Content uploaded by Guy Balme
Author content
All content in this area was uploaded by Guy Balme on Aug 02, 2014
Content may be subject to copyright.
CHAPTER 14
An adaptive management approach
to trophy hunting of leopards
(Panthera pardus): a case study
from KwaZulu-Natal, South Africa
Guy A. Balme, Luke T.B. Hunter, Pete Goodman, Hayden Ferguson,
John Craigie, and Rob Slotow
A radio-collared female
leopard in the Phinda-
Mkhuze ecosystem, South
Africa. #Brett Pearson.
McDonald 14-McDonal-chap14 Revises Proof page 341 22.10.2009 10:52am
In principle, the trophy hunting of large carnivores
has substantial potential to foster their conservation
(Lindsey et al. 2007; Loveridge et al. 2007b). Predators
may be tolerated to a greater extent when they aug-
ment human livelihoods, and hunting may be a
means by which private landowners and local com-
munities can generate revenue from the presence of
carnivores (Leader-Williams and Hutton 2005). If the
profits realized from harvesting a few individuals are
sufficient incentive for people to tolerate the larger
population, the goals of trophy hunting and conser-
vation are compatible. However, poorly managed tro-
phy hunting may produce a variety of deleterious
effects on carnivore populations. Excessive and sus-
tained harvesting of a species can lead to its local
extinction, or reduce numbers to such an extent that
a population is no longer viable in the long term
(Treves and Karanth 2003). Similarly, even when quo-
tas for trophies appear conservative, hunting may con-
stitute additive rather than compensatory mortality,
particularly given that other anthropogenic sources of
mortality are rarely included in the calculation of quo-
tas (Caro et al. 1998). Additionally, uncontrolled tro-
phy hunting may erode the genetic diversity of a
species by consistently selecting the fittest individuals
in a population that would normally survive to propa-
gate the species, particularly adult breeding males and
females (Harris et al. 2002; Coltman et al. 2003). Final-
ly, hunting can alter demographic patterns in carni-
vore populations resulting in impacts on non-hunted
cohorts (Tuyttens and Macdonald 2000). Accordingly,
for trophy hunting to be considered a legitimate tool
in the management and conservation of carnivores, it
is essential that hunting practices avoid detrimental
impacts on populations and are demonstrated to be
biologically sustainable.
The African leopard is one of the most sought-after
big game trophies (Turnbull-Kemp 1967; Grobbelaar
and Masulani 2003). Although resilient in the face of
human pressure, leopards have been eradicated from
vast tracts of their former range due mainly to loss of
habitat, depletion of natural prey, and direct perse-
cution by people (Nowell and Jackson 1996). Leo-
pards once occurred throughout most of the
African continent with the exception of the hyper-
arid interiors of the Sahara and Namib deserts (Fig.
14.1). They are now virtually extinct in North Africa,
have disappeared from most of the West African
coastal belt, and continue to decline outside of pro-
tected areas in large parts of East and southern Africa
(Hunter et al., in press). Ray et al. (2005) estimate that
leopards have been eradicated from 36.7% of their
historic African range. Notwithstanding these recent
declines, leopards are often deemed to warrant low
conservation priority. Their wide geographic range
and ability to persist in regions where other large
carnivores have been extirpated has given rise to a
widespread assumption that their conservation
status is assured. Although listed on appendix I of
the Convention for the International Trade in
Endangered Species (CITES) of flora and fauna, 12
African countries are permitted to export a quota of
leopard skins procured through trophy hunting
(CITES 2007). In 2008, the quota for all states com-
bined was 2648 leopard skins/year, with an addition-
al 10 skins from the Democratic Republic of Congo
and Gabon obtained from sources other than hunt-
ing (Table 14.1).
Trophy hunting is regulated and accounts for a
smaller proportion of leopard deaths than other
human-mediated mortality such as problem animal
control and illegal persecution; for example, an aver-
age of 445 leopards were legally hunted per year in
the 1980s compared to an estimated minimum of
770 killed by other anthropogenic sources in South
Figure 14.1 Current (dark grey) and historical leopard
distribution. Modified from Rayet al. 2005, with permission.
342 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 342 22.10.2009 10:52am
Africa, Botswana, Zimbabwe, and Zambia combined
(Martin and de Meulenaer 1988). Nevertheless, there
is remarkably little scientific input on the allocation
of harvest quotas and the implementation of hunt-
ing practices. Population estimates most widely
quoted by countries proposing to increase or intro-
duce quotas are derived from an overly simplistic
modelling exercise that correlated leopard numbers
with rainfall (Martin and de Meulenaer 1988). This
model was widely criticized for omitting critical
factors such as anthropogenic mortality and prey
availability, and for relying upon questionable as-
sumptions, for example, that leopards occur at maxi-
mum densities in all available habitats. Accordingly,
the final estimate of 714,000 leopards in sub-Saharan
Africa was considered an impossible overestimate
(Jackson 1989; Norton 1990; Bailey 2005). Despite
recent advances in survey techniques that furnish
accurate estimates of leopard numbers at moderate
cost (Henschel 2001; Balme et al., 2009a), few autho-
rities employ these techniques in routine manage-
ment activities such as setting hunting quotas.
Despite the lack of data, the demand for leopard
hunting in Africa appears to be growing. The annual
quotas of four countries (Namibia, South Africa, Tan-
zania, and Mozambique) have increased since 2002,
and Uganda recently introduced trophy hunting of
the species, raising the combined quotas of these
countries from 485 to 1048 (CITES 2007). Further
potential for over-exploitation exists where regula-
tions are abused or ignored. In Tanzania, females
comprised 28.6% of 77 trophies shot between 1995
and 1998, even though only males are legally hunted
there (Spong et al. 2000a). In Zimbabwe, excess skins
from illegal hunts are reportedly smuggled into Zam-
bia and Mozambique to be exported under the quo-
tas of those countries (I. A. Caldwell, personal
communication). As an explicit condition of CITES
certification, African countries that allow leopard
hunting commit to do so in a way that is compatible
with national and sub-national conservation objec-
tives for the species. However, there remains (1) no
rigorous data on the numbers, densities, or popula-
tion trends of leopards anywhere they are hunted;
(2) few empirical data on the impact of hunting on
leopard populations; and, finally, (3) no national,
provincial, or local regulatory framework for harvest-
ing leopards established by an assessment and con-
sideration of numbers 1 and 2.
A variety of management strategies have been pro-
posed to regulate trophy hunting of comparable car-
nivore species that have similar lacunae in
knowledge (Logan and Sweanor 1998; Laundre
˙and
Clark 2003). In lieu of detailed demographic data,
any sensible approach must incorporate an under-
standing of the basic population dynamics of the
target species. In particular, the source–sink meta-
population structure of large felid populations has
been used as a proxy to partition areas into sub-
populations, each with distinct management objec-
tives (Cougar Management Guidelines Working
Group 2005). In this context, source areas are closed
to hunting and function as inviolate refuges where
natural population dynamics can occur not directly
affected by human influence. They serve as robust
biological savings accounts that ensure long-term
population persistence and provide dispersing ani-
mals for numeric and genetic augmentation of
neighbouring sinks and other, more distant sources
(Weaver et al. 1996). Sink areas, in contrast, are those
sub-populations where harvesting is permitted. They
are generally, but not always, less suitable for carni-
vores due mainly to the presence of people, and
concomitant ecological changes such as reduced
Table 14.1 The annual CITES quotas for the 12
African countries permitted to export leopard skins
procured through trophy hunting in 2008.
Country Quota
Botswana 130
Central African Republic 40
Ethiopia 500
Kenya 80
Malawi 50
Mozambique 120
Namibia 250
South Africa 150
Tanzania 500
Zambia 300
Zimbabwe 500
Uganda 28
Trophy hunting of leopards 343
McDonald 14-McDonal-chap14 Revises Proof page 343 22.10.2009 10:52am
prey availability and habitat quality (Cougar Man-
agement Guidelines Working Group 2005).
We adopted a similar approach to develop a novel
protocol for the trophy hunting of leopards in Kwa-
Zulu-Natal (KZN), South Africa. We previously de-
monstrated that poorly regulated harvesting
contributed to the low reproductive output of leo-
pards in the Phinda-Mkhuze ecosystem, one of the
largest leopard populations in the province (Balme
and Hunter 2004; Balme et al., 2009b). In that con-
text, we developed a new protocol for the trophy
hunting of leopards in KZN, intended to be compati-
ble with both the conservation objectives for the
species as defined by the statutory conservation au-
thority, Ezemvelo KwaZulu-Natal Wildlife (EKZNW),
and the desire of private landowners and the hunt-
ing industry to utilize leopards. The chief objectives
of the new hunting protocol were to (1) mitigate the
detrimental effects of leopard hunting in KZN, (2)
enhance both the sustainability and economic pro-
ductivity of leopard hunting in KZN, and (3) ensure
compliance with international and national norms
and standards for leopard hunting (Department of
Environmental Affairs and Tourism, South Africa
2006). In this chapter, we review the historical back-
ground of leopard hunting in KZN, summarize the
biological impacts of hunting on a local leopard
population, and present the new protocol for trophy
hunting the species in the province.
Trophy hunting of leopards in
KwaZulu-Natal
KZN province is situated on the eastern side of South
Africa (26!450–31!100S and 28!450–32!500E;
Fig. 14.1) in the biologically rich transition zone be-
tween the Indian Ocean and the Drakensberg escarp-
ment. Although it is only the sixth largest (92,100
km
2
) of South Africa’s nine provinces, it contains the
most people (estimated at 9.5 million in August 2001;
Statistics South Africa 2005). Altitude ranges from sea
level in the east to over 3450 m above sea level in the
west. Vegetation types vary from tropical dune forests
near the coast to moist lowland and upland grass-
lands, a variety of subtropical forests, and semi-arid
savannas further inland, all of which contain the
megafauna typical of these habitats in Africa, though
many now occur only in protected areas (Goodman
2003). Approximately 58% of the land area in KZN is
used for stock farming (including of wild game), 17%
for crops, 8% for commercial forestry, and roughly 7%
is set aside for conservation (South African National
Biodiversity Institute 2005).
The most suitable leopard habitat in KZN is found in
the mesic northeastern parts of the province on com-
mercial game ranches, cattle farms, and within state-
managed protected areas. Leopardsalso occur on com-
munally owned Zulu lands but generally at lower den-
sities chiefly due to habitat degradation and lack of
natural prey (Balme and Hunter 2004). We refer to
both private and communal land as ‘properties’. Le-
gally, leopards are categorized as ‘specially protected
game’ and fall under the mandate of EKZNWon pub-
lic, private, and communal lands. They are strictly
protected inside provincial public parks (IUCN Pro-
tected Areas Management Category II), and all trophy
hunting of the species currently takes place on private-
ly owned land. Otherwise, leopards cannot legally be
destroyed, unless they represent a threat to life or
property, and then only if the affected party has ap-
plied for and been granted a ‘destruction permit’ by
EKZNW. Despite these restrictions, leopards are killed
opportunistically and illegally on private and commu-
nal lands in the province by livestock owners and
game ranchers (Balme et al., 2009b).
From 1996 onwards, the annual quota of CITES
permits allocated for trophy hunting in KZN has
varied between 5 and 10 leopards (Burgener et al.
2005). Since 2003, it has been stable at 5 animals/
year. By comparison, in 2006, Limpopo Province was
allocated 50 permits, North West Province received
20 permits, and the adjoining Mpumalanga Province
received 10 permits (Burgener et al. 2005). Prior to
2006, CITES permits were allocated in KZN using a
random draw process whereby the first five appli-
cants drawn received a permit and the remaining
applicants were placed sequentially on a waiting
list. If a leopard was not successfully hunted within
2 weeks, the permit defaulted to the next person on
the waiting list, and so on.
For such a widespread species, removing five ani-
mals from the provincial population does not seem
excessive. However, hunting effort was not evenly
distributed throughout KZN. Only 5 of the 22
344 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 344 22.10.2009 10:52am
conservation districts that fall within potential leop-
ard range in the province were awarded CITES per-
mits between 2000 and 2005 (Table 14.2). The
distribution of tags across these five districts was
also uneven, with most (79%) allocated to the neigh-
bouring Nyalazi and Mkhuze districts (Table 14.2),
and specifically to private land surrounding the
Phinda-Mkhuze Game Reserves, which constitute a
single contiguous leopard population. During the
same period, Nyalazi had more than double the
amount of leopards shot on CITES hunts than any
other district (Table 14.2). Sixteen properties in the
province were allocated CITES permits between 2000
and 2005; however, only seven successfully hunted
leopards. The number of leopards shot per property
during this period ranged from 1 to 4 (mean ¼2.29 #
1.21). Additionally, and contrary to national stipula-
tions (Department of Environmental Affairs and
Tourism, South Africa 2006), leopards were frequent-
ly hunted on the same property in consecutive years.
Three privately owned game farms hunted leopards
for 2 successive years and another game farm hunted
leopards for 4 consecutive years. None of these prop-
erties was larger than 25 km
2
, which is slightly less
than the mean home-range size of female leopards
resident in protected areas in the region (mean ¼
27.57 #5.72 km
2
; G.A. Balme, University of
KwaZulu-Natal, unpublished data).
This high offtake contributed to a significant local
effect on the leopard population. The average annual
mortality rate (AMR) of this population between
2002 and 2005 was 0.401 #0.070 (Balme et al.,
2009b), more than double that recorded for the spe-
cies in similar habitat under protection (Bailey 2005).
Male leopards suffered particularly high mortality
(AMR ¼0.538 #0.023), due partly to their higher
vulnerability to trophy hunting (Balme et al., 2009b).
Males are more desirable to trophy hunters because
of their larger size, and males also utilize larger home
ranges and cover greater daily distances than females
(Mizutani and Jewel 1998; Bailey 2005), increasing
their chances of moving into areas where they can be
hunted. In some felid and ursid species, unnaturally
high turnover among males can result in elevated
levels of infanticide (Swenson et al. 1997; Wielgus
and Bunnell 2000; Logan and Sweanor 2001; Whit-
man et al. 2004). Although infanticide is a natural
event in leopards (Ilany 1986; Scott and Scott 2003),
high levels of male turnover promotes a situation in
which females cannot successfully raise cubs because
of constant incursions by new males. The conse-
quences of this in the Phinda-Mkhuze population
included low survival rates among cubs (AMR ¼
0.574 #0.089), delayed age at first parturition
(45.33 #1.76 months), reduced conception rates
(19%), and low annual litter production (0.643 #
0.127 litters/female/year; Balme et al., 2009b). The
combined effects of high mortality and depressed
reproduction led to a negative population growth
rate during the study period (l¼0.978). It is impor-
tant to note that human hunting was not the only
source of leopard mortality and many losses in the
Table 14.2 The number of CITES permits allocated to, and the relative success of, hunts within conservation
districts in KwaZulu-Natal (KZN) province, South Africa, from 2000 to 2005.
Number of permits allocated Successful hunts
Conservation district N%n% Hunt success per district (%)
Dundee 1 2 0— 0
Mdletshe Tribal Authority 1 2 0— 0
Mkhuze 15 32 4 25 27
Nyalazi 22 47 9
a
56 41
Vryheid 8 17 3 19 38
Total 47 16 Mean = 34
a
In addition to this total, an adult female leopard was critically wounded in a hunt in the Nyalazi district in 2003 but not recovered; she is
unlikely to have survived.
Trophy hunting of leopards 345
McDonald 14-McDonal-chap14 Revises Proof page 345 22.10.2009 10:52am
population were the result of illegal persecution rath-
er than legal trophy hunting (Balme et al., 2009b).
Nevertheless, the concentration of trophy hunting
effort on a single population contributed to higher
AMRs than previously documented for the species,
prompting EKZNW to ratify the new protocol to-
wards the end of 2005.
Protocol
Our protocol was based on five key recommenda-
tions intended to achieve objectives 1–3 listed earlier.
These recommendations, and the rationale behind
them, were as follows:
1. Cap the number of CITES permits allocated in
KZN to five each year. When South Africa’s an-
nual CITES quota of 75 leopards was doubled in
October 2004, 10 permits were made available
for trophy hunting in KZN. A population and
habitat viability analysis (PHVA) which evalu-
ated the possible effects of increasing the leop-
ard harvest suggested that the KZN population
had a much greater likelihood of suffering sig-
nificant impacts from hunting than did more
robust populations elsewhere in South Africa
(Daly et al. 2005). This was primarily because
the level of illegal killing of leopards by humans
was estimated among the highest in the coun-
try. According to the PHVA, the risk of extinc-
tion rose from 11% to 62% and the mean
population size declined from an estimated
393 animals to 217 with a quota of 10 leopards
(Fig. 14.2; Daly et al. 2005). Thus, despite the
relatively large estimated population size and
carrying capacity, the removal of just five addi-
tional leopards per year put this population at
substantially greater risk (Daly et al. 2005).
2. Allocate applications for leopard hunts to indi-
vidual properties rather than hunting outfitters.
In the past, any number of hunting outfitters
could apply to hunt leopards on the same prop-
erty, with potential for consecutive hunts dur-
ing the year and multiple animals being
removed from a single farm in a year. Under
the new protocol, landowners applied for a
hunt to take place on a particular property
(this applies whether the land is privately or
communally owned). If successful, the proper-
ty, not the property owner or hunting outfitter,
was allocated the leopard hunt. It then fell to
the property owner to negotiate the best deal
from an outfitter and inform EKZNW of the
name of outfitter and timing of the proposed
hunt so that the CITES permit could be issued.
3. Ensure a more even distribution of CITES per-
mits across the province. As discussed earlier,
CITES permits were formerly allocated uneven-
ly across the province where it was both feasible
to hunt leopards and where landowners sought
to host hunts. To distribute hunting opportu-
nities more evenly, we demarcated five leopard-
hunting zones (LHZ), each with a population
considered extensive and robust enough to sus-
tain hunting (Fig. 14.3). Each LHZ covered an
area of at least 600 km
2
and lay adjacent to a
protected area of at least 290 km
2
in which
hunting was prohibited. Based on density esti-
mates from mark–recapture statistics applied to
camera-trapping we conducted in the main
land use types in which leopards occur in KZN
(11.11 leopards/100 km
2
in protected areas,
7.17 leopards/100 km
2
in game ranches, and
2.49 leopards/100 km
2
in non-protected areas;
Balme et al., in review), we estimated that each
LHZ contained at least 25 adults and was adja-
cent to a protected source population with at
least 32 adults. Each LHZ qualified for a single
permit (i.e. only one animal can be hunted each
year) and had its own individual draw and wait-
ing list. A CITES permit allocated to one LHZ
could not be used in another LHZ, regardless of
demand, or of the success of hunts. If a leopard
was not shot on the first allocated property
within 1 month (rather than the previously
stipulated 2 weeks), the permit was moved to
the next property on the waiting list of that
particular LHZ, and so on.
4. Link the likelihood ofobtaining a tag to the size of
the property on which the hunt will take place.
Properties within an LHZ were still drawn at ran-
dom; however, the chance of a property being
selected was weighted according to its size. This
was based on the assumption that, in the same
area, larger properties are likely to sustain more
346 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 346 22.10.2009 10:52am
leopards. Therefore, for every 1.0 km
2
of land per
property, that property qualified for a single ‘tick-
et’ in the draw for that LHZ. Thus, a 5 km
2
proper-
ty was eligible for 5 tickets, a 25 km
2
property was
eligible for 25 tickets, and so on. In any one draw,
the probability of any property being selected
increased with its size. Once a property had been
allocated a CITES tag, all of its tickets were
removed before the next draw which established
the first property on the waitinglist, and so on.
5. Restrict the trophy hunting of leopards in KZN
to adult males. South Africa is one of the few
countries that have historically allowed the
hunting of female leopards. The PHVA demon-
strated that harvesting only males resulted in
lower extinction rates for all provinces but was
most beneficial in KZN, where the risk of extinc-
tion decreased by almost half (Daly et al. 2005).
Therefore, it was proposed that only adult male
leopards over 3 years old should be legally
hunted in KZN. At this age, male leopards are
easily distinguishable from females; the majori-
ty of 3-year-old males weigh over 60 kg (mean ¼
64.50 #3.7 kg and n¼10; G.A. Balme and L.T.
B. Hunter, Panthera, unpublished data) which
is considerably heavier than females (mean ¼
35 #3.5 kg and n¼7) and they show numerous
obvious physical differences (Fig. 14.4). All
hunters taking leopards in South Africa are re-
quired to be accompanied by a professional
hunter (PH) who is responsible for reporting
the hunt to EKZNW. An EKZNW officer is re-
quired to inspect the trophy within 24 h of the
hunt and, provided all legal stipulations are
met, issue a CITES tag so that the trophy can
be exported. It is the responsibility of the PH
conducting the hunt to judge whether the ani-
mal to be shot is an adult male. A printed guide
is provided by EKZNW (Balme and Hunter, in
press) to PHs and hunting clients to assist them
in sexing and aging animals so that the likeli-
hood of shooting females is reduced.
Discussion
This protocol represents the first rigorous effort to
ensure the trophy hunting of leopards in South
Africa is sustainable. Prior to this, and presently else-
where in South Africa, CITES permits were allocated
on a relatively ad hoc basis that resulted in a concen-
tration of hunting effort on a single leopard popula-
tion. Our protocol addressed this problem by
distributing hunts over a much larger area and is
based on the meta-population management ap-
proach used to manage puma (Puma concolor) and
grizzly bear (Ursus arctos horribilis) populations in
North America (Servheen et al. 1999; Laundre
˙and
Clark 2003). All LHZs in KZN adjoin large protected
areas that potentially act as source populations to
replace animals shot on surrounding farms and
game ranches. The implicit assumption is that each
LHZ is capable of sustaining the removal of one leop-
ard a year to trophy hunting without threatening the
demographic or genetic viability of that sub-popula-
tion or the larger regional population. In other parts
of South Africa, source populations may need to be
identified that do not comprise formally protected
land. In this context, there is considerable scope for
the private sector to contribute; for example, by the
Annual CITES quota
0
0.0
0.2
PE (100 years)
0.4
0.6
0.8
1.0
90 (6) 120 (8) 150 (10)
135 (9)105 (7)75 (5)
Waterberg Free State
KZN E Cape Mountains
E Cape Valleys Wild Coast
Figure 14.2 Effects of CITES quota on probability of
extinction of leopard populations in South Africa (from Daly
et al., 2005, used with permission). KwaZulu-Natal quota in
parentheses.
Trophy hunting of leopards 347
McDonald 14-McDonal-chap14 Revises Proof page 347 22.10.2009 10:52am
consolidation of commercial game ranches and pri-
vate reserves that do not rely on consumptive utili-
zation of leopards into ‘conservancies’, in which
multiple, private landowners develop agreements to
manage and conserve their combined areas as single
unit (the ‘conservancy model’; Hunter et al. 2003).
However, it is critical that source areas are large
enough that losses in adjacent sinks do not have
detrimental effects on the source population. Deter-
mining the size of a source area depends on various
factors such as prey density, habitat quality, and har-
vest rates in neighbouring sink populations. Beier
(1993) and Logan and Sweanor (2001) suggested
that a minimum size of 1000–2200 km
2
was needed
to sustain a viable puma population for 100 years. In
KZN, only two protected areas (Hluhluwe-Imfolozi
Game Reserve and the Mkhuze/Mun-Ya-Wana/Ma-
khasa/St Lucia Complex) adjoining a LHZ were this
size, but leopards reached higher densities in these
areas than pumas in the above examples and all sub-
populations were likely large enough to withstand
the removal of a single animal by hunting.
Although sink areas in our protocol comprised
non-protected farm land, commercial game ranches,
and communal lands, leopard population growth
rate need not be negative and, if the management
objective is to maintain hunting, harvest levels must
not exceed immigration rates (Cougar Management
Guidelines Working Group 2005). Indeed, provided
other anthropogenic sources of leopard mortality are
curtailed, populations occupying game-ranching
areas have the potential for positive growth given
the widespread availability of natural prey and habi-
tat (Lindsey et al. 2005). This may benefit ranch
EKZNW protected areas
South Africa
Commerical game ranches
Leopard hunting zones
KZN provincial boundary
Figure 14.3 The location of
designated leopard-hunting zones
(LHZ) in relation to commercial game
ranches and Ezemvelo KwaZulu-Natal
Wildlife (EKZNW) protected areas and
conservation districts in KwaZulu-Natal
(KZN) Province, South Africa.
348 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 348 22.10.2009 10:52am
owners seeking to profit from hunting leopards but
will only be realized when removal of leopards for
other reasons (primarily by illegal shooting and
trapping in retribution for real and perceived losses
of livestock) are reduced. Most importantly, we do
not accept that a sink is an area where uncontrolled
removal of leopards is permissible. Private land-
owners who hunt leopards and benefit financially
from their presence must foster populations on
their lands if both hunting and leopards are to persist
in those areas. This may be assisted by our recom-
mendation of linking leopard hunting to the size of
properties. As well as rewarding larger landowners
with an increased opportunity of hosting a hunt,
this may encourage smaller landowners with an in-
terest in hunting leopards to increase the size of their
land or consolidate neighbouring properties, and
therefore foster the growth of larger contiguous
areas of suitable habitat for leopards.
While we believe we have been conservative in
planning LHZs, we have not addressed a prevailing
problem anywhere that leopard quotas are assigned,
which is the quantification of additional leopard
deaths due to people. Caro et al. (1998) warned that
high levels of traditional hunting by resident com-
munities and illegal poaching probably resulted in
unsustainable trophy hunting quotas for leopards in
Tanzania, though that country subsequently dou-
bled its quota to 500 in 2002. In KZN and South
Africa generally, leopards are illegally killed primarily
by pastoralists and farmers, though there is some
evidence for organized poaching; for example, skins
of at least 58 individuals were seized in the Mkhuze
district in July 2004, apparently destined for interna-
tional markets (Hunter et al., in press). During a
6-year period, 52.6% of all anthropogenic leopard
mortality (n¼19) recorded in the Nyalazi district
was illegal (Balme et al., 2009b). This is undoubtedly
a minimum estimate given that most illegal killings
are concealed and even accidental deaths of leopards
such as road deaths are rarely reported. The ongoing
sustainability of leopard hunting in each LHZ will
depend on increased quantification and mitigation
of the numbers of individuals killed illegally.
In addition to illegal persecution, significant num-
bers of leopards are legally destroyed every year
through problem animal control usually undertaken
by the landowner. In principle, landowners wishing
to destroy a damage-causing leopard must demon-
strate that a particular individual is preying upon
livestock or creating some other form of damage
and that steps have been taken to bring about a
non-lethal solution to the problem (Ferguson
2006). In practice, destruction permits are regularly
awarded on the basis of little evidence and there is no
mechanism to ensure that landowners remove the
individual responsible for the damage (G.A. Balme
and L.T.B. Hunter, unpublished data). Although re-
cent policy changes by EKZNW attempt to address
this (Ferguson 2006), the number of leopards
killed as problem animals (3.42 #1.21; EKZNW, un-
published data) generally exceeds the numbers
hunted legally every year and often occurs on
the same properties as trophy hunting. Ideally,
(a)
(b)
Figure 14.4 Extract from a guide for PHs to assist in the
accurate identification of suitable leopards for hunting
(G.A. Balme and L.T.B. Hunter, Panthera, unpublished
report): a) adult female leopard highlighting narrower
chest area and proportionally smaller head, and b) adult
( > 3-years-old) male leopard showing developed neck
region with prominent dewlap and much larger head.
Trophy hunting of leopards 349
McDonald 14-McDonal-chap14 Revises Proof page 349 22.10.2009 10:52am
damage-causing leopards could be hunted under the
CITES quota (Daly et al. 2005) but this is not a viable
option in KZN. Trophy hunts are planned months
and sometimes years in advance while stock-killing
incidents are largely unpredictable. With only five
hunts taking place during the course of a year, it is
unlikely that the two events will occur concurrently.
Additionally, for this to work, the allocation of CITES
permits would have to be linked to complaints rather
than by the existing lottery process. Such a system
might foster incentives for false claims about dam-
age-causing leopards to increase the chances of re-
ceiving a CITES tag. Similarly, landowners with a
genuine problem of losing livestock would be re-
warded, potentially promoting lax husbandry. It is
possible that this idea may be more applicable else-
where such as in the Limpopo Province where there
is both a larger quota and greater conflict between
leopards and farmers (Daly et al. 2005). In Namibia, a
‘Hunters’ Hotline’ was established to link farmers
that were losing livestock to leopards with the Pro-
fessional Hunters Association (Stein 2008). While
this partially overcame the problem of sourcing cli-
ents available at short notice, it did not encourage
farmers to improve their management techniques.
Furthermore, a questionnaire survey revealed that
only 12% of farmers were willing to use the hotline
(Stein 2008). Most farmers thought it was easier and
more beneficial to destroy the problem leopard
themselves. In Namibia, a destruction permit can
be obtained after a damage-causing animal has
been killed.
Our protocol was not created to disadvantage hun-
ters or prevent the consumptive utilization of leo-
pards. In fact, the new system may improve the
overall economic productivity of hunting as more
professional hunters and landowners stand to benefit
from the limited CITES permits available. Previously,
the industry was monopolized by a small number of
outfitters operating in the Nyalazi and Mkhuze dis-
tricts. By distributing permits evenly across the prov-
ince, it gives a larger number of outfitters the
opportunity of hosting a leopard hunt. Additionally,
extending the validity of a permit from 2 weeks to 1
month greatly improves the chance of any individual
hunt being successful without increasing the harvest
quota. Individual outfitters therefore stand to gain as
their clients are typically charged a daily rate and there
is now potential for longer visits. Similarly, there is
increased potential for more property owners to
benefit financially over time from having leopards
on their land. The prospect of receiving revenue from
a trophy hunt every few years may mitigate damage
caused by leopards and encourage landowners to
allow the continued presence of leopards on their
farms where formerly they stood no chance to host a
hunt. Thirty-nine percent of landowners interviewed
in the Nyalazi district (n¼18) supported the new
protocol, 11% felt that the allocation of CITES permits
should be linked to the control of damage-causing
leopards, 11% preferred the old process, and 39%
showed no interest in hunting leopards (Balmeet al.,
2009b). More widely, if applied correctly, this frame-
work may ultimately permit increased quotas and
hunting in more areas of the province. If landowners
benefiting from hunts reduce the persecution of leo-
pards and foster population growth on their lands, the
potential exists to increase quotas and expand or add
LHZs to new areas.
Under our protocol, the potential also exists for a
wider cross-section of the community to benefit from
leopard hunts. Historically, leopard hunting has been
restricted largely to private farms and game ranches
and has rarely taken place on communally owned
Zulu land. Hunting outfitters had little incentive to
develop wider relationships as they were guaranteed
access to a small number of private landowners mostly
within the Nyalazi and Mkhuze districts. Under the
new system, outfitters who negotiate with local tribal
authorities (TAs) to conduct hunts will increase their
chances of securing access because communal lands
are generally much larger than individual private
holdings and therefore stand a greater likelihood of
being drawn in the lottery. At least two TAs, the
Mdletshe TA and the Mthembeni TA, already have a
system in place for such negotiations and received a
CITES permit for a leopard hunt in 2001 and 2007,
respectively. Other communities have developed
structures for hunting other trophy species that
could easily be extended to include leopards. For ex-
ample, the Makhasa TA hosted a black rhino (Diceros
bicornis) hunt in 2006 that generated almost US
$11,000 for the community (EKZNW, unpublished
data). Communal areas in KZN comprise approxi-
mately 30,000 km
2
, much of it with considerable po-
tential to be important leopard habitat (McIntoshet al.
350 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 350 22.10.2009 10:52am
1996). This will only likely be realized when commu-
nities are encouraged to change current land-use prac-
tices that limit the viability of the land to harbour
leopards and other wildlife. With its low operating
costs and high profit margin, trophy hunting may
provide one such incentive (Lewis and Jackson 2005).
The implementation of our protocol is only the
start of an adaptive process that will rely upon careful
monitoring of leopard numbers and losses to people.
This currently takes place in only one LHZ as part of a
long-term ecological study on the species (Hunter
et al. 2003; Balme and Hunter 2004; Balme et al.
2009a, Balme 2009b). Our data from this study
suggest the new protocol has been successful in facil-
itating the recovery of the Phinda-Mkhuze leopard
population (Balme et al., 2009b). Since its implemen-
tation in 2006, estimated population growth has
increased by 16% (l¼1.136), annual mortality has
dropped to 0.134 #0.016, and the reproductive
output of the population has improved (Balmeet al.,
2009b). Specifically, mean age at first parturition has
decreased (33.67 #1.85 months), conception rates
have improved (39%), and annual litter production
has increased (0.709 #0.135 litters/female/year;
Balme et al., 2009b). Most notably, all cubs (n¼14)
born to radio-collared females survived to indepen-
dence after the change in hunting protocol. As a result
of lowered harvest rates and increased social stability,
population density increased from 7.17 #1.12
leopards per 100 km
2
in 2005 to 11.21#2.11 leopards
per 100 km
2
in 2009 (Balme et al., 2009b).
Although it is unlikely such intensive monitoring
could be extended to other LHZs, the process may be
assisted by recent improvements in survey meth-
odologies for cryptic species. Camera-trapping is
one such method that furnishes accurate density
estimates for leopards with modest financial re-
sources and relatively little technical expertise
(Henschel and Ray 2003; Balme et al., 2009a). The
monitoring of trophy quality and composition will
also help gauge whether there are any undesirable
demographic shifts within a population (Anderson
and Lindzey 2005). The new reporting requirement
for trophy age and sex will improve monitoring of
trophy quality across all LHZs in KZN. In Botswana
and the Niassa Province in Mozambique, assess-
ments of trophy quality are used to promote sustain-
able leopard hunting (Begg and Begg 2007; Funston,
personal communication). Quotas are assigned inde-
pendently to regions depending on the sex and age
of animals taken in the previous hunting season.
These systems are self-regulating and encourage eco-
logically sound hunting by penalizing hunters that
shoot females or under-age leopards. Although no
similar penalties are currently in place for hunters
in KZN and the hunting of males only (recommen-
dation 5) is simply advised as best practice, more
stringent control is likely in the near future; for ex-
ample, a policy is being drafted that will prevent PHs
receiving a CITES permit for a period of 3 years if they
hunt female or subadult (<3 years old) leopards
(EKZNW, unpublished data). Evaluating and incor-
porating the socio-economic attitudes of human
communities coexisting with leopards will also be
crucial in successfully implementing this protocol.
Aside from anecdotal information (e.g. Hunter et al.
2003), there are no data available on the attitudes of
local communities towards leopards and their man-
agement, nor on the underlying factors influencing
attitudes. Practical solutions exist for conducting
widespread surveys in the region (Lindsey et al.
2005) and should be adopted more widely in parallel
with the implementation of the new protocol in all
LHZs.
Our protocol has potential to act as a standard for
hunting the species in other provinces of South
Africa, and further afield in Africa. Currently, no
other national or regional authority assigns quotas
based on a detailed understanding of the species’
ecology and the potential impacts of hunting as we
have attempted here. Implementing this approach
on a national scale will require that local conserva-
tion authorities clarify leopard population status and
distribution within each province of South Africa. As
a first step, we recommend the delineation of leopard
sub-populations that could act as source or sink areas
utilizing GIS coverages that include habitat and land-
use type, records of leopard presence, patterns of
problem-leopard complaints, historical utilization
patterns of leopards, and the demand for CITES tags
(Cougar Management Guidelines Working Group
2005). Although parameters will vary, this approach
is relevant to most countries that permit sport hunt-
ing of leopards. Similarly, our approach may have
value for other large, wide-ranging felids that are
sport-hunted, for example, cheetahs (Acinonyx
Trophy hunting of leopards 351
McDonald 14-McDonal-chap14 Revises Proof page 351 22.10.2009 10:52am
jubatus) in Namibia which are hunted exclusively on
private land (CITES 1992), pumas in Latin American
populations where sport hunting is growing (Ojasti
1997), and Eurasian lynx (Lynx lynx) in Norway
where quotas are assigned on the basis of depreda-
tion complaints (Andre
`net al. 2006). Except for a
handful of sites where lions have been reintroduced
(Hunter et al. 2007c), the leopard is the apex predator
in many landscapes in KZN and, as such, fulfils a
vital role in functioning ecosystems. We believe the
implementation of this protocol will be a key factor
in maintaining leopards in those landscapes by bal-
ancing the ecological requirements of the species
with the needs of multiple stakeholders.
Acknowledgements
We are grateful to all at EKZNW who helped to de-
velop and implement this protocol, and to the vari-
ous professional hunters, game-ranch managers,
farmers, and landowners who have agreed to adopt
it in the field. We are thankful to CC Africa/&Beyond
and the Wetland Authority for allowing us to con-
duct the research on leopards in their reserves. We
are especially grateful to Kevin Pretorius (CC Africa/
&Beyond) for inviting us to initiate research into the
effects of hunting on leopards in this region, and for
continued logistic and material support. This study
was supported by Albert and Didy Hartog, Panthera,
Christian Sperka and the Wildlife Conservation Soci-
ety. G.A. Balme was supported while writing this
chapter by a Panthera Kaplan Award and received a
NRF bursary from grant #FA2004050400038. Nicole
Williams is thanked for her considerable effort
in proofreading and formatting the manuscript.
We thank David Macdonald, Michael J. Chamber-
lain, Cynthia Jacobson, and five anonymous re-
viewers whose detailed comments improved the
chapter. Our thanks to Brett Pearson for providing
the photograph of Ntombi and to S. Nijhawan
(Panthera) who prepared the maps.
352 Biology and Conservation of Wild Felids
McDonald 14-McDonal-chap14 Revises Proof page 352 22.10.2009 10:52am
