The relative importance of trophy harvest and retaliatory ...
Transcript of The relative importance of trophy harvest and retaliatory ...
The relative importance of trophyharvest and retaliatory killing of large
carnivores: South African leopards as acase study
Lourens H. Swanepoel1, Peter Lindsey3, Michael J. Somers1,4,Wouter Van Hoven1 & Fredrik Dalerum1,2*
1Centre for Wildlife Management, Hatfield Experimental Farm, University of Pretoria, Private Bag X20, Hatfield, 0028 South Africa2Mammal Research Institute, Department of Zoology and Entomology, University of Pretoria,
Private Bag X20, Hatfield, 0028 South Africa3Endangered Wildlife Trust, Private Bag X11, Modderfontein, 1645 South Africa
4Centre for Invasive Biology, University of Pretoria, Private Bag X20, Hatfield, 0028 South AfricaReceived 6 June 2013. Accepted 20 December 2013
There are considerable challenges in the conservation of large carnivores, caused by largearea requirements, low reproduction rates and low population densities coupled with theirtendency to cause conflict with humans. Trophy hunting is one strategy to increase supportfor large carnivore conservation. Leopards, Panthera pardus, rank among the most sought-after trophies in South Africa.However, trophy hunting has been suggested as partly respon-sible for leopard population declines, and leopards are also killed in retaliatory actions. Inthis study we used a stochastic population model to evaluate the relative influences ofretaliatory killing and trophy harvest on leopard population persistence, and to assess thesustainability of the current leopard trophy harvest in South Africa. There was a strongereffect of variation in retaliatory killing than of harvest on population persistence. Althoughwe found low extinction risks for South African leopards within 25 years, high risks ofpopulation declines across a wide range of simulation scenarios call for concern regardingthe viability of the South African leopard population. We suggest that conflict mitigation maybe more effective in promoting leopard persistence than restricting trophy harvest, and thataccurate estimates of retaliatory killing are necessary for assessments of harvestsustainability.
Key words: trophy hunting, carnivore, adaptive management, simulation models, populationviability analysis.
INTRODUCTIONConservation biologists and wildlife managers faceconsiderable challenges in the management andconservation of large carnivores (Noss et al. 1996).Large carnivores have extensive area require-ments, low reproduction rates and normally occurat low densities (Purvis et al. 2000). Furthermore,protected areas are rarely large enough to main-tain viable large carnivore populations (Linnellet al. 2001). Therefore, non-protected areas areimportant in the conservation of large carnivores(Funston et al. 2013). However, carnivore popula-tions in non-protected areas are frequently inconflict with rural communities, commercial farm-ers and game keepers, and retaliatory killing inresponse to human–large carnivore conflict iscommon (Woodroffe & Ginsberg 1998; Dar et al.
2009; Dickman 2010). Such conflict-related killingcan have both direct and indirect demographicconsequences for carnivore populations (Treves& Karanth 2003). For example, legal and illegalretaliatory killing of large carnivores can result inthe removal of mature, reproductively active indi-viduals (Woodroffe & Frank 2005; Baker et al.2008). Retaliatory killing can thus hamper carni-vore population sustainability as well as impedethe recovery of threatened populations (Woodroffe& Frank 2005; Kaczensky et al. 2011).
Commercial exploitation, like trophy hunting, isoften used as an incentive in non-protected areasto increase local tolerance towards large carnivorepopulations (Lewis & Jackson 2005; Treves 2009).Appropriately managed trophy hunting can contrib-ute significantly to local economies, and financialreturns from hunting may in some cases exceedthe cost of predation (Lindsey et al. 2012; Funston
South African Journal of Wildlife Research 44(2): 115–134 (October 2014)
*To whom correspondence should be addressed.E-mail: [email protected]
et al. 2013). Trophy hunting may therefore be apowerful financial tool to increase persistence andpopulation viability of some large carnivores innon-protected areas (Funston et al. 2013). How-ever, the effectiveness of trophy hunting as aviable conservation tool relies on harvest at demo-graphically sustainable levels (Packer et al. 2009).While trophy hunting may not generally pose amajor threat to wildlife populations (Lindsey et al.2007), it has been suggested that poorly managedtrophy hunting quotas can be a contributing factorfor causing declines in populations of some largecarnivores (e.g. Packer et al. 2009; Packer et al.2011). Furthermore, the sustainability of trophyhunting can be adversely affected by retaliatorykilling (Loveridge et al. 2009). It is therefore impor-tant to evaluate the importance of retaliatory killingand trophy hunting simultaneuously when evaluat-ing the sustainablity of harvest levels.
Leopards (Panthera pardus) are widely distrib-uted throughout South Africa (Friedmann & Daly2004) and they are frequently killed in conflict withlivestock or game farmers (Inskip & Zimmerman2009; Thorn et al. 2012). Trophy hunting is oftenseen as an appropiate incentive to conserve leop-ards (Lindsey et al. 2007; Funtson et al. 2013), butconcern over the sustainability of leopard trophyhunting in South Africa has been raised (Balmeet al. 2009). These concerns are founded in adoubling of the harvest quota since 2005 (75 to150 animals), as well as increased incidents oflegal and illegal retaliatory killing (Daly et al.2005).Coupled with recent concerns that poorly man-aged trophy hunting may be partly responsible fordeclining leopard populations (Balme et al. 2009;Packer et al. 2011), there thus appears to be aneed for a formal evaluation of the sustainabilityof the current trophy harvest of South Africanleopards.
Population viability analyses (PVA) are commonlyused to evaluate the effects and importance ofmanagement actions, like trophy hunting, onpopulation persistence (Reed et al. 2002). Theinterpretability of a PVA is, however, dependent onthe quality of underlying population parameters(Dalerum et al. 2008). Since such demographicdata typically are limited for large carnivores,including leopards (Daly et al. 2005), it is oftendifficult to interpret results from PVAs for thesespecies. A potentially effective PVA method forsuch situations is to combine stochastic popula-tion models with logistic regression (McCarthyet al. 1995; McCarthy et al. 1996). This approach
provides a qualitative assessment of the influenceof demographic parameters on extinction risk, andcan be used to estimate risk over a wide range ofpotential demographic and intervention scenarios.It also has the additional advantage that it allowsfor assessments of interactions among parameters(Cross & Beissinger 2001; Dalerum et al. 2008).
In this study we combined binary output from astochastic population model with logistic regressionto evaluate the effects of different scenarios ofharvest and retaliatory killing on leopard popula-tion persistence, with the dual aim of (i) qualita-tively assessing the relative importance of trophyhunting and retaliatory killing for a large carnivorespecies, and (ii) assessing if the current CITESquota is a cause of concern for the South Africanleopard population under realistic ranges of popu-lation sizes and retaliatory killing.
MATERIALS AND METHODS
Model structureRather than trying to provide numerical values
for recommended harvest levels, we used a quali-tative simulation approach where we focused onevaluating the effects of different combinations ofretaliatory killing and harvest scenarios on popula-tion persistence (Starfield 1997; Pe’Er et al. 2013).This approach allowed us to evaluate uncertaintyin retaliatory killing over a range of potential levelsof trophy harvest.
We used a previously described stochasticpopulation model with a simple sex and age struc-ture (Dalerum et al. 2008). The model can bedefined as:
Pt = AMt + AFt + SMt + SFt + Jt (1)
where Pt is population size, AMt and AFt is thenumber of adult males and females, SMt and SFt isthe number of subadult males and females, and Jt
is the number of juveniles at year t, respectively.The model is temporally discrete with time steps ofone year.Following Swanepoel et al. (in press), weregarded animals older than three years to beadults, animals between one and three years to besubadults and animals less than one year old to bejuveniles. Although we capped population size atfour times its initial value, we did not include anydensity dependent effects on demographic param-eters as this theoretical carrying capacity wasapproached. We chose to use a density independ-ent approach partly because data on densitydependent demographic regulation are lacking
116 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
for leopards and similar large felids, and partlybecause we assumed that South African popula-tions are largely limited by a combination of suit-able habitat and persecution, and therefore areunlikely to occur at densities that approach carry-ing capacity defined by resource limitation (seeCaro et al. 2009).
Model parametersWe used published demographic information for
leopards to parameterize the model.A summary ofall model parameters and their sources are givenin Appendix 1. We estimated initial populationstructure to be 30% adult males, 30% adult females,15% subadult males and females respectively,and 10% juveniles. Each time step we drew fecun-dity from Poisson distributions with a mean annuallitter size per female of 1.41 and calculated thetotal number of juveniles in a given year bysummarizing the fecundity of all females. We drewbaseline annual survival probabilities from binomialdistributions with averages reflecting age and sexspecific survival estimated as 91%, 85%, 82%,93% and 33% for adult males and females, sub-adult males and females, and juveniles, respec-tively (Appendix 1). These values were taken froma recent meta-analysis of leopard survival insouthern Africa (Swanepoel et al., in press).Although infanticide has been reported as a majorcause of juvenile mortality in one leopard popula-tion (Balme et al.2013), there are no data availableto quantify the effects of male turnover on such in-fanticide rates (e.g. Swenson et al. 1997). Further-more, infanticidal mortality is included in ourjuvenile survival estimates (Swanepoel et al., inpress). We therefore believe that explicit inclusionof the demographic effects of infanticide in ourmodels will only introduce unquantifiable uncer-tainty, which will hamper the interpretability of thesimulation output. We estimated the number ofemigrants out of each province from binomialemigration probabilities set to 47% for subadultmales, 14% for subadult females and 5% foradults. We calculated the number of immigrantsinto a province as the sum of the number ofemigrants from each neighbouring provincescaled by the proportional length of the land bor-der between each pairs of provinces. For interna-tional borders, we calculated an average numberof immigrants per km of border within South Africafor the respective province, and multiplied thisnumber by the length of the international border.We assumed equal survival for animals migrating
between provinces as for animals remaining withintheir natal province.
We multiplied leopard densities derived fromliterature and unpublished sources by spatiallyexplicit probabilities of occurrences derived from ahabitat suitability model to calculate realisticranges of provincial population sizes (Swanepoelet al. 2013). Source studies used to estimate thesedensities are given in Appendix 2. Each study wasassigned to a major biome, and to incorporateuncertainty in density estimates we calculated aminimum, median and maximum density for eachbiome. We multiplied these densities by the pixelspecific probability of occurrence from our habitatsuitability model to calculate the number of leopardsper pixel, and estimated provincial leopard popula-tion sizes by summarizing the number of animalsfor all pixels in each province (Table 1). Estimatedminimum, median and maximum densities foreach province and biome are given in Appendix 3.
We collected information on leopard harvest andretaliatory killing from provincial nature conserva-tion offices for the period 2000–2010. These datacontained the number of trophy hunted animals,legally killed problem animals (Damage CausingAnimals, DCA) and illegally killed animals. Basedon these data, we estimated the sex ratio of huntedleopards under CITES trophy permits to be 68%males and the age structure to be 95% adults, andthe sex and age structure for retaliatory killing (i.e.estimated from animals killed under DCA permits)to be 65% males and 72% adults (Appendix 1).Due to low compliance to legislation regardingdamage-causing leopards, only a low number ofreports are received for DCA permits issued, aswell as leopards killed under such permits (Chase-Grey 2011). For example, in the Limpopo provincereports were only received for 28% of DCA permitsissued. Furthermore, record-keeping capabilitiesvaried among provinces, with only Limpopo andKwaZulu-Natal having adequate records onDCA permits issued.Because of these two issues,the number of reported animals killed underDCA permits in each province was low. Therefore,to estimate how many leopards were killed underDCA permits we also collected the number ofDCA permits issued for Limpopo and KwaZulu-Natal from 2005 to 2010 (provinces with adequateDCA data). Harvest and retaliatory killing statisticsfor the South African provinces are listed in Appen-dix 4.
We set limits for likely ranges of trophy harvestfrom zero harvest to four times the maximum
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 117
recorded harvest per year in each province. To setan upper limit to harvest for provinces where noharvest had occurred, we first calculated the maxi-mum recorded proportional harvest based on themedian population size from the provinces whereharvest had occurred. This value was then multipliedby the median population size of the respectiveprovince that had been without harvest. We simi-larly set the lower limits for likely ranges of retalia-tory killing at zero killed animals per year in eachprovince. We set the upper limit of likely ranges ofretaliatory killing to the maximum number of issuedDCA permits plus an estimated number of illegallykilled animals. For provinces where we did nothave number of issued permits, we calculatedmaximum proportional annual number of permitsper leopard based on the median population sizesin Limpopo and KwaZulu-Natal (where we didhave information on number of issued permits),and multiplied this number by the median popula-tion sizes for provinces where we did not haveinformation of number of issued permits.
Estimating the number of illegally killed animalswas more challenging because detecting illegalkilling of carnivores is difficult (Kaczensky et al.2011). During 2010 an anonymous amnesty wasgiven to farmers in the Eastern Cape if they gaveinformation on the number of leopards killed ille-gally during the years 2004–2010 (Lindsey et al.2011). We therefore used the amnesty data fromEastern Cape in an attempt to quantify illegalkilling of leopards in other provinces. To do this we
first calculated the proportion of leopards esti-mated from the median population size that wasillegally killed in the Eastern Cape. This proportionwas then multiplied by the median population sizein the other provinces. Likely ranges of harvestand retaliatory killing for each province and forSouth Africa that were used in our simulations aregiven in (Table 1).
Simulations and data analysesWe generated 1000 parameter combinations by
sequentially altering values of harvest, retaliatorykilling and population sizes within ranges describedabove.We ran 100 simulations for each parametercombination, resulting in a total of 100 000 individualsimulations. For each simulation, we capturedpopulation size after 25 years and coded thispopulation size as extinct if fewer than two animalsremained or as having declined if it was belowinitial population size. We used these binaryresponse codes as response variables in general-ized linear models to calculate the probability ofextinction and population declines (McCarthy et al.1995; Cross & Beissinger 2001; Dalerum et al.2008). We used initial population size, trophyharvest, retaliatory killing as well as all interactionterms as predictors. The models were fitted with abinomial error structure and with a logit link function.We have not reported any evaluations of the teststatistics from our logistic models since statisticalinference is dependent on sample size, which insimulated data is arbitrarily determined (e.g.number
118 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
Table 1. The range of leopard population sizes and maximum levels of trophy harvest and retaliatory killing used forthe simulations for each South African province.
Province Population sizea Trophy harvest Retaliatory mortality
Minimum Maximum Maximumb Maximumc
Limpopo 1682 7168 204 348Mpumalanga 338 1851 32 169North West 174 255 40 54Gauteng 25 31 4 8Northern Cape 68 262 14 27Free State 8 26 4 8KwaZulu-Natal 247 1120 20 118Western Cape 200 619 32 65Eastern Cape 71 299 16 33South Africa 2813 11632 366 826
aMinimum and maximum population sizes were calculated by multiplying the lowest and highest recorded density in each major biome byspatially derived explicit probabilities of leopard occurrence derived from a habitat suitability model (Appendix 2, Appendix 3).
bRepresents four times the maximum recorded trophy harvest per year over the period 2002–2010 (Appendix 4), or for provinces withoutexisting harvest the average proportional harvest based on the provinces with existing harvest. Minimum levels were set to zero for eachprovince.
cRepresents the maximum number of damage-causing leopards killed or translocated per year over the period 2002–2010 (Appendix 4) andan estimated number of illegally killed leopards. Minimum levels were set to zero for each province.
of simulation runs; McCarthy et al. 1995; Dalerumet al. 2008). Instead, we have presented standard-ized odds ratios for each model term as a qualita-tive approach of evaluating their relative influenceon extinction risk and risk of population decline(Kaufman 1996). All population simulations andstatistical analyses were conducted using Rversion 2.15.1 for Linux (http://www.r-project.org).
RESULTSFor most provinces, retaliatory killing had substan-tially higher odds ratios than trophy harvest, bothfor extinction risk and risk of population decline(Table 2). This highlights that within the evaluatedranges, variation in retaliatory killing had a morepronounced effect on extinction risk and risk ofpopulation decline than variation in trophy harvest.The odds ratios for all interaction terms were closeto 1 for most provinces (Table 2), which suggestslimited interactive effects between trophy harvest,retaliatory killing and population size. The lowodds ratios for the effect of population size onextinction risk highlight the positive effect of popu-lation size on population persistence. However, forsome provinces, as well as for South Africa as awhole, the odds ratios for the effect of populationsize on risk of population decline was above 1,which suggests a positive association betweenpopulation size and risk of decline in these areas.
At low to intermediate levels of retaliatory killing,neither the South African leopard population as awhole nor any of the provinces seem to be underserious extinction risk under the current level oftrophy harvest (South Africa: 69 animals per year;Limpopo: 51 animals; Mpumalanga: 7 animals;North West: 7 animals; KwaZulu-Natal: 4 animals;Appendix 4), but extinction risk increased acrossall harvest levels with increasing rates of retaliatorykilling in all but two provinces (Fig. 1). Moreover,both the South African leopard population as wellas some provinces, most notably Limpopo,Mpumalanga, KwaZulu-Natal and the WesternCape, showed persistent risks of population declineacross all evaluated scenarios (Fig. 2).
DISCUSSIONOur simulations pointed to substantially highereffects of variation in retaliatory killing than ofvariation in trophy harvest on population persis-tence. We acknowledge that these results could atleast be party caused by the wider ranges of retal-iatory killing that we simulated compared toharvest, for which the upper limits generally did not
reach critical levels to cause increased risks ofextinction or population decline. However, weestimated upper limits of retaliatory killing fromempirical data, and we set upper limits of harvestto four times the current harvest levels, which mustbe regarded as liberate. We therefore argue thatour results are empirically relevant, because theyhighlight that the number of animals killed in retal-iatory actions can widely exceed the number ofanimals killed for recreational harvest. Our studythus conforms to previous concerns about thepotential effects of persecution on large carnivorepopulation persistence (Treves & Karanth 2003).However, our results contradict some recent studiesthat have suggested negative demographiceffects of trophy harvest alone (Caro et al. 2009;Packer et al. 2011), and support suggestions thatretaliatory killing may enhance such negativeeffects (Loveridge et al. 2009).
The limited relative effect of trophy harvest onpopulation persistence highlights its potential as astrategy to increase local tolerance for large carni-vores by providing sustainable incomes (Lewis &Jackson 2005; Treves 2009). However, huntersneed to be actively involved in management plans,landowners need to receive associated benefits,and incomes need to exceed any financial lossesassociated with the presence of predators forpositive effects to occur (Treves & Martin 2011). Inour simulations, the demographic impacts ofretaliatory killing and trophy harvest were alsocompletely additive. The demographic impacts oftrophy harvest could therefore be further minimizedif it is designed to be compensatory with respect tothe killing of conflict animals, i.e. if it is directedtowards areas with high potential for conflict, ortargeting animals that have been identified tocause conflict (Stein et al. 2010).
Our results call for concern regarding thesustainability of the South African leopard popula-tion. Although our simulations suggested that theextinction risk of the South African leopard popula-tion was negligible within the next 25 years, wefound an unequivocal risk of population decline inSouth Africa as a whole as well as for several prov-inces. Because of the dramatic effects of retalia-tory killing on population persistence, our resultshighlight that reliable data on legal and illegal retal-iatory killing are paramount for our ability to assessthe sustainability of any given harvest level. More-over, our results suggest that control of conflict-related killing may be more effective than restrict-ing the CITES quota for harvested animals. Such
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 119
120 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014Ta
ble
2.S
tand
ardi
zed
odds
ratio
sin
dica
ting
the
effe
cts
oftr
ophy
harv
est,
reta
liato
ryki
lling
and
popu
latio
nsi
zeas
wel
las
thei
rin
tera
ctio
nson
extin
ctio
nris
kan
dris
kof
popu
latio
nde
clin
ew
ithin
25ye
ars
ofsi
mul
ated
leop
ard
popu
latio
ns.O
dds
ratio
sw
ere
deriv
edfr
ombi
nom
ialm
odel
sba
sed
onbi
nary
outc
omes
from
ast
ocha
stic
popu
la-
tion
mod
el,a
ndde
scrib
eth
ech
ange
inod
dsof
extin
ctio
nor
popu
latio
nde
clin
eov
eron
est
anda
rdde
viat
ion
unit
ofch
ange
inea
chof
the
pred
icto
rs(i.
e.ha
rves
t,re
talia
tory
killi
ngan
dpo
pula
tion
size
). Lim
popo
Mpu
mal
anga
Nor
thW
est
Gau
teng
Nor
ther
nFr
eeS
tate
Kw
aZul
u-W
este
rnE
aste
rnS
outh
Afr
ica
Pro
vinc
eC
ape
Nat
alC
ape
Cap
e
Ext
inct
ion
Har
vest
1.58
1.10
0.04
0.36
0.97
0.37
1.40
0.97
0.98
4.89
Ret
alia
tory
killi
ng2.
99×
104
1.12
×10
520
.24
0.01
92.7
60.
621.
72×
104
9.28
×10
11.
68×
101
0.11
Pop
ulat
ion
size
8.87
×10
–60.
000.
000.
020.
022.
85×
10–5
0.02
0.02
0.01
8.38
×10
–9
Har
vest
×1.
001.
001.
327.
27×
103
1.08
7.18
×10
20.
961.
081.
031.
00R
etal
iato
ryki
lling
Har
vest
×1.
001.
001.
021.
041.
001.
221.
001.
001.
001.
00P
opul
atio
nsi
ze
Ret
alia
tory
killi
ngx
1.00
1.00
1.02
1.26
0.99
1.35
1.00
0.99
1.00
1.00
Pop
ulat
ion
size
Har
vest
×R
etal
iato
ry1.
001.
001.
000.
991.
000.
851.
001.
001.
001.
00ki
lling
×P
opul
atio
nsi
ze
Dec
line
Har
vest
1.00
1.00
2.27
1.19
1.18
1.89
1.00
1.00
1.00
6.24
Ret
alia
tory
killi
ng1.
001.
002.
66×
103
6.28
×10
16.
30×
102
2.23
1.00
1.00
1.00
1.00
×10
9
Pop
ulat
ion
size
1.00
1.00
0.05
0.07
0.72
0.01
1.00
1.00
1.00
3.33
Har
vest
×R
etal
iato
ry1.
001.
000.
830.
020.
740.
041.
001.
001.
000.
98ki
lling
Har
vest
×P
opul
atio
n1.
001.
001.
001.
001.
000.
931.
001.
001.
001.
00si
ze
Ret
alia
tory
killi
ng×
1.00
1.00
0.98
0.91
0.99
1.09
1.00
1.00
1.00
1.00
Pop
ulat
ion
size
Har
vest
×R
etal
iato
ry1.
001.
001.
001.
001.
001.
041.
001.
001.
001.
00ki
lling
×P
opul
atio
nsi
ze
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 121
Fig. 1. Predicted extinction risk after 25 years of simulations of provincial leopard populations in South Africa underrealistic ranges of trophy harvest and retaliatory killing. Extinction risks were calculated from logistic regressionmodels based on binary outcomes from a stochastic population model.Extinction risks were calculated on the medianestimated population estimates within each province, which are indicated in the figure.
122 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
Fig. 2. Predicted risk of population decline after 25 years of simulations of provincial leopard populations in SouthAfrica under realistic ranges of trophy harvest and retaliatory killing. Risks of decline were calculated from logisticregression models based on binary outcomes from a stochastic population model, where a decline was scored if thepopulation size after 25 years of simulation was lower than the initial population size. Risks of decline were calculatedbased on the median estimated population estimates within each province, which are indicated in the figure.
conflict mitigation could for instance include theuse of livestock protection schemes such as live-stock-guarding dogs (Ogada et al. 2003), but alsoeducational components targeting wildlife associ-ated value norms within local communities (Madden2004). Adequate and functional compensatoryprogrammes have also been shown effective(Swenson & Andrén 2005), but are highly costlyalternatives for financially developing countries.
Some provincial populations experienced highprobabilities of decline without experiencing eitherharvest or retaliatory killing.Since these provincesall had large estimated population sizes, we suggesttwo, not necessarily mutually exclusive, explana-tions. First, we parameterized emigration as aproportion of population size, and calculated immi-gration as the sum of the emigrants from neigh-bouring provinces. Therefore, provinces with largepopulations that were surrounded by muchsmaller ones could experience a net deficit inanimals caused by a net emigration that could beenough to cause high risks of population declines.Second, our median population size estimatescould be above the demographically stable popu-lation sizes given our demographic parameters, sothat the observed declines could be the results of ademographic stabilization process. However, ourmedian estimated population size for South Africa(4476 leopards) falls well below the only previousformal attempt to estimate the South African popu-lation size (23 472 individuals; Martin & deMeulenaar 1988), and generally agrees with esti-mates based on expert opinion (e.g. 2000–3000animals: Norton 1990, and 4250 animals: Dalyet al. 2005). We therefore suggest that factorsother than overestimated population sizes causedthe elevated probabilities of decline at zero harvestand retaliatory killing in these provinces. We notethat three of these provinces, Limpopo, Mpuma-langa and KwaZulu-Natal, currently have leopardharvest, which could call for caution in settingharvest quotas for these areas.
Finally, we identify some caveats with our simu-lation efforts. First, our simulations were based onlimited demographic data, including only SouthAfrican studies (Bailey 2005; Balme et al. 2010).We also indirectly estimated leopard populationsizes from a habitat suitability model (VanDerWalet al. 2009; but see Swanepoel et al. 2013), whichcould have overestimated population size sinceleopard density estimates were biased towardsprotected areas. However, we regard our popula-tion estimates to be good alternatives to expert
opinion estimates, which were our only otheroption, and our estimates generally agree withsuch estimates (Norton 1990; Daly et al. 2005).Second, our model did not include indirect effect ofharvest, for instance related to infanticide (Creel &Rotella 2010). Although such behaviour has beensuggested to influence the demographic impactsof harvest both in leopards (Packer et al. 2009;Balme et al. 2013) and in other large carnivorespecies (e.g. black bears, Ursus americanus:Lecount 1987, brown bears, Ursus arctos: Swensonet al. 1997, and lions, Panthera leo: Whitman et al.2004), there are no data available on how maleturnover influence such infanticide rates in leopards.Since infanticidal mortality is already included inour model through the survival estimates, wedo not believe that explicit incorporation ofinfanticidal effects in our model would render anyimprovements in our interpretations. Moreover,although we recognize that our models may haveunderestimated the influence of male biasedkilling (i.e. harvest), the relative strength of retalia-tory killing versus harvest suggests that the inclu-sion of social processes in our models likely wouldnot have change the qualitative outcome of oursimulations.
To conclude, our study suggests that control oflegal and illegal killing related to human–carnivoreconflict may be more effective in promoting thepersistence of large carnivore populations thansetting restricted trophy harvest quotas. Becausewe found limited effects of variation in harvest onpopulation persistence, trophy harvest could beone potential avenue for such conflict mitigation.Our study further highlights the necessity for accu-rate estimates of legal and illegal retaliatory killingfor any reliable assessment of sustainable harvestlevels.Although our simulations suggested low ex-tinction risk for the South African leopard popula-tion within the next 25 years, they indicated veryhigh probabilities of population decline. Our studytherefore calls for concern regarding the long-termviability of the South African leopard population.However, because of the large effects of retaliatorykilling compared to harvest, we suggest that actionsdirected to decrease conflict-related killing ofleopards may be more effective than restrictingtrophy harvest in securing the long-term viability ofSouth African leopards.
ACKNOWLEDGEMENTSFunding for the project was received from Inter-
national Foundation for Science grant no. D/4984-1,
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 123
Wild Foundation grant no. 2008-011, WilsonFoundation, University of Pretoria and the Endan-gered Wildlife Trust. L.H.S. was further supportedby National Research Foundation of South Africa(NRF) grant no. 74819, F.D. was supported by aresearch fellowship from the University of Pretoria,M.S. by the DST-NRF Centre of Excellence forInvasion Biology and F.D. and M.S. were sup-ported by the NRF.
REFERENCESBAILEY, T.N. 2005. The African leopard: ecology and
behaviour of a solitary felid. Columbia UniversityPress, New York.
BAKER, P.J., BOITANI, L., HARRIS, S., SAUNDERS, G.& WHITE, P.C.L. 2008. Terrestrial carnivores andhuman food production: impact and management.Mamm. Rev. 38: 123–166.
BALME, G.A., SLOTOW, R. & HUNTER, L.T.B. 2009.Impact of conservation interventions on the dynam-ics 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 incarnivore conservation: leopards in the Phinda–Mkhuze Complex, South Africa. Anim. Conserv. 13:315–323.
BALME, G.A., BATCHELOR, A., DE WORONIN BRITZ,N., SEYMOUR, G., GROVER, M., HES, L.,MACDONALD, D.W. & HUNTER, L. 2013. Reproduc-tive success of female leopards Panthera pardus: theimportance of top-down processes. Mamm. Rev. 43:221–237.
CARO, T.M., YOUNG, C.R., CAULDWELL, A. E. &BROWN, D.D.E. 2009. Animal breeding systems andbig game hunting: models and application. Biol.Conserv. 142: 909–929.
CHASE-GREY, J.N. 2011. Leopard population dynamics,trophy hunting and conservation in the SoutpansbergMountains, South Africa. Ph.D. thesis, DurhamUniversity, Durham. Available at Durham E-ThesesOnline: http://etheses.dur.ac.uk/823/ (accessedDecember 2013).
CREEL, S. & ROTELLA, J.J. 2010. Meta-analysis ofrelationships between human offtake, total mortalityand population dynamics of gray wolves (Canislupus). PLOS ONE 5: e12918.
CROSS, P.C. & BEISSINGER, S.R. 2001. Using logisticregression to analyze the sensitivity of PVA models:acomparison of methods based on African wild dogmodels. Conserv. Biol. 15: 1335–1346.
DALERUM, F., SHULTS, B. & KUNKEL, K. 2008. Esti-mating sustainable harvest in wolverine populationsusing logistic regression. J. Wildlife Manage. 72:1125–1132.
DALY, B., POWER, J., CAMACHO, G., TRAYLORHOLZER, K., BARBER, S., CATTERALL, S.,FLECTHER, P., MARTINS, Q., OWEN, C., THAL, T.&FRIEDMAN, Y. 2005. Leopard (Panthera pardus)population and habitat viability assessment. Unpub-lished report. Conservation Breeding Specialist
Group (SSC / IUCN) / CBSG and Endangered Wild-life Trust, Johannesburg.
DAR, N.I., MINHAS, R.A., ZAMAN, Q. & LINKIE, M.2009. Predicting the patterns, perceptions andcauses of human–carnivore conflict in and aroundMachiara National Park, Pakistan.Biol.Conserv.142:2076–2082.
DICKMAN, A.J.2010.Complexities of conflict: the impor-tance of considering social factors for effectivelyresolving human–wildlife conflict. Anim. Conserv. 13:458–466.
FRIEDMANN, Y. & DALY, B. 2004. Red Data Book of themammals of South Africa: a conservation assess-ment. Conservation Breeding Specialist Group(SS/IUCN) and Endangered Wildlife Trust, Johan-nesburg.
FUNSTON, P.J., GROOM, R.J. & LINDSEY, P.A. 2013.Insights into the management of large carnivoresfor profitable wildlife-based land uses in Africansavannas. PLOS ONE 8: e59044.
INSKIP, C. & ZIMMERMAN, A. 2009. Human-felid con-flict: a review of patterns and priorities worldwide.Oryx 43: 18–34.
KACZENSKY, P., JERINA, K., JONOZOVI�, M.,KROFEL, M., SKRBIN�EK, T., RAUER, G., KOS, I.& GUTLEB, B. 2011. Illegal killings may hamperbrown bear recovery in the Eastern Alps. Ursus 22:37–46.
KAUFMAN, R.L. 1996. Comparing effects in dichoto-mous logistic regression: a variety of standardizedcoefficients. Soc. Sci. Q. 77: 90–109.
LECOUNT, A. L. 1987. Causes of black bear cub mortal-ity. Int. Conf. Bear Res. Manage. 7: 75–82.
LEWIS, D. & JACKSON, J. 2005. Safari hunting andconservation on communal land in southern Africa.In:R.Woodroffe, S.J.Thirgood & A.Rabinowitz (Eds),People and wildlife: conflict or coexistence? (pp.239–252). Cambridge University Press, Cambridge.
LINDSEY, P.A., ROULET, P.A.& ROMANACH, S.S.2007.Economic and conservation significance of thetrophy hunting industry in sub-Saharan Africa. Biol.Conserv. 134: 455–469.
LINDSEY, P.A., MARNEWICK, K., BALME, G. &SWANEPOEL, L.H. 2011. Non detriment findingassessment for the trophy hunting of leopards inSouth Africa. Report for the South African NationalBiodiversity Institute (SANBI). Endangered WildlifeTrust, Johannesburg.
LINDSEY, P.A., BALME, G.A., BOOTH, V.R. &MIDLANE, N. 2012. The significance of African lionsfor the financial viability of trophy hunting and themaintenance of wild land. PLOS ONE 7: e29332.
LINNELL, J.D.C., ANDERSEN, R., KVAM, T., ANDREN,H., LIBERG, O., ODDEN, J. & MOA, P.F. 2001. Homerange size and choice of management strategy forlynx in Scandinavia. Environ. Manage. 27: 869–879.
LOVERIDGE, A.J., PACKER, C. & DUTTON, A. 2009.Science and the recreational hunting of lions. In: B.Dickson, J.Hutton & W.M.Adams (Eds), Recreationalhunting, conservation and rural livelihoods’ (pp.108–124). Blackwell Publishing, Oxford.
MADDEN, F. 2004. Creating coexistence betweenhumans and wildlife: global perspectives on local
124 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
efforts to address human–wildlife conflict. Hum.Dimens. Wildl. 9: 247–257.
MARTIN, R.B. & DE MEULENAAR, T. 1988. Survey ofthe status of the leopard (Panthera pardus) insub-Saharan Africa. CITES Secretariat, Lausanne.
McCARTHY, M.A., BURGMAN, M.A. & FERSON, S.1995. Sensitivity analysis for models of populationviability. Biol. Conserv. 73: 93–100.
McCARTHY, M.A., BURGMAN, M.A. & FERSON, S.1996. Logistic sensitivity and bounds for extinctionrisks. Ecol. Model. 86: 297–303.
NORTON, P.M.1990.How many leopards? A criticism ofMartin and de Meulenaer’s population estimates forAfrica. S. Afr. J. Sci. 86: 218–219.
NOSS, R.F., QUIGLEY, H.B., HORNOCKER, M.G.,MERRILL, T. & PAQUET, P.C. 1996. Conservationbiology and carnivore conservation in the RockyMountains. Conserv. Biol. 10: 949–963.
OGADA, M.O., WOODROFFE, R., OGUGE, N.O. &FRANK, L.F. 2003. Limiting depredation by Africancarnivores: the role of livestock husbandry. Conserv.Biol. 17: 1521–1530.
PACKER, C., KOSMALA, M., COLLEY, H.S., BRINK, H.,PINTEA, L., GARSHELIS, D., PURCHASE, G.,STRAUSS, M., SWANSON, A., BALME, G.,HUNTER, L. & NOWELL, K. 2009. Sport hunting,predator control and conservation of large carni-vores. PLOS ONE 4: e5941.
PACKER, C., BRINK, H., KISSUI, B.M., MALITI, H.,KUSHNIR, H. & CARO, T. 2011. Effects of trophyhunting on lion and leopard populations in Tanzania.Conserv. Biol. 25: 142–153.
PE’ER, G.U.Y., MATSINOS, Y.G., JOHST, K., FRANZ,K.W., TURLURE, C., RADCHUK, V., MALINOWSKA,A.H., CURTIS, J.M.R., NAUJOKAITIS-LEWIS, I.,WINTLE, B.A.& HENLE, K.2013.A protocol for betterdesign, application, and communication of popula-tion viability analyses. Conserv. Biol. 27: 644 656.
PURVIS, A., GITTLEMAN, J.L., COWLISHAW, G. &MACE, G. M. 2000. Predicting extinction risk indeclining species. Proc. R. Soc. Lond. B. 267:1947–1952.
REED, J.M., MILLS, L.S., DUNNING JR, J.B., MENGES,E.S., McKELVEY, K.S., FRYE, R., BEISSINGER, S.R.,ANSTETT, M. & MILLER P. 2002. Emerging issues inpopulation viability analysis. Conserv. Biol. 16: 7–19.
STARFIELD, A.M.1997.A pragmatic approach to model-ling for wildlife management. J. Wildl. Manage. 61:261–270.
STEIN, A.B., FULLER, T.K., DAMERY, D.T., SIEVERT, L.
& MARKER, L.L. 2010. Farm management and eco-nomic analyses of leopard conservation in north-central Namibia. Anim. Conserv. 13: 419–427.
SWANEPOEL, L.H, LINDSEY, P., SOMERS, M.J., VANHOVEN, W. & DALERUM, F. 2013. Extent and frag-mentation of suitable leopard habitat in South Africa.Anim. Conserv. 16: 41–50.
SWANEPOEL, L.H., SOMERS, M.J., VAN HOVEN, W.,SCHEISS-MEJIER, M., OWEN, C., SNYMAN, A.,MARTINS, Q., CAMACHO, G., BOSHOFF, W. &DALERUM, F. In press. Survival rates and causes ofmortality of leopards (Panthera pardus) in southernAfrica. Oryx
SWENSON, J.E. & ANDRÉN, H. 2005. A tale of twocountries: large carnivore depredation and compen-sation schemes in Sweden and Norway. In: R.Woodroffe, S. Thirgood & A. Rabinowitz (Eds),People and wildlife: conflict or coexistence? (pp.323–339). Cambridge University Press, Cambridge.
SWENSON, J.E., SANDEGREN, F., SODERBERG, A.,BJARVALL, A., FRANZEN, R. & WABAKKEN, P.1997. Infanticide caused by hunting of male bears.Nature 386: 450–451.
THORN, M., GREEN, M., DALERUM, F., BATEMAN, P.W. & SCOTT, D.M. 2012. What drives human–carni-vore conflict in the North West Province of SouthAfrica? Biol. Conserv. 150: 23–32.
TREVES, A. 2009. Hunting for large carnivore conserva-tion. J. Appl. Ecol. 46: 1350–1356.
TREVES, A. & KARANTH, K.U. 2003. Human-carnivoreconflict and perspectives on carnivore managementworldwide. Conserv. Biol. 17: 1491–1499.
TREVES, A. & MARTIN, K.A. 2011. Hunters as stew-ards of wolves in Wisconsin and the NorthernRocky Mountains, USA. Soc. Natur. Resour. 24:984–994.
VANDERWAL, J., SHOO, L.P., JOHNSON, C.N. &WILLIAMS, S.E. 2009. Abundance and the environ-mental niche: environmental suitability estimatedfrom niche models predicts the upper limit of localabundance. Am. Nat. 174: 282–291.
WHITMAN, K., STARFIELD, A.M., QUADLING, H.S. &PACKER, C. 2004. Sustainable trophy hunting ofAfrican lions. Nature 428: 175–178.
WOODROFFE, R. & GINSBERG, J.R. 1998. Edgeeffects and the extinction of populations insideprotected areas. Science 280: 2126–2128.
WOODROFFE, R. & FRANK, L.G. 2005 Lethal control ofAfrican lions (Panthera leo): local and regional popu-lation impacts. Anim. Conserv. 8: 91–98.
Responsible Editor: G.A. Balme
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 125
126 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
APPENDIX 1. Parameters used for modelling the effects of leopard harvest on the probability of population declineand probability of extinction in South African leopard population. Fecundity, survival and dispersal were in the simula-tions regarded as Poisson and binomially distributed, respectively. Therefore, the variances of these estimates wereregarded to be equal to the mean.Age and sex structure of harvest and retaliatory killing were kept as fixed throughoutall simulations, to enhance the focus on variability in harvest and retaliatory killing per se.
Parameter Value
Fecunditya 1.41Survival
adult maleb 0.91adult femaleb 0.85subadult maleb 0.82subadult femaleb 0.91juvenilesb 0.33
Dispersalc
subadult males 47%subadult females 14%adults 5%
Trophy harvestd
age structure 95%sex ratio 68%
Damage-causing animalse
age structure 72%sex ratio 65%
aMedian from Bailey (2005), Owen (2006), Martin & de Meulenaar (1988), Balme et al. (2009; 2010), expressed as number of offspring perfemale per year.
bSurvival based on protected area estimates from the meta-analysis in Swanepoel et al. (in press).cSubadult dispersal based on 36 collared subadults taken from Bailey (2005), Balme et al. (2010), Bothma et al. (1997), Owen (2006),Stander et al. (1997), expressed as percentage animals dispersing per year.
dData on demographic structure of trophy harvested animals collected from Conservation offices based on hunter return forms in LimpopoProvince, South Africa, n = 47. Age structure is expressed as percentage adults and sex ratio as percentage males.
eData on demographic structure of destroyed problem leopards in Limpopo Province (n =21) and Cape Province (currently Eastern CapeProvince, Western Cape Province and Northern Cape Province) (n = 37). Data from Limpopo Province were collected directly fromconservation offices and data from the Cape Province data taken from C, Stuart and T, Stuart, unpubl. data. Age structure is expressed aspercentage adults and sex ratio as percentage males.
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 127A
PP
EN
DIX
2.Le
opar
dde
nsiti
es(n
o./1
00km
²)in
each
prov
inci
albi
ome
that
was
used
toes
timat
ele
opar
dpo
pula
tion
size
inS
outh
Afr
ica.
Stu
dyno
.P
rovi
nce
Bio
me/
bior
egio
nS
tudy
site
Den
sity
Met
hod
Year
Sou
rce
1M
pum
alan
gaLo
wve
ldS
outh
ern
Kru
ger
Nat
iona
lPar
k15
.63
1a19
72B
aile
y(2
005)
2M
pum
alan
gaLo
wve
ldC
entr
alK
ruge
rN
atio
nalP
ark
15.4
119
72B
aile
y(2
005)
3M
pum
alan
gaLo
wve
ldC
entr
alK
ruge
rN
atio
nalP
ark
4.25
119
72B
aile
y(2
005)
4M
pum
alan
gaLo
wve
ldS
outh
ern
Kru
ger
Nat
iona
lPar
k3.
011
1972
Bai
ley
(200
5)5
Lim
popo
(Bot
swan
a)Lo
wve
ldTu
liG
ame
Res
erve
82b
2005
–200
6S
teyn
(200
8)6
Lim
popo
Low
veld
Nor
ther
nK
ruge
rN
atio
nalP
ark
6.33
119
72B
aile
y(2
005)
7Li
mpo
poB
ushv
eld
Sou
tpan
sber
g19
.97
220
08C
hase
-Gre
y(2
011)
8Li
mpo
poB
ushv
eld
Wel
gevo
nden
Gam
eR
eser
ve4.
332
2008
–201
0L.
HS
wan
epoe
l,un
publ
.9
Lim
popo
Bus
hvel
dW
ater
berg
Bio
sphe
refa
rms
3.18
220
09L.
HS
wan
epoe
l,un
publ
.10
Lim
popo
Bus
hvel
dLa
pala
laW
ilder
ness
3.05
220
08–2
010
L.H
Sw
anep
oel,
unpu
bl.
11K
waZ
ulu-
Nat
alLo
wve
ldM
kuze
Nat
iona
lPar
k11
.11
220
05B
alm
eet
al.(
2010
)12
Kw
aZul
u-N
atal
lLo
wve
ldP
hind
aP
rivat
eG
ame
Res
erve
7.51
220
05B
alm
eet
al.(
2010
)13
Kw
aZul
u-N
atal
Low
veld
Zul
ulan
dR
hino
Res
erve
4.2
220
06C
hapm
an&
Bal
me
(201
0)14
Kw
aZul
u-N
atal
Low
veld
Gam
efa
rmar
eain
KZ
N-N
atal
2.49
220
05B
alm
eet
al.(
2010
)15
Nor
thW
est
Bus
hvel
dM
adik
we
Nat
ure
Res
erve
3.86
3c20
03H
.L.P
.Kel
ly,u
npub
l.16
Nor
thW
est
Bus
hvel
dP
ilans
berg
Nat
ure
Res
erve
3.45
4d20
07H
ayw
ard
etal
.(20
07)
17W
este
rnC
ape
Fyn
bos
Ced
erbe
rgM
ount
ains
2.3
5e20
04–2
007
Mar
tins
(201
0)18
Wes
tern
Cap
eF
ynbo
sC
eder
berg
Mou
ntai
ns1.
85
2004
–200
7M
artin
s(2
010)
19W
este
rnC
ape
Fyn
bos
Ced
erbe
rgM
ount
ains
0.8
319
85N
orto
n&
Law
son
(198
5)20
Wes
tern
Cap
eK
aroo
Ced
erbe
rgM
ount
ains
0.9
520
04–2
007
Mar
tins
(201
0)21
Wes
tern
Cap
eK
aroo
Ced
erbe
rgM
ount
ains
0.25
520
04–2
007
Mar
tins
(201
0)22
Eas
tern
Cap
eF
ynbo
sB
avia
ansk
loof
1.3
520
07M
cMan
us(2
009)
23E
aste
rnC
ape
Fyn
bos
Bav
iaan
sklo
of0.
35
2007
McM
anus
(200
9)24
Eas
tern
Cap
eLo
wve
ldB
avia
ansk
loof
1.3
520
07M
cMan
us(2
009)
25E
aste
rnC
ape
Low
veld
Bav
iaan
sklo
of0.
35
2007
McM
anus
(200
9)26
Nor
ther
nC
ape
Kal
ahar
iK
gala
gadi
Nat
iona
lpar
k0.
625
1984
Bot
hma
&Le
Ric
he(1
984)
27N
orth
ern
Cap
eK
aroo
Ora
nge
Riv
erB
asin
0.35
6f19
88C
.Stu
art&
T.S
tuar
t,un
publ
.a M
etho
d1;
leop
ard
dens
ityw
ases
timat
edby
usin
gho
me
rang
eov
erla
pan
dpr
eyde
nsity
.b M
etho
d2,
leop
ard
dens
ityw
ases
timat
edby
appl
ying
mar
kre
capt
ure
fram
ew
ork
toca
mer
atr
appi
ngda
ta.
c Met
hod
3,le
opar
dde
nsity
was
estim
ated
byus
ing
hom
era
nge
anal
ysis
and
over
lap.
d Met
hod
4,le
opar
dde
nsity
was
estim
ated
bym
akin
gpr
edic
tions
onpr
eyde
nsity
and
prey
pref
eren
ces.
e Met
hod
5,le
opar
dde
nsity
was
estim
ated
with
GP
Sco
llar
hom
era
nge
data
.f M
etho
d6,
leop
ard
dens
ityw
ases
timat
edby
usin
gsi
gnsu
rvey
data
.A
dd
itio
nal
refe
ren
ces
for
Ap
pen
dix
es1
and
2B
OT
HM
A,J
.D.P
.&LE
RIC
HE
,E.A
.N.1
984.
Asp
ects
ofth
eec
olog
yan
dth
ebe
havi
our
ofth
ele
opar
dP
anth
era
pard
usin
the
Kal
ahar
iDes
ert.
Koe
doe
27:2
59–2
79.
CH
AP
MA
N,S
.&B
ALM
E,G
.201
0.A
nes
timat
eof
leop
ard
popu
latio
nde
nsity
ina
priv
ate
rese
rve
inK
waZ
ulu-
Nat
al,S
outh
Afr
ica,
usin
gca
mer
a-tr
aps
and
capt
ure–
reca
ptur
em
odel
s.S
.Afr.
J.W
ildl.
Res
.40:
114–
120.
HAY
WA
RD
,M.W
.,O
’BR
IEN
,J.&
KE
RLE
Y,G
.I.H
.200
7.C
arry
ing
capa
city
ofla
rge
Afr
ican
pred
ator
s:P
redi
ctio
nsan
dte
sts.
Bio
l.C
onse
rv.1
39:2
19–2
29.
MA
RT
INS
,Q.E
.201
0.T
heec
olog
yof
the
leop
ard
Pan
ther
apa
rdus
inth
eC
eder
berg
Mou
ntai
ns.P
.hD
.the
sis
Uni
vers
ityof
Bris
tol,
Bris
tol,
U.K
.M
cMA
NU
S,S
.J.2
009.
Spa
tiale
colo
gyan
dac
tivity
patte
rns
ofle
opar
ds(P
anth
era
pard
us)
inth
eB
avia
ansk
loof
and
Gre
ater
Add
oE
leph
antN
atio
nalP
ark
(GA
EN
P),
Eas
tern
Cap
eP
rovi
nce,
Sou
thA
fric
a.M
.Sc.
thes
is,R
hode
sU
nive
rsity
,Gra
ham
stow
n,S
outh
Afr
ica.
OW
EN
,C.2
006.
Rep
rodu
ctiv
ebi
olog
yan
dpo
pula
tion
ecol
ogy
ofle
opar
ds(P
anth
era
pard
us)
onK
aron
gwe.
M.S
c.th
esis
,Uni
vers
ityof
Kw
aZul
u-N
atal
,Dur
ban,
Sou
thA
fric
a.S
TAN
DE
R,P
.E.,
HA
DE
N,P
.J.,
KA
QE
CE
&G
HA
U19
97.T
heec
olog
yof
asoc
ialit
yin
Nam
ibia
nle
opar
ds.J
.Zoo
l.24
2:34
3–36
4.S
TE
YN
,V.P
.200
8.T
heso
cial
-spa
tialo
rgan
izat
ion
ofle
opar
ds(P
anth
era
pard
us)
inth
eno
rthe
rnTu
liG
ame
Res
erve
.M.S
c.th
esis
,Tsh
wan
eU
nive
rsity
ofTe
chno
logy
,Pre
toria
,Sou
thA
fric
a.
128 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
APPENDIX 3. Minimum, median and maximum leopard densities used in each provincial biome to calculateleopard population sizes in each South African province.
Province Biome/bioregion Density (leopards/100km²)
Minimum Median Maximum
Mpumalangaa Lowveld 3.01 9.80 15.63
Mpumalangab Bushveld 3.05 3.76 19.97
Mpumalangac Grasslandd 0.25 0.60 0.90
Limpopoe Lowveld 6.33 7.20 15.60
Limpopof Bushveld 3.01 3.76 19.97
KwaZulu Natalg Lowveld 2.49 5.86 11.21
KwaZulu Natalc Grassland d 0.25 0.60 0.90
North Westh Bushveldi 3.45 3.66 3.86
North Westj Kalahari 0.31 0.62 1.24
North Westc Grasslandd 0.25 0.60 0.90
Gautengh Bushveldi 3.45 3.66 3.86
Gautengc Grasslandd 0.25 0.60 0.90
Free Stateh Bushveldi 3.45 3.66 3.86
Free Statec Grasslandd 0.25 0.60 0.90
Free Statec Karoo 0.25 0.60 0.90
Northern Capek Fynbosl 0.80 1.30 2.30
Northern Capej Kalahari 0.30 0.62 1.24
Northern Capec Karoo 0.25 0.60 0.90
Western Capek Fynbosl 0.80 1.30 2.30
Western Capec Karoo 0.25 0.60 0.90
Eastern Capem Fynbos 0.30 0.80 1.30
Eastern Capem Lowveld 0.30 0.80 1.30
Eastern Capec Karoo 0.25 0.60 0.90
Eastern Capec Grasslandd 0.25 0.60 0.90
Study numbers below taken from literature sources outlined in Appendix 2.aMin taken from study 4, median calculated from studies 1–4, and max taken from study 1.bMin taken from study 10, median calculated from studies 7–10, and max taken from study 7.cMin taken from study 21, median calculated from studies 20, 21, 27, and max taken from study 20.dWe assumed equal leopard densities in the Grassland biome as in the Karroo biome in all provinces.eMin taken from study 6, median calculated from studies 5–6, and max taken from study 5.fMin taken from study 10, median calculated from studies 7–10, and max taken from study 7.gMin taken from study 14, median calculated from studies 11–14, and max taken from study 11.hMin taken from study 16, median calculated from studies 15–16, and max taken from study 15.iWe assumed equal leopard densities in the North West, Free State and Gauteng provinces in the bushveld biomes.jMin calculated as half density from 26, median taken from study 26, and max calculated as the double density from study 26.kMin taken from study 19, median calculated from studies 17,18,19, 22, 23, and max taken from study 17.lWe assumed equal leopard densities in Northern Cape and Western Cape provinces in the Fynbos biome.mMin taken from study 23, median calculated from studies 22,23,24,25, and max taken from study 22.Since we did not have estimated leopard densities for all biomes or provinces, we assumed equal leopard densities in the karoo andgrasveld biome in all provinces, equal densities in the bushveld biome in the North West, Free State and Gauteng provinces, and equaldensities for the fynbos biome in the Northern Cape and Western Cape. For the Kalahari biome, which only had one density estimate, weused half the estimated density as the minimum and double the estimated density as the maximum density.
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 129
APPENDIX 4. Number of animals of each sex and age class killed for trophy harvest and under Damage CausingAnimal (DCA) permits for each South African province during 2002–2010, as well as number of issued DCA permits inLimpopo (2002–2010) and KwaZulu-Natal (2005–2010).
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
Limpopo2002 Adult 1 0 0 0 0 0 22
Subadult 0 0 0 0 0 0Unknown 0 0 27 0 0 3
2003 Adult 4 1 0 1 0 0 13Subadult 0 0 0 0 0 0Unknown 0 0 30 0 0 0
2004 Adult 1 1 0 0 0 0 22Subadult 0 0 0 0 1 0Unknown 0 1 30 0 0 0
2005 Adult 1 1 0 0 0 0 17Subadult 0 0 0 0 0 0Unknown 0 0 12 0 0 1
2006 Adult 4 2 0 0 0 0 21Subadult 1 0 0 0 0 0Unknown 0 0 9 0 0 1
2007 Adult 6 2 0 0 0 0 46Subadult 1 0 0 0 0 0Unknown 0 0 13 0 0 2
2008 Adult 2 0 0 1 0 0 66Subadult 0 0 0 0 0Unknown 0 0 47 0 0 65
2009 Adult 0 0 0 0 0 0 56Subadult 0 0 0 0 0 1Unknown 0 0 65 0 0 57
2010 Adult 10 5 0 0 0 0 63Subadult 0 1 0 0 0 0Unknown 1 0 34 0 0 0
Mpumalanga2002 Adult 4 2 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2003 Adult 3 3 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 5 2 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 1 0 0 0
2005 Adult 4 3 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 4 2 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 6 0 0 0 0 0Subadult 0 0 0 0 0 0
Continued on p. 130
130 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
Unknown 0 0 0 0 0 0
2008 Adult 5 2 2 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 1
2009 Adult 2 2 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 7 0 0 0
North West2002 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 19 0 0 0
2004 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 17 0 0 0
2005 Adult 7 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 11 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 3 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 2 0 0 0
2008 Adult 8 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 8 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 2 0 0 0
2010 Adult 4 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 3 0 0 0
Gauteng2002 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 1 0 0 0
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 1 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Continued on p. 131
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 131
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
2005 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 2 0 0 0
2006 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2008 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Northern Cape2002 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2005 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 1
2007 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 1
2008 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 3
2009 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 1
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Free State2002 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Continued on p. 132
132 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
Unknown 0 0 0 0 0 0
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2005 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2008 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
KwaZulu-Natal2002 Adult 1 1 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2003 Adult 1 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 1 1 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2005 Adult 1 1 0 0 0 0 7Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 1 0 0 2 0 0 2Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 2
2007 Adult 0 1 0 0 0 0 5Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2008 Adult 1 1 0 1 0 0 4Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 5 0 0 1 0 0 8Continued on p. 133
Swanepoel et al.: Trophy harvest and retaliatory killing of large carnivores: South African leopards 133
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 5 0 0 1 0 0 8Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Western Cape1977 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 2 4 0
1978 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 9 4 0
1979 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 9 1 0
1980 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 5 10 0
2002 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2005 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2008 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Eastern Cape2002 Adult 0 0 0 0 0 0
Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
Continued on p. 134
134 South African Journal of Wildlife Research Vol. 44, No. 2, October 2014
Year Age Harvest DCA DCA
Male Female Unknown Male Female Unknown permits
2003 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2004 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2005 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2006 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2007 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2008 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2009 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0
2010 Adult 0 0 0 0 0 0Subadult 0 0 0 0 0 0Unknown 0 0 0 0 0 0