Population recovery highlights spatial organization dynamics in adult leopards

Abstract

Polygynous species follow sex-specific spacing patterns to maximize reproductive
success, and changes in population density under otherwise stable environmental
conditions likely provoke sex-specific responses in spacing patterns. A classical
dual reproductive strategy hypothesis posits that female home range size and overlap
are set by habitat productivity and remain stable under increasing population
density, whereas male home range size and overlap decrease with increased mate
competition. An alternative dispersal-regulated strategy predicts that females relinquish
part of their home range to philopatric daughters and form matrilineal clusters,
while adult male spacing is stable with density-dependent subadult male
emigration rates. We used 11 years of telemetry data to assess the response of
adult leopard Panthera pardus spacing following the release of harvest pressure.
Female annual home ranges and core areas were smaller than in males. Intersexual
overlap was larger than intra-sexual overlap in males or in females. As leopard
density increased, female home range size and inter-annual fidelity in home range
use decreased, and females formed matrilineal kin clusters. In contrast, male leopards
maintained large home ranges, and did not track female home range contraction.
Spacing dynamics in adult leopards was consistent with dispersal-regulated
strategies, and did not support a classical dual reproductive strategy. Our study
suggests possible hidden lag effects of harvest disturbance on spacing dynamics
that are not necessarily apparent when only assessing demographic recovery of harvested
populations.

Full text

Population recovery highlights spatial organization
dynamics in adult leopards
J. Fattebert
1,2
, G. A. Balme
1,3
, H. S. Robinson
1,4
, T. Dickerson
1
, R. Slotow
2,5
& L. T. B. Hunter
1,2
1 Panthera, New York, NY, USA
2 School of Life Sciences, University of KwaZulu-Natal, Durban, South Africa
3 Department of Biological Sciences, University of Cape Town, Cape Town, South Africa
4 College of Forestry and Conservation, University of Montana, Missoula, MT, USA
5 Department of Genetics, Evolution and Environment, University College, London, UK
Keywords
density-dependence; dual reproductive strategy;
home range; Panthera pardus; South Africa.
Correspondence
Julien Fattebert, School of Life Sciences,
University of KwaZulu-Natal, Durban,
South Africa
Email: julien.fattebert@gmail.com
Editor: Matthew Hayward
Received 9 November 2015; revised 2 March
2016; accepted 2 March 2016
doi:10.1111/jzo.12344
Abstract
Polygynous species follow sex-specific spacing patterns to maximize reproductive
success, and changes in population density under otherwise stable environmental
conditions likely provoke sex-specific responses in spacing patterns. A classical
dual reproductive strategy hypothesis posits that female home range size and over-
lap are set by habitat productivity and remain stable under increasing population
density, whereas male home range size and overlap decrease with increased mate
competition. An alternative dispersal-regulated strategy predicts that females relin-
quish part of their home range to philopatric daughters and form matrilineal clus-
ters, while adult male spacing is stable with density-dependent subadult male
emigration rates. We used 11 years of telemetry data to assess the response of
adult leopard Panthera pardus spacing following the release of harvest pressure.
Female annual home ranges and core areas were smaller than in males. Intersexual
overlap was larger than intra-sexual overlap in males or in females. As leopard
density increased, female home range size and inter-annual fidelity in home range
use decreased, and females formed matrilineal kin clusters. In contrast, male leop-
ards maintained large home ranges, and did not track female home range contrac-
tion. Spacing dynamics in adult leopards was consistent with dispersal-regulated
strategies, and did not support a classical dual reproductive strategy. Our study
suggests possible hidden lag effects of harvest disturbance on spacing dynamics
that are not necessarily apparent when only assessing demographic recovery of har-
vested populations.
Introduction
Patterns of animal social and spatial organization (hereafter:
‘spacing’) are assumed to reflect individual attempts to maxi-
mize fitness while accounting for the distribution of limiting
resources and competing conspecifics (Mitchell & Powell,
2012). Broad-scale intra-specific variations in spacing often fol-
low gradients of habitat productivity and resource availability
across a species’range (McLoughlin, Ferguson & Messier,
2000; Marker & Dickman, 2005; Nilsen, Herfindal & Linnell,
2005; Hayward et al., 2009). Population density can also influ-
ence spacing patterns (Dahle & Swenson, 2003), and high
levels of harvest have been shown to affect spacing patterns in
carnivores (Woodroffe et al., 2006; Davidson et al., 2011). In
polygynous species, each sex follows a different spacing strat-
egy to maximize reproductive success (dual reproductive strat-
egy: Clutton-Brock & Harvey, 1978; Sandell, 1989), and
changes in population density likely provokes sex-specific
responses on spacing (Maletzke et al., 2014).
As large carnivore population densities are driven largely by
prey availability (Carbone & Gittleman, 2002; Hayward,
O’Brien & Kerley, 2007), it remains challenging to disentangle
the effects of resource availability and population density on
spacing (Benson, Chamberlain & Leopold, 2006). Experimental
manipulations of carnivore populations to enable rigorous
hypothesis testing are limited by significant logistical and ethi-
cal constraints, but the study of harvested populations can give
insight into the effect of population density on a dual strategy
of spacing (Maletzke et al., 2014; Keehner et al., 2015). Clas-
sically, the dual reproductive strategy predicts that female
space-use patterns follow a foraging optimality rule of area
minimization, and that females achieve higher reproductive
success and maximize fitness by using resources within famil-
iar, stable home ranges (Sandell, 1989). Therefore, when a
population is reduced below carrying capacity under otherwise
stable environmental conditions, female spacing should remain
stable when per capita resource availability is stable or
increased (Logan & Sweanor, 2001). In contrast, males
Journal of Zoology  (2016) – ª2016 The Zoological Society of London 1
Journal of Zoology. Print ISSN 0952-8369

maximize reproductive success by securing access to several
females, and mate competition constrains male space-use (San-
dell, 1989). When voids are created by the removal of male
conspecifics through harvest, mate competition predicts an
increase in male home range size and intra-sexual overlap
(Logan & Sweanor, 2001).
Alternatively, spacing dynamics might be regulated by disper-
sal and philopatry. In polygynous mammals, natal dispersal is
generally male-biased and females tend to be philopatric (Green-
wood, 1980). Under a dispersal-regulated alternative to the classi-
cal dual strategy, females are predicted to relinquish part of their
home range to philopatric daughters at independence, and overlap
more with related females that can benefit from local knowledge
of resources, and inclusive fitness (Anderson, 1989; Lambin,
Aars & Piertney, 2001). In contrast, mate competition predicts
density-dependent subadult male emigration rates (Fattebert
et al., 2015a) with stable adult male spacing.
Understanding how animals adjust their spacing to changes in
population density that can impact their mating system has impli-
cations for how quickly populations may rebound from stochastic
events, or heavy harvest. Among solitary felids, sex-specific spac-
ing response to population decline following the classical dual
strategy has been documented in cougars Puma concolor ( Logan
& Sweanor, 2001; Maletzke et al., 2014). Conversely, in other
populations where harvest pressure has been reduced or an
expanding population is monitored, spacing dynamics have been
less predictable and more variable (bobcat Lynx rufus: Benson
et al., 2006; tiger Panthera tigris: Goodrich et al., 2010; Eura-
sian lynx Lynx lynx: Pesenti & Zimmermann, 2013). To address
this research gap, we assessed the two alternatives to the dual
strategy hypothesis in a recovering leopard Panthera pardus pop-
ulation with increasing density following release from harvest
(Balme, Slotow & Hunter, 2009).
Leopards are polygynous, solitary large carnivores with den-
sities and spacing related to prey densities at the inter-popula-
tion level (Marker & Dickman, 2005). Generally, subadult
female leopards are philopatric, and subadult males disperse
(Fattebert et al., 2013, 2015a). In this study, we measured
adult female and male leopard’s home range size, inter-indivi-
dual overlap, and fidelity of home range use over contrasting
population densities during 11 years. As the population density
increased, we predicted under a classical dual reproductive
strategy (1) female home range size and overlap to be set by
habitat productivity, and therefore to remain stable under stable
environmental conditions; (2) male home range size and over-
lap to decrease due to increased mate competition. Alterna-
tively, under a dispersal-regulated strategy, we predicted (3)
female home range size to decrease, and intra-sexual overlap
to increase while forming matrilineal kin clusters; and (4) male
spacing to remain stable (Table 1).
Materials and methods
Study area and study population
We studied leopard spatial organization in Phinda Private
Game Reserve (hereafter ‘Phinda’; 234 km
2
) in northern Kwa-
Zulu-Natal, South Africa (27°33-27°55
0S; 32°060–32°260E)
Table 1 Predictions and corresponding expected observations for the two competing dual reproductive strategy hypotheses for home range size and overlap patterns in adult female male
and male leopards in Phinda Game Reserve, 2002–2012
Sex Hypothesis Predictions Expected observations
Female Classical dual
reproductive
strategy
Home range size and overlap are limited by resource availability.
Home range size and overlap are not density-dependent
No change in home range size and overlap
under stable prey availability and increasing
population density
Dispersal-regulated
strategy
Female philopatry leads to the formation of matrilineal kin clusters.
Adult female relinquish portion of their home range to philopatric
female progeny
Decrease in home range size and larger overlap
among related female conspecifics under
stable prey availability and increasing population density
Male Classical dual
reproductive
strategy
Home range size and overlap are limited by social interactions and
mate competition. Home range size and overlap are density
dependent under stable resource availability
Decrease in home range size and overlap under
increasing population density and stable prey availability
Dispersal-regulated
strategy
Mate competition drives male density-dependent male dispersal.
Home range size and overlap are not density-dependent
No change in home range size and overlap under
stable prey availability and increasing population density
2Journal of Zoology  (2016) – ª2016 The Zoological Society of London
Leopard spatial organization dynamics J. Fattebert et al.

from 2002 to 2012. Climate is subtropical with hot, humid
summers (October–March), and warm, dry winters (April–
September). Mean monthly temperatures range from 33°C (Jan-
uary) to 19°C (July). Annual rainfall ranges from 550 to
1150 mm, and occurs mainly in summer (Fattebert et al.,
2013). Although strictly protected in Phinda, leopards move
freely across boundary fences and face greater mortality risk in
surrounding non-protected areas, from a combination of trophy
hunting, problem animal control, and illegal killing for skins
(Balme et al., 2009).
Based on the population demographic parameters documented
elsewhere (Balme et al., 2009; Balme, Slotow & Hunter, 2010),
we defined three separate temporal periods: (1) a Disturbance
period (2002–2004), when the population was in decline (annual
population growth rate k=0.978) due to high levels of mortality
[annual mortality rate (AMR) =0.401]; (2) a Recovery period
(2005–2008) following the implementation of sustainable harvest
protocols along with other conservation interventions in 2005,
when AMR declined to 0.134, annual population growth rate
increased (k=1.136), and leopard density increased from
7.2 1.1 to 11.2 2.1 individuals/100 km
2
; and (3) a Stabi-
lization period (2009–2012), when the population density pla-
teaued at putative carrying capacity (Figure S1). Populations of
the three main leopard prey species in this system (nyala Tragela-
phus angasii, impala Aepyceros melampus and warthog Phaco-
choerus africanus: Balme, Hunter & Slotow, 2007) were stable
over the study period (Figure S1). Overall competitor populations
(lions Panthera leo and spotted hyena Crocutta crocutta) were
also stable (Balme et al., 2010), and the leopard demographic
recovery between 2005–2008 was attributed to the decline of
human-caused mortality (Balme et al., 2009). During the course
of the study, all radio-tracked subadult females remained philopa-
tric, while emigration rate and natal dispersal distance in subadult
males increased with population density (Fattebert et al., 2015a).
Capture and tracking
We captured leopards in Phinda with a combination of cage-
trapping, free-darting, and soft-hold foot-snaring as described
by Balme et al. (2007). We estimated the age of leopards
using morphological cues such as teeth wear and yellowing,
head and body size, weight, facial scarring, tearing of the ears,
nose pigmentation, and development of a dewlap (Stander,
1997; Balme, Hunter & Braczkowski, 2012). We classified
leopards >3 years old as adults. Depending on the accessibility
for radiotracking within or outside the protected area, we fitted
adult leopards with VHF (250 g, Sirtrack Ltd., New Havelock
North, Zealand, 0.5% of adult female body mass) or GPS col-
lars (420 g, Vectronic-Aerospace, Berlin, Germany, 1.2% of
adult female body mass).
We located VHF-collared individuals to approximately
100 m using ground homing or triangulation on average every
3 days. We programmed GPS collars to acquire 2–6fixes
daily. Mean fix acquisition success rate was 86.3 2.8%
(n=17 collars), and we further filtered GPS data for poten-
tially large locational errors removing 3D fixes with Positional
Dilution of Precision (PDOP) >15 and 2D fixes with
PDOP >5 (Lewis et al., 2007). Ezemvelo KwaZulu-Natal
Wildlife provincial authority granted permit for this research
(permit HO/4004/07), and the Animal Ethics Subcommittee of
the University of KwaZulu-Natal Ethics Committee approved
animal handling procedures (approval 051/12/Animal). All
transmitters were retrieved at the end of the study.
Home range size
We used telemetry relocations to compute seasonal (dry season:
April–September; wet season: October–March), and annual
(April–March) fixed kernel utilization distributions (UD; Ker-
nohan, Gitzen & Millspaugh, 2001). We restricted our analysis to
adult individuals with ≥50 relocations available in a given year
(Kernohan et al., 2001). We calculated smoothing factor band-
width h using the ‘solve-the-equation plug-in’method (Gitzen,
Millspaugh & Kernohan, 2006) using the Geospatial Modeling
Environment program (Beyer, 2012). We defined home ranges
and core areas using the 90 and 50% UD isopleths, respectively.
We used annual UDs for all subsequent analyses, as a paired t-test
showed no significant difference in seasonal home range size
(Table S1). Vegetation greenness is an important driver of leop-
ard habitat use (Fattebert et al., 2015b), and we extracted the
mean annual Enhanced Vegetation Index (EVI; 537 m resolution
raster; WAMIS 2015) to each home range and core area as a
proxy for prey density at the individual range level. Vegetation
indices are commonly used to quantify vegetative productivity
(Duncan et al., 2015), and it is often assumed that a relationship
exists between productivity and prey density (Nilsen et al.,
2005). For comparison with the literature, we also present 95%
and 50% Minimum Convex Polygon (MCP; Hayne, 1949) annual
range size (Table S2).
Home range overlap
To assess the inter-individual overlap between concurrently
tracked individuals, we computed the three-dimensional utiliza-
tion distribution overlap index (UDOI; Fieberg & Kochanny,
2005) using the R package adehabitatHR (Calenge, 2006):
UDOI ¼Ai;jZ1
1 Z1
1 d
UDiðx;yÞd
UDjðx;yÞdxdy (1)
where d
UDiðx;yÞis the value of the UD of the animal iat the
point x, y. UDOI yields a single value per interaction, ranging
from 0 (no overlap) to 1 when two uniformly distributed UDs
completely overlap, and takes values >1 when the two UDs
overlap extensively but are not uniformly distributed (Fieberg
& Kochanny, 2005). To assess fidelity of home range use in
individual leopards tracked during consecutive years, we calcu-
lated intra-individual UDOI. For comparison with the literature,
we present the two-dimensional home range overlap indices
(Tables S3, S4).
Statistical analyses
To test for the effect of covariates on home range, and core
area size and overlap, we fitted generalized linear mixed-effect
models (Bolker et al., 2009) with a Gaussian distribution using
Journal of Zoology  (2016) – ª2016 The Zoological Society of London 3
J. Fattebert et al. Leopard spatial organization dynamics

the R package lme4 (Bates, Maechler & Bolker, 2013).To
account for leopards tracked for more than 1 year, we fitted
the identity of the individuals (range size) or the interacting
dyad (overlap) as a random intercept. To assess the general
spacing patterns of the leopard population, we first tested for
sexual differences in home range and core area sizes, and for
differences in sex-specific overlap.
To test the classical and the dispersal-regulated dual strategy
hypotheses, we used an information theoretic approach. For
each sex, we explored the effect of study period (population
density) and EVI (habitat productivity) on home range size,
core area size, or overlap index, or their additive effect in a
full model:
Spatial measurement b0þb1study period þb2EVI þð1jIDÞ
(2)
Through direct observation of telemetered adult females with
cubs, we defined mother and daughters, or female siblings of
same or successive litters as related, and other females as of
unknown relatedness. To assess the existence of matrilineal kin
clusters, we included relatedness as a factor (‘related’vs. ‘un-
known’) in the model set for female–female UDOI, adding one
parameter to the full model:
female-female UDOI b0þb1study period þb2EVI
þb3relatedness þð1jIDÞ(3)
For each spatial measurement, we used the Akaike’s infor-
mation criterion corrected for small sample size (AICc) to
select for the most parsimonious model, including a null model
with no explanatory variable (Burnham & Anderson, 2002). If
top-ranking candidate models were within ΔAICc <2, we per-
formed model averaging to estimate unbiased coefficients of
the parameters using the R package MuMIn (Barton, 2013). To
identify informative parameters in final models, coefficients
were deemed significant when corresponding 90% confidence
interval (CI) did not include 0 (Arnold, 2010). We ran all sta-
tistical analyses in R version 3.0.0 (R Core Team 2013). We
report mean standard error (SE) throughout unless stated
otherwise.
Results
We used telemetry data from 10 adult female leopards tracked
over 1–9 years, and 11 adult males tracked over 1–4 years in
Phinda between 2002 and 2012. We obtained 42 annual female
ranges (393 71 relocations over 9.3 0.5 months; Table S5)
and 25 annual male ranges (509 90 relocations over 6.8 
0.7 months). Overall, female home ranges (29.5 1.6 km
2
)
and core areas (7.3 0.5 km
2
) were significantly smaller than
male home ranges (74.0 7.5 km
2
;b
females
=42.6; 90% CI
60.1, 25.1) and core areas (20.2 2.5 km
2
;b
females
=
12.1; 90% CI 18.2, 6.1).
Inter-individual UDOI was larger in female–male
(0.172 0.027, n=56) than in female–female (0.060 
0.018, n=31; b
female-female
=0.113; 90% CI 0.181, 0.046),
or in male–male overlap (0.045 0.015, n=11; b
male-male
=
0.123; 90% CI 0.216, 0.031). Both sexes showed high
levels of annual home range use fidelity, and there was no sig-
nificant difference in intra-individual UDOI between females
(0.763 0.049, n=31) and males (0.686 0.074, n=13;
b
females
=0.131; 90% CI 0.048, 0.310).
Female spacing dynamics
In the Stabilization period, average female home range size
(23.3 2.1 km
2
,n=14) was smaller than during the Distur-
bance (33.7 3.7 km
2
,n=9; b
disturbance
=10.5; 90% CI 3.9,
17.0) and Recovery periods (32.1 2.2 km
2
,n=19;
b
recovery
=8.9; 90% CI 3.7, 14.1; Fig. 1) with no effect of
EVI (Table 2). Effect of study period on female core area size
was not informative (b
disturbance
=0.7; 90% CI 1.3, 2.7;
b
recovery
=0.6; 90% CI 1.1, 2.2; Fig. 1, Table 2). Inter-indi-
vidual female–female overlap was higher among related
females (0.133 0.089, n=5) than among females of
unknown relatedness (0.046 0.013, n=26; b
related
=0.088;
90% CI 0.010, 0.165), with no effect of study period or EVI
(Fig. 2, Table 3). Neither study period nor EVI affected inter-
individual female–male overlap (Table 3). Intra-individual
female UDOI in the Disturbance period (0.982 0.035,
n=9) was higher than during the Recovery period
(0.645 0.086, n=14; b
recovery
=0.273; 90% CI 0.466,
0.084), and the Stabilization periods (0.722 0.060, n=8;
b
stabilization
=0.206; 90% CI 0.428, 0.017; Fig. 3), the latter
being not informative, with no effect of EVI (Table 4).
Male spacing dynamics
There was no effect of study period (population density) on
any of the spatial measurement in males (Figs 1-3, Tables 2-
4), although inter-individual male overlap seemed to decrease
over time (Fig. 2). There was no informative effect of EVI on
male home range size (b
EVI
=0.04; 90% CI 0.002, 0.008;
Table 2). Male core area increased with EVI (b
EVI
=0.01;
90% CI 0.0009, 0.02; Table 2). Inter-individual male–male
overlap decreased with EVI (b
EVI
=0.00008; 90% CI
0.0001, 0.00006; Table 3). There was no informative effect
of EVI on intra-individual male UDOI (b
EVI
=0.0006; 90%
CI 0.0010, 0.00005; Table 4).
Discussion
Adult female and male leopards in Phinda showed different
spacing dynamics in response to changes in population density.
As harvest declined and leopard density increased, females
contracted home range size and appeared to form matrilineal
kin clusters. This was reflected by greater overlap among
related females compared to unrelated females, and decreasing
inter-annual fidelity in home range use as females relinquished
a portion of their home range to philopatric daughters. In con-
trast, male leopards did not track the contraction of female
home ranges. Males maintained large home ranges, and
overlapped less among each other. The observed dynamics in
spacing were consistent with dispersal-regulated sex-specific
4Journal of Zoology  (2016) – ª2016 The Zoological Society of London
Leopard spatial organization dynamics J. Fattebert et al.

strategies, and did not support a classical dual reproductive
strategy.
The overall spacing of adult leopards in Phinda was typical
of polygynous, solitary carnivores in which larger male home
ranges overlapped with several smaller female home ranges to
increase mating opportunities (Sandell, 1989). We also found
limited intra-sexual overlap among adult leopards of both sexes
under all levels of anthropogenic disturbance, indicative of
mutual avoidance and territoriality in both males and females
(Marker & Dickman, 2005). In both sexes, lack of seasonal
difference in home range size, and high fidelity in inter-annual
home range use reflected the value of local knowledge and
defense of resources (‘resident fitness hypothesis’: Anderson,
1989).
Over time, the observed changes in leopard population den-
sity revealed subtle sex-specific dynamics in spacing of adult
individuals. As population density increased, female home
range size decreased independent of habitat productivity. When
given the opportunity, female leopards seem to use larger
home ranges than required for purely energetic needs (Grigione
et al., 2002; Jetz et al., 2004), to then apportion some of this
space to their female progeny (Fattebert et al., 2015a). On the
surface, this appears to contradict the rule of area minimization
(Grigione et al., 2002) and stable space-use patterns under a
stable environment (Maletzke et al., 2014). However, female
core area size did not change over the course of the study,
consistent with territoriality and exclusive use of resources at
the core area level (Benson et al., 2006).
We link such female spatial behavior to obligate female
philopatry documented in this leopard population over the
course of the study (Fattebert et al., 2015a). In a perturbed
system, filling the voids created by the removal of conspecifics
is a potentially profitable tactic for females to secure sufficient
space to later relinquish to philopatric daughters (Goodrich
et al., 2010). This led to the formation of matrilineal kin clusters,
Figure 1 Empirical mean (SE) annual home range (95% kernel) and core area (50% kernel) size of adult female and male leopards in Phinda
GR, 2002–2012. Temporal segments of the study were defined based on demographic parameters and estimated population density:
Disturbance (2002–2004), Recovery (2005–2008), and Stabilization (2009–2012).
Table 2 Selection of mixed-effect regression models exploring the
effect of study period and vegetative productivity (EVI) on home
range and core area size in adult female and male leopards in Phinda
Game Reserve, South Africa, 2002–2012. Temporal periods of the
study were defined based on demographic parameters and estimated
population density: Disturbance (2002–2004), Recovery (2005–2008),
and Stabilization (2009–2012)
Model Model parameters AICc ΔAICc w
Female home range Period 313.05 0 0.69
Period +EVI 315.67 2.62 0.19
null 317.00 3.94 0.10
EVI 319.07 6.01 0.03
Male home range null 250.15 0 0.53
EVI 250.76 0.61 0.39
Period 254.14 3.99 0.07
Period +EVI 257.04 6.90 0.02
Female core area null 217.41 0 0.52
Period 218.80 1.39 0.26
EVI 219.85 2.44 0.15
Period +EVI 221.50 4.09 0.07
Male core area EVI 186.18 0 0.49
null 186.32 0.14 0.46
Period 191.39 5.20 0.04
Period +EVI 192.79 6.61 0.02
AICc, Akaike Information Criteria adjusted for small sample sizes;
ΔAICc, (AICc) –(AICc)
min
; w, Akaike weight.
Journal of Zoology  (2016) – ª2016 The Zoological Society of London 5
J. Fattebert et al. Leopard spatial organization dynamics

as evidenced by higher overlap among related adult females
compared to females with unknown relatedness. Philopatric
individuals should benefit from local knowledge of resources,
in accordance with resident fitness (Anderson, 1989). Females
are more likely to tolerate each other and benefit from inclu-
sive fitness, and should face the costs of dispersal only when
the costs of resource competition in the natal range are higher
(Lambin et al., 2001). Female dispersal might therefore only
occur under long-lasting stability of both elevated population
density, and socio-spatial organization (Fattebert et al., 2015a).
Spatial tactics of female leopards in Phinda were congruent
with the mounting evidence that matrilines spatially structure
the female segment of the population in other polygynous
mammals (brown bear Ursus arctos: Støen et al., 2005; tiger:
Goodrich et al., 2010; wild boar Sus scrofa: Podg
orski, Scan-
dura & Je
zdrzejewska, 2014).
Male leopards in Phinda did not follow the contraction of
female home ranges over time, as would be expected if male
home range size had been set by female availability (Grigione
et al., 2002). Under increased population density, males
appeared to maintain large home ranges and decrease inter-
individual overlap, although not significantly, in order to maxi-
mize mating opportunities, while limiting access of other males
to these same females (Clutton-Brock & Harvey, 1978). It
would be easier for males to gain exclusive access to fewer
females in smaller home ranges, as large home ranges are
energetically costly to maintain (Jetz et al., 2004). However,
males did not retain exclusive access to all females, overlap-
ping only partially with some female home ranges (Fig. 2,
Table S3). Although complete overlap could ensure exclusive
access, it would yield mating opportunities with fewer female
individuals (Logan & Sweanor, 2001). Such partial inter-sexual
overlap is congruent with the promiscuous mating behavior of
female leopards to possibly confuse paternity as a strategy to
reduce the risk of infanticide (Balme & Hunter, 2013). Male
spacing stability also paralleled elevated subadult male emigra-
tion probability, and longer dispersal distances reported else-
where (Fattebert et al., 2015a).
Adult leopards followed a dual reproductive strategy, with
females and males showing different spatial dynamics under
increasing population density, and generally stable resource
availability. While the male segment of this leopard population
was the most impacted demographically by high human-
mediated mortality (Balme et al., 2009), female spacing
Table 3 Selection of mixed-effect regression models exploring the effect of study period and vegetative productivity (EVI) on adult leopard inter-
individual utilization distribution overlap index (UDOI) in Phinda Game Reserve, South Africa, 2002–2012. Temporal segments of the study were
defined based on demographic parameters and estimated population density: Disturbance (2002–2004), Recovery (2005–2008), and Stabilization
(2009–2012). For female intra-sexual overlap, the effect of individuals’ relatedness was also tested
Model Model parameters AICc ΔAICc w
Female–Female UDOI Relatedness
a
50.09 0 0.43
null 49.19 0.9 0.27
EVI +Relatedness 47.56 2.53 0.12
EVI 46.55 3.54 0.07
Period +Relatedness 45.89 4.2 0.05
Period +EVI +Relatedness 44.11 5.98 0.02
Period 43.94 6.15 0.02
Period +EVI 40.86 9.23 0.00
Female–Male UDOI null 16.33 0 0.68
EVI 14.02 2.30 0.22
Period 12.07 4.25 0.08
Period +EVI 9.62 6.71 0.02
Male–Male UDOI EVI 37.49 0 0.93
null 32.28 5.31 0.07
Period 20.79 16.70 0.00
Period +EVI 19.48 18.01 0.00
a
Mothers and adult daughters or female siblings were defined as related, otherwise unrelated.
AICc, Akaike Information Criteria adjusted for small sample sizes; ΔAICc, (AICc) –(AICc)
min
; w, Akaike weight.
Table 4 Selection of mixed-effect regression models exploring the
effect of study period and vegetative productivity (EVI) on intra-
individual utilization distribution overlap index (UDOI) in adult female
and male leopards in Phinda Game Reserve, South Africa, 2002–
2012. Temporal segments of the study were defined based on
demographic parameters and estimated population density:
Disturbance (2002–2004), Recovery (2005–2008), and Stabilization
(2009–2012)
Model Model parameters AICc ΔAICc w
Female UDOI Period 4.73 0 0.55
null 6.82 2.10 0.19
EVI 7.43 2.70 0.14
Period +EVI 7.79 3.07 0.12
Male UDOI null 10.02 0 0.70
EVI 11.76 1.74 0.29
Period 18.96 8.94 0.01
Period +EVI 24.62 14.6 0.00
AICc, Akaike Information Criteria adjusted for small sample sizes;
ΔAICc, (AICc) –(AICc)
min
; w, Akaike weight.
6Journal of Zoology  (2016) – ª2016 The Zoological Society of London
Leopard spatial organization dynamics J. Fattebert et al.

dynamically adjusted to population density following release
from harvest. These patterns do not support the classical pre-
dictions in response to changes in population density observed
in cougars where spacing patterns are stable in females, and
density-dependent in males (Logan & Sweanor, 2001; Malet-
zke et al., 2014). Spacing dynamics in this leopard population
Figure 2 Empirical mean (SE) inter-individual utilization distribution overlap index (UDOI) between adult female and male leopards in Phinda
GR, 2002–2012. Temporal segments of the study were defined based on demographic parameters and estimated population density:
Disturbance (2002–2004), Recovery (2005–2008), and Stabilization (2009–2012). FF, female–female; FM, female–male; MM, male–male. Note
that during the Disturbance period only one overlapping male dyad was recorded.
Figure 3 Empirical mean (SE) intra-individual utilization distribution overlap index (UDOI) as a measure of fidelity of home range use in adult
female and male leopards in Phinda GR, 2002–2012. Temporal segments of the study were defined based on demographic parameters and
estimated population density: Disturbance (2002–2004), Recovery (2005–2008), and Stabilization (2009–2012).
Journal of Zoology  (2016) – ª2016 The Zoological Society of London 7
J. Fattebert et al. Leopard spatial organization dynamics

appeared to be regulated by dispersal patterns, that is resident
fitness through philopatry in females, and mate competition
and emigration in males (Fattebert et al., 2015a).
Alternatively, an increase in population density led to a reduc-
tion in home range size in both males and females in bobcats
(Benson et al., 2006), and Eurasian lynx (Pesenti & Zimmer-
mann, 2013). The observed lack of change in male leopard home
range size might indicate that the male segment has not yet
reached a threshold of inverse density-dependence. This could
highlight possible lag in resilience of spatial patterns, as docu-
mented in a Eurasian lynx population, where social organization
pattern returned to the pre-disturbance after a delay of 3 years
(Breitenmoser-W€
ursten et al., 2007). These various responses to
disturbance and changes in population density by solitary felids
suggests complexity of resilience mechanisms, and caution
should be exercised in extrapolating results across species, and
even across populations of the same species.
We demonstrated that release of anthropogenic harvest pres-
sure had different consequences on the socio-spatial organiza-
tion of female and male leopards. While male spacing
remained stable, female spacing continually decreased for
7 years following release from harvest. Despite the potential
for rapid numerical response to improved protection (Balme
et al., 2009), demographic perturbations in species like leop-
ards appears to have long-lasting effects on socio-spatial pat-
terns. Our study suggests possible hidden lag effects of harvest
disturbance that likely affect the broader population dynamics
beyond locally impacted populations (Fattebert et al., 2015a;
Pitman et al., 2015). Caution should therefore be exercised
when assessing impact of conservation measures as such lag
effects are not apparent when only assessing demographic
recovery.
Acknowledgements
The study was funded by Panthera, Albert and Didy Hartog,
the Stichting Timbo Foundation, the National Research Foun-
dation (FA 2004050400038 to GB), and the University of
KwaZulu-Natal Gay Langmuir Bursary (to JF). We thank
&Beyond, EKZNW and the neighboring land owners for
allowing us to conduct research on their reserves and are grate-
ful to everyone that assisted with fieldwork, particularly V.
Mitchell, K. Pretorius, J. Mattheus and S. Naylor. Suggestions
from Lourens Swanepoel and Craig Tambling helped improve
the paper significantly.
References
Anderson, P.K. (1989). Dispersal in rodents: a resident fitness
hypothesis. Utah: American Society of Mammalogists Provo.
Arnold, T.W. (2010). Uninformative parameters and model
selection using Akaike’s information criterion. J. Wildl. Mgmt.
74, 1175–1178.
Balme, G.A. & Hunter, L.T.B. (2013). Why leopards commit
infanticide. Anim. Behav. 86, 791–799.
Balme, G., Hunter, L. & Slotow, R. (2007). Feeding habitat
selection by hunting leopards Panthera pardus in a woodland
savanna: prey catchability versus abundance. Anim. Behav. 74,
589–598.
Balme, G., Slotow, R. & Hunter, L.T.B. (2009). Impact of
conservation interventions on the dynamics and persistence of
a persecuted leopard (Panthera pardus) population. Biol.
Conserv. 142, 2681–2690.
Balme, G.A., Slotow, R. & Hunter, L.T.B. (2010). Edge effects
and the impact of non-protected areas in carnivore
conservation: leopards in the Phinda-Mkhuze Complex, South
Africa. Anim. Conserv. 13, 315–323.
Balme, G.A., Hunter, L. & Braczkowski, A.R. (2012).
Applicability of age-based hunting regulations for African
leopards. PLoS ONE 7, e35209.
Barton, K. (2013) MuMIn: multi-model inference. R package
version 1.9.5. Available at: https://cran.rproject.org/web/pack
ages/MuMIn.
Bates, D., Maechler, M. & Bolker, B. (2013) lme4: Linear
mixed-effects models using S4 classes. R package version
0.999999-2. Available at: https://cran.rproject.org/web/pack
ages/lme4.
Benson, J.F., Chamberlain, M.J. & Leopold, B.D. (2006).
Regulation of space use in a solitary felid: population density
or prey availability? Anim. Behav. 71, 685–693.
Beyer, H.L. (2012) Geospatial modelling environment (Version
0.7.2.0). Available at: http://www.spatialecology.com/gme/.
Bolker, B.M., Brooks, M.E., Clark, C.J., Geange, S.W., Poulsen,
J.R., Stevens, M.H.H. & White, J.-S.S. (2009). Generalized
linear mixed models: a practical guide for ecology and
evolution. Trends Ecol. Evol. 24, 127–135.
Breitenmoser-W€
ursten, C., Zimmermann, F., Stahl, P., Vandel,
J.-M., Molinari-Jobin, A., Molinari, P., Capt, S. &
Breitenmoser, U. (2007). Spatial and social stability of a
Eurasian lynx Lynx lynx population: an assessment of 10 years
of observation in the Jura Mountains. Wildl. Biol. 13, 365–
380.
Burnham, K.P. & Anderson, D.R. (2002). Model selection and
multi-model inference. A Practical Information-Theoretic
Approach. New York: Springer.
Calenge, C. (2006). The package adehabitat for the R software:
a tool for the analysis of space and habitat use by animals.
Ecol. Modell. 197, 516–519.
Carbone, C. & Gittleman, J.L. (2002). A common rule for the
scaling of carnivore density. Science 295, 2273–2276.
Clutton-Brock, T.H. & Harvey, P.H. (1978). Mammals,
resources and reproductive strategies. Nature 273, 191–195.
Dahle, B. & Swenson, J.E. (2003). Home ranges in adult
Scandinavian brown bears (Ursus arctos): effect of mass, sex,
reproductive category, population density and habitat type. J.
Zool. (Lond.) 260, 329–335.
Davidson, Z., Valeix, M., Loveridge, A.J., Madzikanda, H. &
Macdonald, D.W. (2011). Socio-spatial behaviour of an
African lion population following perturbation by sport
hunting. Biol. Conserv. 144, 114–121.
Duncan, C., Nilsen, E.B., Linnell, J.D.C. & Pettorelli, N. (2015).
Life-history attributes and resource dynamics determine
8Journal of Zoology  (2016) – ª2016 The Zoological Society of London
Leopard spatial organization dynamics J. Fattebert et al.

intraspecific home-range sizes in Carnivora. Remote Sens.
Ecol. Cons. 1,39–50.
Fattebert, J., Dickerson, T., Balme, G., Slotow, R. & Hunter, L.
(2013). Long-distance natal dispersal in leopard reveals
potential for a three-country metapopulation. S. Afr. J. Wildl.
Res. 43,61–67.
Fattebert, J., Balme, G., Dickerson, T., Slotow, R. & Hunter, L.
(2015a). Density-dependent natal dispersal patterns in a
leopard population recovering from over-harvest. PLoS ONE
10, e0122355.
Fattebert, J., Robinson, H.S., Balme, G., Slotow, R. & Hunter,
L. (2015b). Structural habitat predicts functional dispersal
habitat of a large carnivore: how leopards change spots. Ecol.
Appl. 25, 1911–1921.
Fieberg, J. & Kochanny, C.O. (2005). Quantifiying home-range
overlap: the importance of the utilization distribution. J. Wildl.
Mgmt. 69, 1346–1359.
Gitzen, R.A., Millspaugh, J.J. & Kernohan, B.J. (2006).
Bandwidth selection for fixed-kernel analysis of animal
utilization distributions. J. Wildl. Mgmt. 70, 1334–1344.
Goodrich, J.M., Miquelle, D.G., Smirnov, E.N., Kerley, L.L.,
Quigley, H.B. & Hornocker, M.G. (2010). Spatial structure of
Amur (Siberian) tigers (Panthera tigris altaica) on Sikhote-
Alin Biosphere Zapovednik. Russia. J. Mammal. 91, 737–748.
Greenwood, P.J. (1980). Mating systems, philopatry and
dispersal in birds and mammals. Anim. Behav. 28, 1140–1162.
Grigione, M.M., Beier, P., Hopkins, R.A., Neal, D., Padley,
W.D., Schonewald, C.M. & Johnson, M.L. (2002). Ecological
and allometric determinants of home-range size for mountain
lions (Puma concolor). Anim. Conserv. 5, 317–324.
Hayne, D.W. (1949). Calculation of size of home range. J.
Mammal. 30,1–18.
Hayward, M.W., O’Brien, J. & Kerley, G.I.H. (2007). Carrying
capacity of large African predators: Predictions and tests. Biol.
Conserv. 139, 219–229.
Hayward, M.W., Hayward, G.J., Druce, D.J. & Kerley, G.I.H.
(2009). Do fences constrain predator movements on an
evolutionary scale? Home range, food intake and movement
patterns of large predators reintroduced to Addo Elephant
National Park, South Africa. Biodivers. Conserv. 18, 887–
904.
Jetz, W., Carbone, C., Fulford, J. & Brown, J.H. (2004). The
scaling of animal space use. Science 306, 266–268.
Keehner, J.R., Wielgus, R.B., Maletzke, B.T. & Swanson, M.E.
(2015). Effects of male targeted harvest regime on sexual
segregation in mountain lion. Biol. Conserv. 192,42–47.
Kernohan, B.J., Gitzen, R.A. & Millspaugh, J.J. (2001).
Analysis of animal space use and movements. In Radio
tracking and animal populations: 125–166. Millspaugh, J.J. &
Marzluff, J.M. (Eds). San Diego: Academic Press.
\Lambin, X., Aars, J. & Piertney, S.B. (2001). Dispersal,
intraspecific competition, kin competition and kin facilitation:
a review of the empirical evidence. In Dispersal: Individual,
Population, and Community: 110–122. Clobert, E.D.,
Dhondt, A.A. & Nichols, J.D. (Eds). Oxford: Oxford
University Press.
Lewis, J.S., Rachlow, J.L., Garton, E.O. & Vierling, L.A.
(2007). Effects of habitat on GPS collar performance: using
data screening to reduce location error. J. Appl. Ecol. 44,
663–671.
Logan, K.A. & Sweanor, L.L. (2001). Desert puma:
evolutionary ecology and conservation of an enduring
carnivore. Washington: Island Press.
Maletzke, B.T., Wielgus, R., Koehler, G.M., Swanson, M.,
Cooley, H. & Alldredge, J.R. (2014). Effects of hunting on
cougar social organization. Ecol. Evol. 4, 2178–2185.
Marker, L.L. & Dickman, A.J. (2005). Factors affecting leopard
(Panthera pardus) spatial ecology, with particular reference to
Namibian farmlands. S. Afr. J. Wildl. Res. 35, 105–115.
McLoughlin, P., Ferguson, S. & Messier, F. (2000). Intraspecific
variation in home range overlap with habitat quality: a
comparison among brown bear populations. Evol. Ecol. 14,
39–60.
Mitchell, M.S. & Powell, R.A. (2012). Foraging optimally for
home ranges. J. Mammal. 93, 917–928.
Nilsen, E.B., Herfindal, I. & Linnell, J.D.C. (2005). Can intra-
specific variation in carnivore home-range size be explained
using remote-sensing estimates of environmental productivity?
Ecoscience 12,68–75.
Pesenti, E. & Zimmermann, F. (2013). Density estimations of
the Eurasian lynx (Lynx lynx) in the Swiss Alps. J. Mammal.
94,73–81.
Pitman, R., Swanepoel, L., Hunter, L., Slotow, R. & Balme, G.
(2015). The importance of refugia, ecological traps, and scale for
large carnivore management. Biodivers. Conserv. 24,1975–1987.
Podg
orski, T., Scandura, M. & Je
zdrzejewska, B. (2014). Next of
kin next door –philopatry and socio-genetic population
structure in wild boar. J. Zool. (Lond.) 294, 190–197.
R Core Team. (2013) R: A language and environment for
statistical computing. Vienna: Foundation for Statistical
Computing. Available at: http://www.R-project.org/.
Sandell, M. (1989). The mating tactics and spacing patterns of
solitary carnivores. In Carnivore Behavior, Ecology, and
Evolution: 164–182. Gittleman, J.L. (Ed). Ithaca: Cornell
University Press.
Stander, P.E. (1997). Field age determination of leopards by
tooth wear. Afr. J. Ecol. 35, 156–161.
Støen, O.-G., Bellemain, E., Sæbø, S. & Swenson, J.E. (2005).
Kin-related spatial structure in brown bears Ursus arctos.
Behav. Ecol. Sociobiol. 59, 191–197.
WAMIS (2015) Wide Area Monitoring Information System.
Available at: http://wamis.meraka.org.za/products/long-term-
time-series (Accessed on 28 January 2015).
Woodroffe, R., Donnelly, C.A., Cox, D.R., Bourne, F.J.,
Cheeseman, C.L., Delahay, R.J., Gettinby, G., McInerney, J.P.
& Morrison, W.I. (2006). Effects of culling on badger Meles
meles spatial organization: implications for the control of
bovine tuberculosis. J. Appl. Ecol. 43,1–10.
Journal of Zoology  (2016) – ª2016 The Zoological Society of London 9
J. Fattebert et al. Leopard spatial organization dynamics

Supporting Information
Additional Supporting Information may be found in the online
version of this article:
Figure S1. Leopard density and prey biomass estimates,
Phinda Private Game Reserve, South Africa, 2002–2012.
Table S1. Summary statistics of consecutive seasonal home
range and core area size of adult females and adult male leop-
ards in Phinda Private Game Reserve, South Africa, 2002–
2012.
Table S2. Summary statistics of annual home range size esti-
mated using the minimum convex polygon (MCP) method in
adult female and male leopards, Phinda Private Game Reserve,
South Africa, 2002–2012.
Table S3. Summary statistics of the inter-individual annual
home range percent area overlap in adult female and male
leopards, Phinda Private Game Reserve, South Africa, 2002–
2012.
Table S4. Summary statistics of the intra-individual annual
home range percent area overlap in adult female and male
leopards, Phinda Private Game Reserve, South Africa, 2002–
2012.
Table S5. Annual number of VHF and GPS telemetry reloca-
tions collected from adult leopards sampled in Phinda Game
Reserve, South Africa, 2002–2012.
10 Journal of Zoology  (2016) – ª2016 The Zoological Society of London
Leopard spatial organization dynamics J. Fattebert et al.