one or both individuals were then seen to actively change direction to avoid each other,or retreat in the direction they had just come. During night follows in Chobe, hyenas were sometimes observedto be startled by the unexpected appearance of a conspeci�c (noted by the sudden jerk backwards and therapidly departing speed of both individuals in opposite directions). It is potentially likely that hyenas exhibitingsimilar foraging strategies may unknowingly come close together in dense, riparian habitats, which wouldaccount for the retreating behaviours between hyenas that we observed in Chobe, unlike Etosha which lacksextensive dense riparian habitat.

The increased proportion of overlaps among home ranges observed in Chobe hyenas coincides with the“uninterrupted mosaic” of hyena territories from the Maasai Mara in Kenya [41], and in the Savuti region of theChobe National Park [83]. Hyenas were found to scavenge more food from lions, and had higher food intake inareas of low lion densities relative to areas of higher lion densities [69], and adjusted their grouping patterns inresponse to feeding competition [43]. In contrast to a previous study in Etosha which never saw hyenas stealfood from lions [68], the hyenas in this study were observed to successfully appropriate food from female lionsin both ecosystems. Thus, hyenas presumably derive bene�ts from conspeci�cs to both appropriate resourcesand for protection from lions.

Coexistence among species is possible when a trade-off occurs between exploitation and interference [84].Carnivore species that employ costly interference mechanisms (i.e. territoriality or resource guarding) are able tocoexist with competitors provided that they also engage in interference mechanisms that confer a bene�t tothemselves (i.e., predation or kleptoparastism) [84]. Our analysis of interspeci�c pair movements demonstratesthat a majority of hyenas exhibit retreat behaviours away from lions, while a third of lions exhibited approachbehaviours towards hyenas. This is in line with Vanak et al. [13], who demonstrated that lions consistently move

Page 16/32

towards other carnivore species that turn away from the lions, although their study lacked data on spottedhyenas. Similarly, cheetahs (Acinonyx jubatus) have been observed to avoid immediate risk by positioningthemselves at further distances from lions and hyenas than lions and hyenas do towards cheetahs [9], despitethis behaviour not being emulated by the hyenas in response to the lions in that study. Furthermore, lions andhyenas were found to actively track each other [28], and hyenas appeared to lose more of their kills to lions,although not in terms of biomass [33]. As lions are signi�cantly larger and stronger than hyenas, and thesuccess of acquiring food from hyenas is largely dependent on the ratio of lions to hyenas [66, 67], it followsthat lions would likely approach hyenas at the prospect of encountering a feeding opportunity.

In our data, we observed that the retreat behaviour of hyenas in the presence of lions persisted even at largedistances of 1–10 km. This suggests that the strength of pairwise interactions is not associated with thedistance between the paired individuals. Kittle et al. [35] suggested that the observed positive associationbetween lions and spotted hyenas in his study resulted from one species tracking the other. Our results supportthis hypothesis, with long periods of interaction occurring between the two species (upwards of 300 consecutiveminutes). Despite considerable variation among individual animals, more than half of the hyena individualsconsistently show retreat behaviours away from lions. Similarly, cheetahs were observed to exhibit localizedreactive avoidance to lions in the Serengeti [28]. However, lions rarely retreated and, instead, demonstratedmainly movements towards hyenas, as hyenas simultaneously moved away.

Alternatively, if both species interfere less with each other than they do among themselves with regard to theacquisition of resources, then the higher intensity of intraspeci�c competition relative to interspeci�ccompetition enables species coexistence [85, 86]. We also applied our analyses to intraspeci�c interactionsbetween conspeci�c pairs. Our results indicated that both species differed in the utilization of approach andretreat movements towards conspeci�cs, with lions utilizing approach movements in nearly three quarters ofcases, and hyenas retreating in more than half of the cases. However, lions were found to spend considerablymore time together than hyenas did. Lions have been documented in other studies to have �ssion-fusionsocieties in which pride members form subgroups of various sizes [73]. Larger prides were found to have acompetitive advantage in intergroup competition for territory [87], and better able to defend againstkleptoparasitism from other predators [66]. Prides with more adult females were more likely to gain access tohigher quality habitats and had increased reproductive success [78], and were able to better defend cubsagainst predation and infanticidal males [88]. Similarly, intraspeci�c interactions in a reintroduced population oflions resulted in shifts in the habitat selection among lion groups, relegating subordinate members tosuboptimal habitats [77]. Thus, for the lions in our study, it is likely that the direct effect of intraspeci�cinteractions exerted a stronger in�uence on lion movement behaviour than the indirect effect of interspeci�cinteractions.

The opposite, however, is true for spotted hyenas. Our �ndings effectively demonstrated that, despite variationamong individuals, lions actively tracked spotted hyenas across much of their shared ranges, while hyenasutilized local reactive avoidance behaviours in response to being in close proximity to lions. The activebehavioural avoidance of lions by spotted hyenas demonstrated in this study is similar to the behavioursexhibited by cheetahs and wild dogs (Lycaon pictus) from other studies which have actively avoided lions [9, 13,28, 89]. In studies that examined the behavioural response of hyenas to lion calls, hyenas were found to varytheir level of vigilance and responses to lion calls based on individual characteristics of risk-taking

Page 17/32

temperaments [70, 90]. In the Hwange National Park, spotted hyenas selected den sites and altered theirpatterns of den attendance, and shifted their foraging strategy from active predation to scavenging in responseto the potential of competitive interactions with lions [71, 72]. Therefore, and as observed in this study, spottedhyenas have the capacity to behaviourally mediate their responses to a dynamic environment, includingwhether they choose to track, or to retreat from lions.

Measures of dynamic interactions have been proposed to allow researchers to quantify the type of interactionsbetween dyads based on movement data (see [91, 92] for a complete review). The majority of these approachesfocuses on proximity which re�ects the proportion of time the two individuals spent together [76, 93–95], and isgenerally given as a range of 0 to 1 [93, 96–99], or re�ects the movement coordination within a range of -1 to 1[95, 100, 101]. Some of these metrics are either di�cult to interpret [95, 100], require a reference area [96, 98], ordepend on a parameter with its underlying assumptions [93, 95, 97, 99, 100]. Benhamou et al. [76] usessuccessive distances between lion individuals while moving to determine joint movements of a dyad. To ourknowledge, previous measures of dynamic interaction do not illustrate following/leadership behaviour, orwhether individuals exhibit approach/retreat behaviours. Here, we have used a reliable inferential framework foranalyzing the movement interactions of paired individuals, to understand the dynamics of asymmetricalmovement interactions.

The relative-motion method we used allows us to analyze approach and retreat behaviours as a means ofidentifying the type of interaction occurring between individuals. The application of this method also allows usto ascertain speci�c nuances of the animal’s behaviour in relation to the other, so that we could test hypothesesrelated to how movement interactions are dictated by species that compete for the same resources, and howthey are in�uenced by sex and social bonds within conspeci�cs. In addition, if analyses are performed onrelatively high frequency relocation data (e.g., 5 min sampling intervals) then movement interactions are able toclearly reveal approach and avoidance behaviours when competing individuals encounter each other as theymove across shared space-use areas. Animals that segregate their shared space use, or visit the same areas atdifferent times, never meet. Our �ndings indicate that this is not the case for lions and spotted hyenas, whichnot only have shared space-use areas, but also often encounter one another within these jointly used areas.Consequently, knowledge of the nuances with which individuals utilize approach or retreat behaviours isessential for understanding the dynamic interactions between pairs.

Furthermore, with respect to large-scale spatiotemporal partitioning exhibited by sympatric lions and spottedhyenas occupying the same area at different times, we conclude that �ne scale temporal partitioning of activityperiods, and reactive local avoidance employed by hyenas, acts as the main mechanisms for facilitating the co-existence of the two species. Speci�cally, spotted hyenas minimize the potential for interference competition bybehaviourally adjusting their movements to effectively avoid encountering lions. Understanding the space-usepatterns and movement characteristics of sympatric predators on a �ne-scale provides insight into the dynamicinteractions between co-existing species [76, 91, 92, 102–105]. Such dynamic interactions between individualscan reveal how behaviour choices are made by predators and scavengers of a guild, and has profoundimplications for the management strategies for both species [32, 86, 106]. For instance, conservation managersmay consider incorporating landscape features at arti�cial water sources to provide competing species withopportunities for spatial partitioning, or to supplement alternative water sources in areas with low densities ofwater points where species demonstrate high levels of temporal partitioning behaviours. In light of a changing

Page 18/32

global climate and increasing land-use pressures, factors in�uencing the spatial and temporal constraintsimposed by predator movements can inform future area and con�guration requirements of co-existing largepredators [107–111], which ultimately may prove vital towards the conservation of multi-trophic communitiesand the maintenance of ecosystem function.

ConclusionsWe have provided evidence for the presence of detailed interactions between lions and spotted hyenas withintheir home ranges where they overlap spatially and temporally. Areas of overlap among individuals typicallyoccurred at the outer edges of individual ranges. Although the majority of the minimum time lag betweenspatially overlapping hulls among lion and spotted hyena dyads was 12–24 and > 24 hr, there were severalincidences of overlapping hulls without any time lag, which indicates the occurrence of a shared space use atthe same times. Lions and spotted hyenas evidently interacted within these jointly-used areas, both acrossseasons and ecosystems. Furthermore, in the context of lion-hyena interactions, our �ndings revealed that lionsare signi�cantly more often tracking spotted hyenas, while hyenas simultaneously utilize local reactiveavoidance behaviours to reduce the probability of encountering or coming into close contact with lions. Thus,certain species mediate the potential effects of interspeci�c competition by utilizing localized habitat shifts,temporal shifts in activity periods, or behaviourally adjusting their trajectory in response to reducing the risk ofpotential interactions with a competitor. Quantifying these types of movement behaviour can be a usefulindication of the type of interaction occurring in a dyad, and may prove essential to guide conservation effortsof certain species. Understanding how apex predators move across the landscape in response to interactionswith competitors will ultimately affect the maintenance of diverse and functional ecosystems in the face ofincreasing climatic variability and dwindling wild lands.

AbbreviationsGPS: Global Positioning System; GIS: Geographic Information System; UD: Utilization Distribution; a-LoCoH:adaptive Local Convex Hull; T-LoCoH: Time-Local Convex Hull, TSD: Time Scaled Distance

DeclarationsAcknowledgements

We thank Dr. Frans Joubert, Christopher Diab, Danica Metlay, Jacqueline Moser, Julie Fabricius Faustrup, AstriFrafjord, Linda Molloy, Christopher Mastropietro for dedicated �eld work, and Marthin Kasaona, GabrielShatumbu, Shayne Kotting, Mathews Daniel, Fredrick Khumub, Malakia Hango, Isaskar Uahoo, Dr. Rob Jackson,and Dr. Larry Patterson for support and assistance in the �eld.  We also thank the Ministry of Environment andTourism Namibia (Etosha National Park) and Boas Erckie, the Department of Wildlife and National ParksBotswana (Chobe National Park) and Dr. Michael Flyman, and the Chobe Enclave Community Trust (LinyantiConservancy) for infrastructure support and permissions.   We are grateful to the Etosha Ecological Instituteand CARACAL Biodiversity Center for the use of lab facilities, and to Dr. Richard Fynn and the OkavangoResearch Institute for materials.  We thank Chris Coetzee, Fritz Stolzenberg, and Trish Williams for supportoutside the park.   We are indebted to Drs. Jerry and Jana Lackey for their constant and unwavering support. We are also grateful to Lotek Wireless Inc., African Lion and Environmental Research Trust (ALERT), Namibian

Page 19/32

Environment & Wildlife Society (NEWS), Gert Uls and Resun GmbH for sponsoring some of the �eld equipment.  We would also like to thank Andy Lyons for assistance with T-LoCoH, and to Dana Seidel for guidance withcoding R analyses.

Authors’ contributions

NAB was involved with the design of the study, obtained funds and equipment sponsorship, secured therelevant permits, collected and analyzed the data, provided the graphics and wrote the manuscript.  VSdeveloped the code to analyze the data and the graphics for the manuscript.  LLV developed the methods andcode to quantify the types of movement interaction, calculated the 95% CI and provided the graphics describingthe method.  KAA supplied the IACUC approval for Botswana.  WMG supplied the IACUC approval for Namibia,and was involved with the design of the study and developments of new methods to analyze the data.  RS andWMG were involved in securing funds for the study, provided guidance on analysis of the data and providededitorial guidance on the draft manuscript.  All authors read and approved the �nal manuscript.

Funding

The study was supported through a doctoral award from the Natural Sciences and Engineering ResearchCouncil of Canada (NSERC), and funds from the School of Mathematics, and the School of Life Sciences fromthe University of KwaZulu-Natal, South Africa and the Department of Environmental, Science, Policy andManagement from the University of California at Berkeley.  The role of these funding bodies were limited toproviding funds and were not involved in the design of the study, data collection, analysis, interpretation of data,nor in writing the manuscript.  

Availability of data and materials

Data collected for this study is available at https://doi.org/10.5281/zenodo.6476304 

Ethics approval and consent to participate

All animal handling procedures were conducted with the ethical clearance of the Animal Research EthicsCommittee of the University of KwaZulu-Natal, South Africa.  Permission to conduct this study were obtainedfrom the Ministry of Environment and Tourism, Namibia (Research/Collecting Permits 1724/2012, 1834/2013,1956/2014) and from the Department of Wildlife and National Parks, Botswana (Research Permit EWT 8/36/4XXVIII (35)).  

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References

Page 20/32

1. Caro TM, Stoner CJ. The potential for interspeci�c competition among African carnivores. BiologicalConservation [Internet]. 2003;110:67–75. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0006320702001775

2. Linnell JDC, Strand O. Interference interactions, co-existence and conservation of mammalian carnivores.Diversity and Distributions. 2000;6:169–76.

3. Palomares F, Caro TM. Interspeci�c Killing among Mammalian Carnivores. The American Naturalist.1999;153:492–508.

4. Terborgh J, Estes JA, Paquet P, Ralls K, Boyd-Heger D, Miller BJ, et al. The Role of Top Carnivores inRegulating Terrrestial Ecosystems. Continental Conservation: Scienti�c Foundations of Regional ReserveNetworks. Washington, D.C.: Island Press; 1999. p. 39–64.

5. Hebblewhite M, White CA, Nietvelt CG, McKenzie JA, Hurd TE, Fryxell JM, et al. Human Activity Mediates aTrophic Cascade caused by Wolves. Ecology [Internet]. 2005;86:2135–44. Available from:http://doi.wiley.com/10.1890/04-1269

�. Lehmann KDS, Montgomery TM, MacLachlan SM, Parker JM, Spagnuolo OS, VandeWetering KJ, et al.Lions, hyenas and mobs (oh my!). Curr Zool [Internet]. 2017;63:313–22. Available from:http://cz.oxfordjournals.org/lookup/doi/10.1093/cz/zow073

7. Nelson JL, Cypher BL, Bjurlin CD, Creel S. Effects of Habitat on Competition Between Kit Foxes andCoyotes. Journal of Wildlife Management [Internet]. 2007 [cited 2019 May 16];71:1467–75. Available from:http://www.bioone.org/doi/abs/10.2193/2006-234

�. Tannerfeldt M, Elmhagen B, Angerbjörn A. Exclusion by interference competition? The relationship betweenred and arctic foxes. Oecologia [Internet]. 2002 [cited 2019 May 16];132:213–20. Available from:http://link.springer.com/10.1007/s00442-002-0967-8

9. Broekhuis F, Cozzi G, Valeix M, McNutt JW, Macdonald DW. Risk avoidance in sympatric large carnivores:reactive or predictive? Journal of Animal Ecology [Internet]. 2013 [cited 2019 May 16];82:1098–105.Available from: http://doi.wiley.com/10.1111/1365-2656.12077

10. Hayward MW, Slotow R. Temporal Partitioning of Activity in Large African Carnivores: Tests of MultipleHypotheses. South African Journal of Wildlife Research [Internet]. 2009 [cited 2019 May 16];39:109–25.Available from: http://www.bioone.org/doi/abs/10.3957/056.039.0207

11. Haswell PM, Jones KA, Kusak J, Hayward MW. Fear, foraging and olfaction: how mesopredators avoidcostly interactions with apex predators. Oecologia [Internet]. 2018 [cited 2019 May 16];187:573–83.Available from: http://link.springer.com/10.1007/s00442-018-4133-3

12. Ramesh T, Kalle R, Sankar K, Qureshi Q. Spatio-temporal partitioning among large carnivores in relation tomajor prey species in Western Ghats: Spatio-temporal partitioning in large carnivores and prey. Journal ofZoology [Internet]. 2012;287:269–75. Available from: http://doi.wiley.com/10.1111/j.1469-7998.2012.00908.x

13. Vanak AT, Fortin D, Thaker M, Ogden M, Owen C, Greatwood S, et al. Moving to stay in place: behavioralmechanisms for coexistence of African large carnivores. Ecology [Internet]. 2013 [cited 2019 May16];94:2619–31. Available from: http://doi.wiley.com/10.1890/13-0217.1

14. Thurfjell H, Ciuti S, Boyce MS. Applications of step-selection functions in ecology and conservation.Movement Ecology [Internet]. 2014 [cited 2019 May 16];2:4. Available from:

Page 21/32

https://movementecologyjournal.biomedcentral.com/articles/10.1186/2051-3933-2-4

15. Andersen GE, Johnson CN, Jones ME. Sympatric predator odour reveals a competitive relationship in size-structured mammalian carnivores. Behavioral Ecology and Sociobiology [Internet]. 2016 [cited 2019 May16];70:1831–41. Available from: http://link.springer.com/10.1007/s00265-016-2189-9

1�. Switalski TA. Coyote foraging ecology and vigilance in response to gray wolf reintroduction in YellowstoneNational Park. Canadian Journal of Zoology [Internet]. 2003 [cited 2019 May 16];81:985–93. Availablefrom: http://www.nrcresearchpress.com/doi/10.1139/z03-080

17. Dröge E, Creel S, Becker MS, M’soka J. Spatial and temporal avoidance of risk within a large carnivore guild.Ecology and Evolution [Internet]. 2017 [cited 2019 May 16];7:189–99. Available from:http://doi.wiley.com/10.1002/ece3.2616

1�. Darnell AM, Graf JA, Somers MJ, Slotow R, Szykman Gunther M. Space Use of African Wild Dogs inRelation to Other Large Carnivores. PLoS ONE [Internet]. 2014 [cited 2021 Jul 22];9:e98846. Available from:https://dx.plos.org/10.1371/journal.pone.0098846

19. Swanson A, Caro T, Davies-Mostert H, Mills MGL, Macdonald DW, Borner M, et al. Cheetahs and wild dogsshow contrasting patterns of suppression by lions. Journal of Animal Ecology [Internet]. 2014 [cited 2019May 16];83:1418–27. Available from: http://doi.wiley.com/10.1111/1365-2656.12231

20. Kushata JNT, Périquet S, Tarakini T, Muzamba M, Mafuwa B, Loveridge AJ, et al. Drivers of diurnal rest siteselection by spotted hyaenas. Journal of Zoology [Internet]. 2018 [cited 2019 May 16];304:132–40.Available from: http://doi.wiley.com/10.1111/jzo.12504

21. Davies AB, Tambling CJ, Marneweck DG, Ranc N, Druce DJ, Cromsigt JPGM, et al. Spatial heterogeneityfacilitates carnivore coexistence. Ecology [Internet]. 2021 [cited 2022 Mar 11];102:e03319. Available from:https://onlinelibrary.wiley.com/doi/10.1002/ecy.3319

22. Périquet S, Fritz H, Revilla E, Macdonald DW, Loveridge AJ, Mtare G, et al. Dynamic interactions betweenapex predators reveal contrasting seasonal attraction patterns. Oecologia [Internet]. 2021 [cited 2021 Jul22];195:51–63. Available from: http://link.springer.com/10.1007/s00442-020-04802-w

23. Lovari S, Minder I, Ferretti F, Mucci N, Randi E, Pellizzi B. Common and snow leopards share prey, but nothabitats: competition avoidance by large predators?: Interactions between two leopard species. Journal ofZoology [Internet]. 2013 [cited 2019 May 16];291:127–35. Available from:http://doi.wiley.com/10.1111/jzo.12053

24. Karanth KU, Srivathsa A, Vasudev D, Puri M, Parameshwaran R, Kumar NS. Spatio-temporal interactionsfacilitate large carnivore sympatry across a resource gradient. Proceedings of the Royal Society B:Biological Sciences [Internet]. 2017 [cited 2019 May 16];284:20161860. Available from:http://rspb.royalsocietypublishing.org/lookup/doi/10.1098/rspb.2016.1860

25. Manlick PJ, Woodford JE, Zuckerberg B, Pauli JN. Niche compression intensi�es competition betweenreintroduced American martens (Martes americana) and �shers (Pekania pennanti). Journal ofMammalogy [Internet]. 2017 [cited 2019 May 16];98:690–702. Available from:https://academic.oup.com/jmammal/article-lookup/doi/10.1093/jmammal/gyx030

2�. Mumma MA, Holbrook JD, Rayl ND, Zieminski CJ, Fuller TK, Organ JF, et al. Examining spatial patterns ofselection and use for an altered predator guild. Oecologia [Internet]. 2017 [cited 2019 May 16];185:725–35.Available from: http://link.springer.com/10.1007/s00442-017-3971-8

Page 22/32

27. Petrov PR, Popova ED, Zlatanova DP. Niche Partitioning among the Red Fox Vulpes vulpes (L.), StoneMarten Martes foina (Erxleben) and Pine Marten Martes martes (L.) in Two Mountains in Bulgaria. ActaZoologica Bulgarica. 2016;68:375–90.

2�. Swanson A, Arnold T, Kosmala M, Forester J, Packer C. In the absence of a “landscape of fear”: How lions,hyenas, and cheetahs coexist. Ecology and Evolution [Internet]. 2016 [cited 2019 May 16];6:8534–45.Available from: http://doi.wiley.com/10.1002/ece3.2569

29. Hearn AJ, Cushman SA, Ross J, Goossens B, Hunter LTB, Macdonald DW. Spatio-temporal ecology ofsympatric felids on Borneo. Evidence for resource partitioning? PLOS ONE [Internet]. 2018 [cited 2019 May16];13:e0200828. Available from: https://dx.plos.org/10.1371/journal.pone.0200828

30. Durant SM. Living with the enemy: avoidance of hyenas and lions by cheetahs in the Serengeti. BehavioralEcology [Internet]. 2000 [cited 2019 May 16];11:624–32. Available from:https://academic.oup.com/beheco/article-lookup/doi/10.1093/beheco/11.6.624

31. de Satgé J, Teichman K, Cristescu B. Competition and coexistence in a small carnivore guild. Oecologia[Internet]. 2017 [cited 2019 May 16];184:873–84. Available from: http://link.springer.com/10.1007/s00442-017-3916-2

32. Moleón M, Sánchez-Zapata JA, Selva N, Donázar JA, Owen-Smith N. Inter-speci�c interactions linkingpredation and scavenging in terrestrial vertebrate assemblages. Biological Reviews [Internet]. 2014 [cited2019 May 16];89:1042–54. Available from: http://doi.wiley.com/10.1111/brv.12097

33. Périquet S, Fritz H, Revilla E. The Lion King and the Hyaena Queen: large carnivore interactions andcoexistence. Biological Reviews [Internet]. 2015 [cited 2019 May 16];90:1197–214. Available from:http://doi.wiley.com/10.1111/brv.12152

34. Hayward MW, Kerley GIH. Prey preferences and dietary overlap amongst Africa’s large predators. SouthAfrican Journal of Wildlife Research [Internet]. 2008 [cited 2019 May 16];38:93–108. Available from:http://www.bioone.org/doi/abs/10.3957/0379-4369-38.2.93

35. Kittle AM, Bukombe JK, Sinclair ARE, Mduma SAR, Fryxell JM. Landscape-level movement patterns by lionsin western Serengeti: comparing the in�uence of inter-speci�c competitors, habitat attributes and preyavailability. Movement Ecology [Internet]. 2016 [cited 2019 May 16];4:17. Available from:http://movementecologyjournal.biomedcentral.com/articles/10.1186/s40462-016-0082-9

3�. Loveridge AJ, Valeix M, Davidson Z, Murindagomo F, Fritz H, Macdonald DW. Changes in home range sizeof African lions in relation to pride size and prey biomass in a semi-arid savanna. Ecography [Internet].2009;32:953–62. Available from: http://doi.wiley.com/10.1111/j.1600-0587.2009.05745.x

37. Yiu S-W, Parrini F, Karczmarski L, Keith M. Home range establishment and utilization by reintroduced lions(Panthera leo) in a small South African wildlife reserve. Integrative Zoology [Internet]. 2017 [cited 2019 May16];12:318–32. Available from: http://doi.wiley.com/10.1111/1749-4877.12243

3�. Zehnder A, Henley S, Weibel R. Home ranges of lions in the Kalahari, Botswana exhibit vast sizes and hightemporal variability. Zoology [Internet]. 2018 [cited 2019 May 16];128:46–54. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0944200617301575

39. Mills MGL. Kalahari Hyenas, The Comparative Behavioral Ecology of Two Species. London, UnitedKingdom: Chapman and Hall; 1990.

Page 23/32

40. Tilson RL, Henschel JR. Spatial arrangement of spotted hyaena groups in a desert environment, Namibia.African Journal of Ecology [Internet]. 1986 [cited 2019 May 16];24:173–80. Available from:http://doi.wiley.com/10.1111/j.1365-2028.1986.tb00358.x

41. Boydston EE, Morelli TL, Holekamp KE. Sex Differences in Territorial Behavior Exhibited by the SpottedHyena (Hyaenidae, Crocuta crocuta). Ethology [Internet]. 2001 [cited 2019 May 16];107:369–85. Availablefrom: http://doi.wiley.com/10.1046/j.1439-0310.2001.00672.x

42. Henschel JR, Skinner JD. Territorial Behaviour by a Clan of Spotted Hyaenas (Crocuta crocuta). Ethology.1991;88:223–35.

43. Smith JE, Kolowski JM, Graham KE, Dawes SE, Holekamp KE. Social and ecological determinants of�ssion–fusion dynamics in the spotted hyaena. Animal Behaviour [Internet]. 2008 [cited 2019 May16];76:619–36. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0003347208001814

44. Lyons AJ, Turner WC, Getz WM. Home range plus: a space-time characterization of movement over reallandscapes. Movement Ecology [Internet]. 2013 [cited 2019 May 16];1:2. Available from:https://movementecologyjournal.biomedcentral.com/articles/10.1186/2051-3933-1-2

45. Getz WM, Wilmers CC. A local nearest-neighbor convex-hull construction of home ranges and utilizationdistributions. Ecography [Internet]. 2004 [cited 2019 May 16];27:489–505. Available from:http://doi.wiley.com/10.1111/j.0906-7590.2004.03835.x

4�. Getz WM, Fortmann-Roe S, Cross PC, Lyons AJ, Ryan SJ, Wilmers CC. LoCoH: Nonparameteric KernelMethods for Constructing Home Ranges and Utilization Distributions. PLoS ONE [Internet]. 2007 [cited 2019May 16];2:e207. Available from: https://dx.plos.org/10.1371/journal.pone.0000207

47. Lyons A. Visualizing Wildlife Conservation and Development in Southern Africa: A Multi-Optic Approach[PhD Dissertation]. University of California, Berkeley; 2012.

4�. Vissat LL, Blackburn JK, Getz WM. A relative-motion method for parsing spatio-temporal behaviour ofdyads using GPS relocation data [Internet]. Ecology; 2021 Jan. Available from:http://biorxiv.org/lookup/doi/10.1101/2021.01.04.425238

49. Benbow EM, Tomberlin JK, Tarone AM. Carrion Ecology, Evolution, and Their Applications. Boca Raton,London, New York: CRC Press Tayloy & Francis Group; 2016.

50. Moleón M, Sánchez-Zapata JA, Sebastián-González E, Owen-Smith N. Carcass size shapes the structureand functioning of an African scavenging assemblage. Oikos [Internet]. 2015 [cited 2019 May16];124:1391–403. Available from: http://doi.wiley.com/10.1111/oik.02222

51. Bagamian KH, Alexander KA, Had�eld TL, Blackburn JK. Ante- and postmortem diagnostic techniques foranthrax: Rethinking pathogen exposure and the geographic extent of the disease in wildlife. Journal ofWildlife Diseases [Internet]. 2013 [cited 2022 Mar 28];49:786–801. Available from:http://www.bioone.org/doi/10.7589/2013-05-126

52. Getz WM. Biomass transformation webs provide a uni�ed approach to consumer-resource modelling:Biomass transformation webs. Ecology Letters [Internet]. 2011 [cited 2020 Jun 21];14:113–24. Availablefrom: http://doi.wiley.com/10.1111/j.1461-0248.2010.01566.x

53. Dorai-Raj S. Binomial Con�dence Intervals For Several Parameterizations. CRAN Package “binom”; 2014.

54. Schoener TW. The Compression Hypothesis and Temporal Resource Partitioning. Proceedings of theNational Academy of Sciences of the United States of America [Internet]. 1974;71:4169–72. Available from:

Page 24/32

http://www.jstor.org/stable/64150

55. Hassell MP, Comins HN, May RM. Species coexistence and self-organizing spatial dynamics. Nature[Internet]. 1994 [cited 2020 Mar 19];370:290–2. Available from: http://www.nature.com/articles/370290a0

5�. Fedriani JM, Palomares F, Delibes M. Niche relations among three sympatric Mediterranean carnivores.Oecologia [Internet]. 1999;121:138–48. Available from: http://www.jstor.org/stable/4222449

57. Albrecht M, Gotelli NJ. Spatial and Temporal Niche Partitioning in Grassland Ants. Oecologia [Internet].2001;126:134–41. Available from: http://www.jstor.org/stable/4222828

5�. Redfern JV, Ryan SJ, Getz WM. De�ning herbivore assemblages in the Kruger National Park: a correlativecoherence approach. Oecologia [Internet]. 2006 [cited 2020 Feb 17];146:632–40. Available from:http://link.springer.com/10.1007/s00442-005-0235-9

59. Valeix M, Chamaillé-Jammes S, work(s): HFR. Interference Competition and Temporal Niche Shifts:Elephants and Herbivore Communities at Waterholes. Oecologia [Internet]. 2007;153:739–48. Availablefrom: http://www.jstor.org/stable/40213022

�0. Ritchie EG, Martin JK, Johnson CN, Fox BJ. Separating the in�uences of environment and speciesinteractions on patterns of distribution and abundance: competition between large herbivores. Journal ofAnimal Ecology [Internet]. 2009 [cited 2019 May 16];78:724–31. Available from:http://doi.wiley.com/10.1111/j.1365-2656.2008.01520.x

�1. Darmon G, Calenge C, Loison A, Jullien J-M, Maillard D, Lopez J-F. Spatial distribution and habitat selectionin coexisting species of mountain ungulates. Ecography [Internet]. 2012 [cited 2019 May 16];35:44–53.Available from: http://doi.wiley.com/10.1111/j.1600-0587.2011.06664.x

�2. Durant SM. Competition refuges and coexistence: an example from Serengeti carnivores. Journal of AnimalEcology [Internet]. 1998 [cited 2019 May 16];67:370–86. Available from:https://onlinelibrary.wiley.com/doi/abs/10.1046/j.1365-2656.1998.00202.x

�3. Edwards S, Gange AC, Wiesel I. Spatiotemporal resource partitioning of water sources by African carnivoreson Namibian commercial farmlands: Carnivore resource partitioning at water sources. Journal of Zoology[Internet]. 2015 [cited 2019 May 16];297:22–31. Available from: http://doi.wiley.com/10.1111/jzo.12248

�4. Ramesh T, Kalle R, Downs CT. Staying safe from top predators: patterns of co-occurrence and inter-predatorinteractions. Behavioral Ecology and Sociobiology [Internet]. 2017 [cited 2019 May 16];71:41. Availablefrom: http://link.springer.com/10.1007/s00265-017-2271-y

�5. Atwood TC, Fry TL, Leland BR. Partitioning of anthropogenic watering sites by desert carnivores. TheJournal of Wildlife Management [Internet]. 2011 [cited 2020 Feb 17];75:1609–15. Available from:http://doi.wiley.com/10.1002/jwmg.225

��. Cooper SM. Optimal hunting group size: the need for lions to defend their kills against loss to spottedhyaenas. African Journal of Ecology. 1991;29:130–6.

�7. Honer OP, Wachter B, East ML, Hofer H. The response of spotted hyaenas to long-term changes in preypopulations: functional response and interspeci�c kleptoparasitism. Journal of Animal Ecology [Internet].2002 [cited 2019 May 16];71:236–46. Available from: http://doi.wiley.com/10.1046/j.1365-2656.2002.00596.x

��. Trinkel M, Kastberger G. Competitive interactions between spotted hyenas and lions in the Etosha NationalPark, Namibia. African Journal of Ecology [Internet]. 2005 [cited 2019 May 16];43:220–4. Available from:

Page 25/32

http://doi.wiley.com/10.1111/j.1365-2028.2005.00574.x

�9. Watts HE, Holekamp KE. Interspeci�c competition in�uences reproduction in spotted hyenas. Journal ofZoology [Internet]. 2008 [cited 2019 May 16];276:402–10. Available from:http://doi.wiley.com/10.1111/j.1469-7998.2008.00506.x

70. Watts HE, Blankenship LM, Dawes SE, Holekamp KE. Responses of Spotted Hyenas to Lions Re�ectIndividual Differences in Behavior. Ethology [Internet]. 2010 [cited 2019 May 16];116:1199–209. Availablefrom: http://doi.wiley.com/10.1111/j.1439-0310.2010.01833.x

71. Périquet S, Valeix M, Claypole J, Drouet-Hoguet N, Salnicki J, Mudimba S, et al. Spotted hyaenas switchtheir foraging strategy as a response to changes in intraguild interactions with lions: Intraguild interactionsin�uence diet in large carnivores. Journal of Zoology [Internet]. 2015 [cited 2019 May 16];297:245–54.Available from: http://doi.wiley.com/10.1111/jzo.12275

72. Périquet S, Mapendere C, Revilla E, Banda J, Macdonald DW, Loveridge AJ, et al. A potential role forinterference competition with lions in den selection and attendance by spotted hyaenas. MammalianBiology [Internet]. 2016 [cited 2019 May 16];81:227–34. Available from:https://linkinghub.elsevier.com/retrieve/pii/S1616504715001032

73. Schaller GB. The Serengeti lion: a study of predator-prey relations. Chicago, Ill.; Bristol: University ofChicago Press; 1972.

74. Packer C, Pusey AE. Divided We Fall: Cooperation among Lions. Sci Am [Internet]. 1997 [cited 2022 Apr5];276:52–9. Available from: https://www.scienti�camerican.com/article/divided-we-fall-cooperation-among-l

75. Barker NA, Joubert FG, Kasaona M, Shatumbu G, Stowbunenko V, Alexander KA, et al. Coursing hyenas andstalking lions: the potential for inter- and intraspeci�c interactions [Internet]. Ecology; 2022 [cited 2022 Apr21]. Available from: http://biorxiv.org/lookup/doi/10.1101/2022.02.23.481665

7�. Benhamou S, Valeix M, Chamaillé-Jammes S, Macdonald DW, Loveridge AJ. Movement-based analysis ofinteractions in African lions. Animal Behaviour [Internet]. 2014 [cited 2020 Mar 19];90:171–80. Availablefrom: https://linkinghub.elsevier.com/retrieve/pii/S0003347214000645

77. Yiu SW, Karczmarski L, Parrini F, Keith M. Resource selection in reintroduced lions and the in�uence ofintergroup interactions. J Zool [Internet]. 2019 [cited 2020 Jul 2];307:111–24. Available from:https://onlinelibrary.wiley.com/doi/abs/10.1111/jzo.12620

7�. Mosser A, Packer C. Group territoriality and the bene�ts of sociality in the African lion, Panthera leo. AnimalBehaviour [Internet]. 2009 [cited 2020 Mar 19];78:359–70. Available from:https://linkinghub.elsevier.com/retrieve/pii/S000334720900219X

79. Bailey I, Myatt JP, Wilson AM. Group hunting within the Carnivora: physiological, cognitive andenvironmental in�uences on strategy and cooperation. Behavioral Ecology and Sociobiology [Internet].2013 [cited 2019 May 16];67:1–17. Available from: http://link.springer.com/10.1007/s00265-012-1423-3

�0. Trinkel M, Fleischmann PH, Kastberger G. Comparison of land-use strategies of spotted hyenas (Crocutacrocuta, Erxleben) in different ecosystems. African Journal of Ecology [Internet]. 2006 [cited 2019 May16];44:537–9. Available from: http://doi.wiley.com/10.1111/j.1365-2028.2006.00658.x

�1. Trinkel M, Fleischmann PH, Steindorfer AF, Kastberger G. Spotted hyenas (Crocuta crocuta) followmigratory prey. Seasonal expansion of a clan territory in Etosha, Namibia. Journal of Zoology [Internet].

Page 26/32

2004 [cited 2019 May 16];264:125–33. Available from: http://doi.wiley.com/10.1017/S0952836904005588

�2. Hofer H, East ML. The commuting system of Serengeti spotted hyaenas: how a predator copes withmirgratory prey. I. Social organization. Animal Behaviour. 1993;46:547–57.

�3. Cooper DSM. Clan sizes of Spotted Hyaenas in the Savuti Region of the Chobe National Park, Botswana.Botswana Notes and Records. 1989;21:121–33.

�4. Amarasekare P. Interference competition and species coexistence. Proceedings of the Royal Society ofLondon Series B: Biological Sciences [Internet]. 2002 [cited 2019 May 16];269:2541–50. Available from:http://www.royalsocietypublishing.org/doi/10.1098/rspb.2002.2181

�5. Vance RR. Interference Competition and the Coexistence of Two Competitors on a Single LimitingResource. Ecology [Internet]. 1984;65:1349–57. Available from: http://doi.wiley.com/10.2307/1939115

��. Zemel A, Lubin Y. Inter-group competition and stable group sizes. Animal Behaviour [Internet]. 1995 [cited2019 May 16];50:485–8. Available from: https://linkinghub.elsevier.com/retrieve/pii/S000334728570262X

�7. McComb K, Packer C, Pusey AE. Roaring and numerical assessment in contests between groups of femalelions, Panthera leo. Animal Behaviour. 1994;47:379–87.

��. Grinnell J, McComb K. Maternal grouping as a defense against infanticide by males: evidence from �eldplayback experiments on African lions. Behavioral Ecology [Internet]. 1996 [cited 2021 Feb 28];7:55–9.Available from: https://academic.oup.com/beheco/article-lookup/doi/10.1093/beheco/7.1.55

�9. Cozzi G, Broekhuis F, McNutt JW, Turnbull LA, Macdonald DW, Schmid B. Fear of the dark or dinner bymoonlight? Reduced temporal partitioning among Africa’s large carnivores. Ecology [Internet]. 2012 [cited2019 May 16];93:2590–9. Available from: http://doi.wiley.com/10.1890/12-0017.1

90. Webster H, McNutt JW, McComb K. Eavesdropping and Risk Assessment Between Lions, Spotted Hyenasand African Wild Dogs. Ethology [Internet]. 2010 [cited 2019 May 16];116:233–9. Available from:http://doi.wiley.com/10.1111/j.1439-0310.2009.01729.x

91. Long JA, Nelson TA, Webb SL, Gee KL. A critical examination of indices of dynamic interaction for wildlifetelemetry studies. J Anim Ecol [Internet]. 2014 [cited 2020 Mar 19];83:1216–33. Available from:http://doi.wiley.com/10.1111/1365-2656.12198

92. Joo R, Etienne M-P, Bez N, Mahévas S. Metrics for describing dyadic movement: a review. Mov Ecol[Internet]. 2018 [cited 2020 Mar 19];6:26. Available from:https://movementecologyjournal.biomedcentral.com/articles/10.1186/s40462-018-0144-2

93. Bertrand MR, DeNicola AJ, Beissinger SR, Swihart RK. Effects of Parturition on Home Ranges and SocialA�liations of Female White-Tailed Deer. The Journal of Wildlife Management [Internet]. 1996 [cited 2022Apr 4];60:899. Available from: https://www.jstor.org/stable/3802391?origin=crossref

94. Doncaster CP. Non-parametric estimates of interaction from radio-tracking data. Journal of TheoreticalBiology [Internet]. 1990 [cited 2022 Apr 4];143:431–43. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0022519305800207

95. Kenward RE, Marcstrom V, Karlbom M. Post-nestling behaviour in goshawks, Accipiter gentilis: II. Sexdifferences in sociality and nest-switching. Anim Behav. 1993;46:371–8.

9�. Brotherton PNM, Pemberton JM, Komers PE, Malarky G. Genetic and behavioural evidence of monogamy ina mammal, Kirk’s dik–dik (;Madoqua kirkii). Proc R Soc Lond B [Internet]. 1997 [cited 2022 Apr 4];264:675–81. Available from: https://royalsocietypublishing.org/doi/10.1098/rspb.1997.0096

Page 27/32

97. Long JA, Webb SL, Nelson TA, Gee KL. Mapping areas of spatial-temporal overlap from wildlife trackingdata. Mov Ecol [Internet]. 2015 [cited 2022 Apr 4];3:38. Available from:http://www.movementecologyjournal.com/content/3/1/38

9�. Minta SC. Tests of Spatial and Temporal Interaction Among Animals. Ecological Applications [Internet].1992 [cited 2022 Apr 4];2:178–88. Available from: http://doi.wiley.com/10.2307/1941774

99. Richman JS, Moorman JR. Physiological time-series analysis using approximate entropy and sampleentropy. American Journal of Physiology-Heart and Circulatory Physiology [Internet]. 2000 [cited 2022 Apr4];278:H2039–49. Available from: https://www.physiology.org/doi/10.1152/ajpheart.2000.278.6.H2039

100. Long JA, Nelson TA. Measuring Dynamic Interaction in Movement Data: Measuring Dynamic Interaction.Transactions in GIS [Internet]. 2013 [cited 2022 Apr 4];17:62–77. Available from:https://onlinelibrary.wiley.com/doi/10.1111/j.1467-9671.2012.01353.x

101. Dodge S, Weibel R, Forootan E. Revealing the physics of movement: Comparing the similarity of movementcharacteristics of different types of moving objects. Computers, Environment and Urban Systems [Internet].2009 [cited 2022 Apr 4];33:419–34. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0198971509000556

102. Edelhoff H, Signer J, Balkenhol N. Path segmentation for beginners: an overview of current methods fordetecting changes in animal movement patterns. Mov Ecol [Internet]. 2016 [cited 2021 Jan 23];4:21.Available from: http://movementecologyjournal.biomedcentral.com/articles/10.1186/s40462-016-0086-5

103. Michelot T, Langrock R, Patterson TA. moveHMM: an R package for the statistical modelling of animalmovement data using hidden Markov models. Methods in Ecology and Evolution [Internet]. 2016 [cited2019 May 16];7:1308–15. Available from: http://doi.wiley.com/10.1111/2041-210X.12578

104. Miller JA. Towards a Better Understanding of Dynamic Interaction Metrics for Wildlife: a Null ModelApproach: Null model approach to dynamic interactions. Transactions in GIS [Internet]. 2015 [cited 2019May 16];19:342–61. Available from: http://doi.wiley.com/10.1111/tgis.12149

105. Gurarie E, Bracis C, Delgado M, Meckley TD, Kojola I, Wagner CM. What is the animal doing? Tools forexploring behavioural structure in animal movements. Journal of Animal Ecology [Internet]. 2016 [cited2019 May 16];85:69–84. Available from: http://doi.wiley.com/10.1111/1365-2656.12379

10�. Wilson EE, Wolkovich EM. Scavenging: how carnivores and carrion structure communities. Trends inEcology & Evolution [Internet]. 2011 [cited 2019 May 16];26:129–35. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0169534710003034

107. Ripple WJ, Estes JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, et al. Status and EcologicalEffects of the World’s Largest Carnivores. Science [Internet]. 2014 [cited 2019 May 16];343:1241484.Available from: http://www.sciencemag.org/cgi/doi/10.1126/science.1241484

10�. Riggio J, Jacobson A, Dollar L, Bauer H, Becker M, Dickman A, et al. The size of savannah Africa: a lion’s(Panthera leo) view. Biodiversity and Conservation [Internet]. 2013 [cited 2019 May 16];22:17–35. Availablefrom: http://link.springer.com/10.1007/s10531-012-0381-4

109. Gurd DB, Nudds TD, Rivard DH. Conservation of Mammals in Eastern North American Wildlife Reserves:How Small Is Too Small? Conservation Biology [Internet]. 2001;15:1355–63. Available from:http://www.jstor.org/stable/3061491

Page 28/32

110. Tuqa JH, Funston P, Musyoki C, Ojwang GO, Gichuki NN, Bauer H, et al. Impact of severe climate variabilityon lion home range and movement patterns in the Amboseli ecosystem, Kenya. Global Ecology andConservation [Internet]. 2014 [cited 2019 May 16];2:1–10. Available from:https://linkinghub.elsevier.com/retrieve/pii/S2351989414000122

111. Zielinski WJ, Tucker JM, Rennie KM. Niche overlap of competing carnivores across climatic gradients andthe conservation implications of climate change at geographic range margins. Biological Conservation[Internet]. 2017 [cited 2019 May 16];209:533–45. Available from:https://linkinghub.elsevier.com/retrieve/pii/S0006320716308588

Figures

Figure 1

Location of the study areas. The map of the African continent shows the countries of Namibia and Botswanashaded, and the protected areas within these countries where the study was conducted. Maps were generatedwith ArcGIS (ESRI ArcMap v10.0).

Page 29/32

Figure 2

A description of the relative-motion method in using approach and retreat behaviours to identify the type ofinteraction occurring between individuals. (a) Absolute heading of individual A (headA) and B (headB) areshown in blue (see text for details). Relative head of A with respect to B (dirAB) and B with respect to A (dirBA)are shown in green (see text for details), and the difference between absolute and relative for individual A (diffA)and B (diffB) are shown in red. (b) Legend for barplots depicting the simultaneous movement behaviours ofboth individuals A and B in the dyad. The blue represents movements towards, red is movements away, andpurple orthogonal movements.

Page 30/32

Figure 3

Hull association metrics of a single lion (NU-33865) and spotted hyena (GO-33869) dyad from Etosha,Namibia. Overlapping hulls for the individual in the dyad are shown for the lion (a, c) and the hyena (b, d). Plotsdepict the frequency of (a, b), and minimum time lag (c, d) in spatially overlapping hulls for the dyad. In (a, b),red colours indicate the greatest number of overlapping hulls, and purple the least. In (c, d), purple coloursindicate a time lag of >24 hr between overlapping hulls, and red colours indicate when the dyad occurred in thesame area at the same time. Histograms show the pairwise distribution of the mean centroid distance (seeSupplementary Appendix 1, Additional File 3) for temporally overlapping hulls in the home range (e), and core

Page 31/32

area overlap (f) between the lion-hyena dyad. The red line is the histogram envelope from an equivalent numberof randomly matched pairs of hulls, and represents the null model of no interaction. An asterisk denotessigni�cance (p < 0.05).  

Figure 4

Movement distribution and ratio plots for two lion and spotted hyena dyads (OF-34309 & SA-33872, SU-33868 &SA-33872) from Etosha, Namibia. Panels on the far left indicate the difference in the heading of one individual

Page 32/32

towards the other individual, while panels on the far right indicate the dyad’s simultaneous movement types forthe distance interval of 0-10 km. Panels second from the left indicate the ratio of the blue/(blue + red) areas forthe individual, and panels second from the right indicate the distribution for the ratio. Upper three panels in both�gures (a, b) represent the movements of the lion towards the hyena in the dyad, and the lower three panels inboth �gures (a, b) represent the movements of the hyena towards the lion in the dyad. In (a), the lion movesmostly towards the hyena while the hyena moves mostly away, and in (b) both the lion and the hyenademonstrate mostly random movements towards each other.  

Supplementary Files

This is a list of supplementary �les associated with this preprint. Click to download.

AdditionalFile1.pdf

AdditionalFile2.pdf

AdditionalFile3.pdf

AdditionalFile4.pdf

AdditionalFile5.pdf

AdditionalFile6.pdf

AdditionalFile7.pdf

AdditionalFile8.pdf

AdditionalFile9.pdf

AdditionalFile10.pdf

AdditionalFile11.pdf

AdditionalFile12.pdf