ResearchPDF Available

Expedition report: A game of cats & elephants: safeguarding big cats, elephants and other species of the African savannah, Namibia (August - November 2013)

Authors:
  • Biosphere Expeditions

Abstract and Figures

Abstract This research project started in 2012 and was based on Okambara Elephant Lodge, a game farm located 85 km south of Windhoek’s international airport, in the Khomas region of central Namibia. Okambara is game-fenced and comprises an area of 150 km2. This report covers the survey work conducted during the period of August–November 2013. The key study species were the African leopard (Panthera pardus) and the African elephant (Loxodonta africana). Leopards are protected animals and listed as “Near Threatened” by the IUCN (International Union for Conservation of Nature). However, the conservation of leopards outside of protected areas in Namibia is not assured. Their “problem predator” image and high trophy value, together with habitat loss, habitat fragmentation and local outbreaks of wildlife diseases, are the main threats. These threats and the lack of scientific data on this species living on commercial farmland demonstrate the need for research. Elephants are an increasingly desired game species on farms which host tourists regularly due to their high attractiveness for visitors. However, the impact of elephants on confined areas and the comprised ecosystem are rarely studied and more information is needed in order to create appropriate management guidelines. This study focussed on the spatial ecology and prey preferences of leopards on Namibian farmland. Invasive as well as non-invasive methods were used; invasive methods included trapping and collaring of leopards, whilst non-invasive methods included camera traps, track counts, search for prey remains and faeces collection. Regarding elephants, methods such as tracking of the herd as well as directly observing feeding behaviour were carried out. Data collected on Okambara showed differences in the ecology of leopards living on farmland and in protected areas. Home range sizes differed between the study sites and were bigger than those of leopards living in protected areas, most likely due to habitat preferences, variation in prey availability and lower predator densities compared to protected areas. The camera trap surveys on Okambara yielded a density of 1.9 individuals per 100 km2, a lower density compared to protected areas, thereby confirming the assumption that home range size is related to density. The camera trap surveys also revealed the existence of additional carnivores and related interspecific behaviour showing that they seem to avoid each other and thereby reduce direct competition and conflict. Despite and indeed perhaps because of this, different strands of evidence show that the habitat on Okambara is suitable for the survival and reproduction of different predator species. A study on elephant feeding ecology was also initiated and confirmed the importance of even a small herd of nine individuals as significant ecosystem engineers. This study has important implications for the increasing trend of stocking Namibian game farms with elephants and as such should be continued and expanded. Zusammenfassung Im Jahr 2012 startete dieses Forschungsprojekt auf der Okambara Elephant Lodge, einer Wildtierfarm etwa 85 km südlich von Windhoeks internationalem Flughafen, in der Khomas Hochland Region in Zentral-Namibia. Okambara ist von einem Wildtierzaun umgeben und deckt ein Gebiet von 150 km² ab. Dieser Bericht befasst sich mit Untersuchungen, die dort im Zeitraum August - November 2013 durchgeführt wurden. Im Fokus der Studie standen, der Leopard (Panthera pardus) sowie der Afrikanische Elefant (Loxodonta africana). Der Leopard ist eine geschützte Art und als "potenziell gefährdet" von der IUCN (International Union for Conservation of Nature) eingestuft. Jedoch kommt ein Großteil der namibischen Leopardenpopulation auf kommerziell genutztem Farmland und somit außerhalb geschützter Gebiete vor. Dadurch ist die Erhaltung dieser Art in Namibia nicht gesichert. Ihr Status als "Problem-Beutegreifer", ein hoher Trophäenwert, fortschreitender Verlust von Lebensraum und Wildtierkrankheiten sind ihre stärksten Bedrohungen. Diese Bedrohungen sowie der Mangel an wissenschaftlichen Daten machen es sinnvoll und notwendig, diese Spezies im Lebensraum Farmland besser zu erforschen. Die Haltung von Elefanten wird insbesondere von Touristenfarmen stetig mehr angestrebt, die diese Tiere eine hohe Anziehungskraft auf Besucher besitzen. Jedoch ist bisher unklar, welche Auswirkungen die Haltung von Elefanten in einem eingezäunten Gebiet auf das dort beherbergte Ökosystem hat und zusätzliche Informationen sind notwendig, um artgerechte und langfristig funktionierende Richtlinien zu entwerfen. In dieser Studie standen die räumliche Ökologie von Leoparden auf nambianischen Farmland sowie deren Beutepräferenzen im Mittelpunkt. Sowohl invasive als auch non-invasive Methoden wurden angewandt; invasive Methoden beinhalteten den Fang und die Besenderung von Leoparden, während non-invasive Methoden die Nutzung von Kamerafallen und die Suche nach Kot, Spuren und Überresten von Beutetieren umfassten. Um Daten bezüglich der Elefanten aufzunehmen, wurden Methoden wie das aktive Aufspüren der Herde und direkte Beobachtung der Fressgewohnheiten angewendet. Die auf Okambara aufgenommenen Daten zeigten, dass sich die räumliche Ökologie von Leoparden auf namibianischen Farmland von der in geschützten Gebieten vorkommenden Leoparden unterscheidet. Streifgebiete waren größer als in Schutzgebieten und ist vermutlich auf Habitatpräferenzen, variierende Beutetierdichte sowie geringere Beutegreiferdichte im Vergleich zu geschützten Gebieten zurückzuführen. Der Kamerafallenstudie zufolge weist Okambara eine Leopardendichte von 1,9 Tieren pro 100 km2 auf. Die Leopardendichte auf Farmland ist somit geringer als in geschützten Gebieten und unterstützt die Vermutung eines Zusammenhangs von Streifgebietsgrößen mit vorkommender Dichte. Weiterhin zeigte der Einsatz von Kamerafallen das Vorkommen weiterer Beutegreifer auf Okambara sowie damit verbundenes interspefizisches Verhalten, da sich die verschiedenen Arten in Raum und Zeit zu meiden scheinen, um Konflikte zu vermeiden. Damit ist Okambara ist ein geeignetes Habitat für den Fortpflanzung und Bestand verschiedener Beutegreifer. Eine ernährungsökologische Studie von Elefanten wurde weitergeführt und bestätigte den wichtigen Einfluss selbst der kleinen Okambara-Herde von neun Tieren auf das Gesamtökosystem. In Namibia existiert eine zunehmende Tendenz, Wildtierfarmen mit Elefanten zu bestücken. Die angestoßene Studie liefert hierfür wichtige Rückschlüsse und sollte fortgesetzt werden.
Content may be subject to copyright.
EXPEDITION REPORT
Expedition dates: 4 August
29
November
2013
Report published:
J
une
201
5
ts: safeguarding big cats, elephants
and other species of the African savannah, Namibia
1
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, France, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council &
Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
EXPEDITION REPORT
A game of cats & elephants: safeguarding big cats, elephants
and other species of the African savannah, Namibia
Expedition dates:
4
August
29
November
201
3
Report published:
J
une
201
5
Author
s
:
V
era Menges
Biosphere Expedition
s
Jörg Melzheimer
Biosphere Expeditions
Matthias Hammer (editor)
Biosphere Expeditions
2
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Abstract
This
research project started in 2012 and was based on
Okambara
Elephant Lodge
,
a
game farm located 85 km south of Windhoe
k’s international airport,
in the
Khomas
region
of central Namibia
.
Okambara is game
-
fenced and
comprises
an area of 150
km
2
.
This
report covers the survey w
ork conducted during the period
of
August
November 2013
.
The key study species were
the
African
leo
pard (
Panthera pardus
)
and the African
elephant (
Loxodonta a
fricana
)
.
Leopards are protected animals and listed as “
N
ear
T
hreatened” by the IUCN
(International Union for Conservation of Nature). However
,
the conservation of leopards
outside of protected a
reas in Namibia is not assured. Their “problem predator” image
and
high trophy value,
together with
habitat loss, habitat fragmentation and local outbreaks of
wildlife diseases
,
are the main threats. These threats and the lack of scientific data on this
sp
ecies living on commercial farmland
demonstrate
the need for research.
Elephants are
an incre
asingly desired game species on
farms
which host tourists regularly
due to their
high attractiveness for visitors. However, the impact of elephants on confined are
as and
the comprised ecosystem are rarely studied and more information is needed in order to
create appropriate management guidelines.
This study focussed on the spatial ecology and prey preferences of leopards on Namibian
farmland.
Invasive
as well as
no
n
-
invasive
methods were used;
invasive
methods included
trapping and collaring of leopards, whilst
non
-
invasive
methods included camera traps,
track
counts
, search for prey remains
and faeces collection.
Regarding elephants,
methods such as tracking of the
herd as well as directly observing feeding behaviour were
carried out.
Data collected
on Okambara
show
ed
differences in the ecology of leopards living on
farmland and in protected
areas. Home range sizes
differ
ed
between the study
sites and
were bigger
t
han those of
leop
ards living in protected areas,
most likely due to habitat
preferences
, variation in prey availability
and lower predator densities compared to
protected areas.
The camera trap surveys
on Okambara
yielded a density of 1.9
individuals per
100
km
2
,
a
lower density
compared
to
protected areas
, thereby confirming
the assumption that home
range size is related to density.
The
camera trap surveys also revealed
the existence of
additional carnivores and related
interspecific behaviour showing
tha
t they
seem to
avoid
each other and there
by reduce
direct competition and conflict.
Despite and indeed perhaps
because of this, d
ifferent
strands of
evidence
show
that the habitat on Okambara is
suitable for
the
survival and
reproduction of different preda
tor species.
A study on elephant feeding ecology was also initiated and confirmed the importance of
even
a small herd of nine individuals
as significant ecosystem engineers. This study has
important implications for the increasing trend of stocking Namibi
an game farms with
elephants and as such should be continued and expanded.
3
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Z
usammenfassung
Im Jahr 2012
startete
dieses
Forschungsprojekt
auf
der Okambara Elephant Lodge
,
einer
Wildtierfarm etwa 85 km südlich von
Windhoek
s internationalem Flughafen
,
in
der
Khomas
Hochland
Region
in
Zentral
-
Namibia
.
Okambara ist von einem Wildtierzaun umgeben und
deckt ein Gebiet von 150 km² ab.
Dieser
Bericht befasst sich
mit Untersuchungen, die dort im
Zeitraum
August
-
November 201
3
durchgeführt
wurden
. Im
Foku
s der
Studie standen, der
Leopard (
Panthera pardus
)
sowie der A
frikanische
Elefant
(
Loxodonta
africana
)
.
Der Leopard ist eine geschützte Art und als "potenziell gefährdet" von der IUCN (International
Union for Conserv
ation of Nature) eingestuft. Jedoch kommt ei
n
Großteil der namibischen
Leopardenpopulation auf kommerziell genutztem Farmland
und
somit
außerhalb geschützter
Gebiete
vor
. D
adurch ist die Erhaltung dieser Art in Namibia nicht gesichert. Ihr
Status als
"Problem
-
Beutegreifer
", ein
hoher Trophäenwert,
f
ortschreitender
Verlust von Lebensraum
und
Wildtierkrankheiten sind
ihre
stärksten Bedrohungen. Diese Bedrohungen
sowie
der
Mangel an wissenschaftlichen Daten
machen es sinnvoll und notwendig, diese Spezies im
Lebensraum Farmland besser zu erforschen.
Die
Haltung von Elefant
en wird insbesondere
von Touristenfarmen stetig mehr angestrebt, die diese Tiere eine hohe Anziehungskraft auf
Besucher besitzen. Jedoch ist bisher unklar, welche Auswirkungen die Haltung von Elefanten
in einem eingezäunten Gebiet auf da
s dort beherbergte Ökosystem hat und zusätzliche
Informationen sind notwendig, um artgerechte und langfristig funktionierende Richtlinien zu
entwerfen.
In dieser
Studie standen die räumliche Ökologie von Leoparden auf nambianischen Farmland
sowie der
en
Be
utepräferenzen im Mittelpunkt.
Sowohl
invasive
als auch
non
-
invasive
Methoden wurden angewandt;
invasive
Methoden beinhalteten den Fang und die
Besenderung von Leoparden, während
non
-
invasive
Methoden die Nutzung von Kamerafallen
und die Suche nach Kot, Sp
uren und Überresten von Beutetieren umfassten.
Um Daten
bezüglich der Elefanten aufzunehmen, wurden Methoden wie das
aktive
Aufspüren der Herde
und direkte Beobachtung der Fressgewohnheiten angewendet.
Die auf Oka
mbara aufgenommenen Daten zeigten, dass si
ch die räumliche Ökologie von
Leoparden auf namibianischen Farmland von der in geschützten Gebieten vorkommenden
Leoparden unterscheidet. Streifgebiete waren größer als in Schutzgebieten und ist vermutlich
auf Habitatpräferenzen, variierende Beutetierdicht
e s
owie geringere
Beutegreifer
dichte
im
Vergleich zu geschützten Gebieten zurückzuführen.
Der
Kamerafallenstudie
zufolge weist Okambara
eine
Leopardendichte von 1,9
Tieren
pro
100
km
2
auf
.
Die Leopardendichte auf Farmland ist somit geringer als in geschüt
zten Gebieten und
unterstützt die Vermutung eines Zusammenhangs von Streifgebietsgrößen mit
vorkommender
Dichte. Weiterhin zeigte der Einsatz von Kamerafallen das Vorkommen weiterer
Beutegreifer
auf Okambara sowie
damit verbundenes
interspefizisches Verhal
ten, da sich die
verschiedenen Arten in Raum und Zeit zu meiden scheinen, um Konflikte zu vermeiden
. Damit
ist
Okambara is
t ein geeignetes Habitat für d
en Fortpflanzung und Bestand
verschiedener
Beutegreifer.
Eine ernährungsök
o
logische Studie von Elefante
n wurde
weitergeführt
und bestätigte
den
wichtigen Einflu
ss
selbst der
kleine
n Okambara
-
Herde von neu
n Tieren
auf das
Gesamtökosystem. In Namibia existiert eine zunehm
en
de Tendenz, Wildtierfarmen mit
Elefanten zu bestücken. Die angesto
ß
ene Studie liefert h
ierfür wichtige Rückschlüsse und
sollte
fortgesetzt
werden.
4
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Contents
Abstract
2
Zusammenfassung
3
Contents
4
1. Expedition r
eview
5
1.1. Back
groun
d
5
1.2. Research a
rea
6
1.3. Dates
7
1.4. Local conditions & s
upport
7
1.5.
E
xpedition
s
cie
ntist
s
8
1.6. Expedition l
eader
9
1.7. Expedition
t
eam
9
1.8
. Expedition b
udget
1
0
1.9. Ack
nowledgements
1
1
1.1
0
. Further i
nfo
rmation & e
nquiries
1
1
2.
African leopard ecology on
a
Namibian game farm
1
2
2.1. Introduction
and background
1
2
2.2.
Study site and training of expedition participants
1
5
2.3.
Study
animal
1
5
2.4.
Methods
1
7
2.5.
Results
2
3
2.6
. Discussion and conclusion
s
3
5
2.
7
.
Literature
cited
4
4
3. Elephant study
5
1
3.1. Introduction
51
3.2. Methods
5
1
3
.3. Results
5
1
3.4. Discussion
5
4
Appendix I: Studies reporting mean home range sizes and densities
5
5
Appendix II: Expedition diaries
& reports
5
6
5
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Please note: Each expedition report is written as a stand
-
alone document that can be read
with
out having to refer back to previous reports. As such, much of this section, which
remains valid and relevant, is a repetition from previous reports, copied here to provide the
reader with an uninterrupted flow of argument and rationale.
1.
Expedition
r
ev
iew
Matthias Hammer
Biosphere Expedition
s
1.1. Background
Biosphere Expeditions runs wildlife conservation research expeditions to all corners of the
Earth. Our projects are not tours, photographic safaris or excursions, but genuine research
expeditions
placing ordinary people with no research experience alongside scientists who
are at the forefront of conservation work. Our expeditions are open to all and there are no
special skills (scientific or otherwise) required to join. Our expedition team members
are
people from all walks of life, of all ages, looking for an adventure with a conscience and a
sense of purpose. More information about Biosphere Expeditions and its research
expeditions can be found at
www.biosphere
-
expeditions.org
.
This expedition report deals with an expedition to Namibia that ran from
4 August to 29
November
2013
. The expedition was part of a long
-
term research project and assisted the
local scientist in ascertaining the status of t
he African leopard (
Panthera pardus
) living in
parts of mountainous game farmland in the Khomas region of Namibia. The expedition
s
emphases were on capture activities, GPS
-
tracking, searching for leopard
signs such as
counting tracks and collecting scats,
identifying individuals with the help of camera trap
surveys, and on recording prey animals by hide
-
based observationsat water points and on
game study drives. Additionally, a herd of
African
elephants
(
Loxodonta africana
)
was
observed daily to obtain
in
formation
about their feeding and social behaviour
within the
confines of the fenced farm area study site
.
Namibia is one of a few African countries that support six species of large carnivores.
Lions, spotted hyaenas and wild dogs are mainly restricted t
o protected areas, but
cheetahs, leopards and brown hyaenas still occur on areas with intensive livestock and/or
game farming. The leopard is currently not listed as an IUCN endangered species in
Namibia. However, we believe that high trophy take
-
off toget
her with “problem predator”
reduction
,
combined with habitat loss and fragmentation
,
may put the local leopard
population under threat. There is thus an urgent need to gain a better scientific insight into
both leopard demographics
and
ecology outside prot
ected areas in Namibia.
A good knowledge of leopard ecology on Namibian game farmland will help to conserve
and protect the predator.
In 2011, the Ministry of Environment and Tourism conducted a
leopard population density estimate (national estimate: 14,1
54 leopards)
throughout
Namibia
on which the hunting quota for leopards was based (250 individuals per annum)
(Stein & Aschenborn 2012)
.
However, the
removal through human
-
wildlife
conflict is poorly
monitored
and currently no reliable numbers are accessib
le.
6
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
1.2. Research area
At 825,418 km
2
Namibia is the world's
34th
largest country
(
F
igure 1.2a)
. However, after
Mongolia,
it
is the
second
least densely populated country in the world (2.5 inhabitants per
km
2
). About 40% of the total area in Namibia is
used for commercial livestock farming,
while
communal areas comprise another 40%
and
national parks and restricted areas
mak
e
up the remaining 20%
(
Berry 1990
)
. It is estimated that commercial farmland hosts
about 80% of the commercially useable larger gam
e species (
Brown 1992
) and also
represents most important habitat types.
Flag and location of Namibia and study site.
An overview of Biosphere Expeditions’ research sites,
assembly points, base camp and office locations is at
Google Maps
.
Figure 1.2a.
Map and flag of
Namibia and
location of
study site
.
The study area was cent
red on Okambara
G
ame Reserve in the Khomas region very
close to the Omaheke region in the east
(
F
igure 1.2b)
. The Khomas region spans 36,804
km²
(
F
igure 1.2b
;
Mendelsohn 2009
)
and, due to the inclusion of Windhoek, Namibia’s
capital, has the highest human
population of any region in Namibia.
Figure 1.2b
.
Regional government areas and study site
(red dot)
in Namibia
.
7
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
1.3. Dates
The expedition ran
from August to November 2013, split into
seven two
-
week groups:
4 August
16 August | 18
30
Augu
st | 8
20
Septem
ber | 22
September
4 October |
13 October
25 October | 27 October
8
November
| 17
29
November
2013
All groups
were composed of a team of international research assistants, guides, support
personnel and an expedition leader (see below fo
r team details).
1.4. Local conditions & s
upport
Expedition base
The expedition team was based at
the Okambara Bush Camp
on the Okambara
Game
Reserve
, about 85
km southeast
of
Windhoek’s
international Hosea Kutako
airport
,
in the
Khomas region
. The
camp
(
S
22.69227
,
E
18.21029
) was situated
in the southern part
of
the
Reserve
.
Team members stayed in
chalets
equipped with beds, mosquito n
ets,
basic f
urniture
and
en
-
suit
e
bathrooms. Breakfast and all meals were prepared by the expedition cooks, who
could
cater for vegetarians and
some other special diets. Chalets
had 220V mains
electricity from European style sockets.
There was also a communal building called
lapa
with a dining room, rest areas with sofas, and a fireplace with a view of a waterhole.
Weath
er
The climate is
semi
-
arid savannah
type
with three distinct seasons
. The
hot, dry season
runs
from September to December when t
emperatures
can reach
40ºC
or more
during the
day and
plummet
at night
,
sometimes
to levels
below zero
. Second is a hot, wet s
eason
f
rom January to April and third
is a cold, dry season from May to August with warm days
,
which are contrasted by very cold nights, when temperatures often drop to below freezing.
The expedition
started at the end
of winter in
August
2013
. Annual rain
fall
was
highly
variable
, but in general rainfall in 2013 was very low compared to previous years
.
Average
daily temperatures duri
ng the expedition ranged from 19
to 36.7
º
C
.
Field communications
There wa
s good mobile coverage around the camp but no
cove
rage
in
the mountains.
Regular expedition diary updates were uploaded to
the
Biosphere Expeditions blog
,
Facebook
and
Google+
for friends and family to access.
Transport & vehicles
Team members made their own way to the Windhoek assembly point. From there onwards
and back to the assembly point all transport and vehicl
es were provided for the expedition
team, for expedition support and for emergency evacuations.
8
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Cars used during the expedition were Toyota Landcruisers, provided by Christian Schmitt,
the owner of Okambara
G
ame
R
eserve.
Team members wishing to drive the
cars
had to
be older than 21, have a full clean driving licen
c
e and a new style EU or equivalent credit
-
card sized driving licen
c
e document. Off
-
road driving and safety training was part of the
expedition.
Medical support and insurance
The expedition l
e
ader was a trained first aider
and the expedition carried a comprehensive
medical kit. Namibia’s healthcare system is of an excellent standard and the nearest
doctor and hospital were in Windhoek. All team members were required to carry adequate
travel ins
urance covering emergency medic
al evacuation and repatriation
,
and emergency
procedures were in place, but did not have to be invoked
.
There were
some stomach
upsets, but
no serious medical incidents during the
expedition.
1.5.
E
xpedition
s
cientist
s
Vera Meng
es, born and educated in Germany, joined Biosphere Expeditions
in 2013
. After
spending a couple of years abroad (UK & New Zealand), she graduated from the
Westphalian Wilhelms
-
University
Mue
nster in Germany with a Bachelor’s Degree in
Biology and from Edin
burgh Napier University in Scotland with a Master’s Degree in
Conservation and
Management of Protected Areas.
The latter was based on research
of
brown bears in Sweden in collaboration with the Scandinavian Brown Bear Research
Project. Since then,
she has
worked for this
bear project as well as for a lynx/roe deer
research project in the Bavarian Forest National Park, Germany
.
Now s
he is putting her
skills and passion for wildlife research and conservation towards pursuing a PhD
on
leopard ecology within th
e spatial ecology working group
of the Leibniz Institute for Zoo
and Wildlife Research in Berlin
,
as well as mitigating
the local
human
wildlife conflict by
working
o
n
the big cat and elephant project in Namibia
.
Jörg Melzheimer is a keen biologist and co
nservationist and runs different projects in
Namibia. He was raised in the German countryside and developed his interest in nature
early. He studied spatial ecology and conservation management at the University of
Potsdam (Germany), Universidade Federal de
Santa Catarina (Brazil), University of the
Witwatersrand (South Africa) and the Free University of Berlin. Currently his main research
focus is a cheetah research project of the Leibniz Institute for Zoo and Wildlife Research in
Berlin, where he heads the
spatial ecology working group, coordinates the project’s field
work and acts as its PR and liaison manager, responsible for stakeholder involvement and
media work. On the field science side he is involved in research on spatial ecology of
cheetahs, leopar
ds, wild dogs, brown hyaenas, kudus, oryx, jackals and bat
-
eared foxes.
Jörg also chairs the management boards of two conservancies and is the talks & event
coordinator of the Namibian Environmental and Wildlife Society (NEWS).
9
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
1.6. Expedition l
eader
Al
isa Clickenger was born in the United States and educated at Bennington College in
Vermont. After many successful years in the corporate world, she fell in love with the path
less travelled. She now lives a life of travel and adventure, and writes about it
for several
magazines. An experienced overlander on two and four wheels, Alisa
has a
love of nature
and foreign cultures
which in 2009
brought her on a seven
-
month solo journey through
Central and South America seeking wildlife and wild places. An experie
nced tour guide in
the adventure travel field, at Biosphere Expeditions Alisa realises a dream
that of
combining her love of people with her love of wildlife and conservation.
1.7. Expedition t
eam
The expedition team was recruited by Biosphere Expediti
ons and consisted of a mixture of
all ages, nationalities and backgrounds. They were (with countries of residence):
4
16 August 2013
:
Karin Kindbom (Sweden), Renate Kupke (Germany), Shelagh
Macmillan (UK), John Munthe (Sweden), Martin Quigley (USA), Jan
e Quigley (USA),
Virinia Renaud (Switzerland), Paul Schneider (USA), Philip Worden (UK), Martina Zeuner
(Germany).
18
30
Augu
st 2013:
Susanne Ahlquist (Norway), Dianne Aitken (Australia), Barbara Allen
(Australia), Theresa Bowman (Germany), H C Connor (
USA), Andrew Coogan (UK),
journalist Matthew Havercroft (UK), Beate Hinterreither (Austria), Shelagh Macmillan (UK),
John Rawnsley (UK), Gary Schiavi (USA), Rose Tapp (Australia), Claire Waring (UK).
8
20
Septem
ber 2013:
Sandra Bartenbach (The Netherlan
ds), Helen Cory (Australia),
Frank Hahm (Germany), Sandra Hogben (UK), Gary Hogben (UK), Gabriele Koßmann
(Germany), Anand Nadathur (Singapore), Piotr Piesik (Poland), Suresh Rajagopalan
(India), Brigitte Schuberth (Germany), Anna Zagorowska
-
Piesik (Poland
).
22
September
4 October 2013:
Joan Arbuthnot (USA), Susie Barrett (UK), Louise Barton
(Australia), Amanda Chao (USA), Peter Gorr (USA), Susan Gorr (USA), Renate Hall
(USA), Louize Hermitage
-
Holt (UK), Gabriele Koßmann (Germany), Jackie Saxon (USA),
Da
ve Stamm (USA).
13
25 October 2013:
Mary Alford (UAE), Geoff Badham (Australia), Greg Deming (USA),
Julia Johnson (UAE), Don Macpherson (Australia), Peter Martin (Australia), Kim
McCormack (Australia), Eric Swenson (USA), Jill Swenson (USA), Verena Thue
rey (The
Netherlands), Stefan Thuerey (The Netherlands), Ngoc anh Tran (France).
27 October
8
November
2013:
Kate Bass (UK), Nicole Berthier (Switzerland), Jim
Blomgren (USA), Eliza Dlugolecka (UK), John Haddon (UK), Gabriele Krimpmann
(Germany), Christ
ine Marklow (UK), Sigrun Metz (Germany), Jairun Naisha (USA),
Andrew Porritt (UK), Ngoc anh Tran
(France), Nancy Tran (USA).
17
29
November
2013:
Valerie Boquet (USA), Wayne Curley (USA), Barbara Felitti
(USA), Eva Jung (Germany), Charlene Kalin (USA),
Stephanie Kappus (Switzerland),
Seema Mathew (China), Carole Metour (USA), Morgan Pegg (UK), Ann Setser (USA),
Sheila Tolley (UK), Ritva Viitala (Finland).
10
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
1.8. Expedition
b
udget
Each team member paid towards expedit
ion costs a contribution of £
1,750
per
two
-
week
slot. The contribution covered accommodation and meals, supervision and induction, all
maps and special non
-
personal equipment,
and
all transport from and to the team
assembly point. It did not cover excess luggage charges, travel insurance, perso
nal
expenses
such as
telephone bills, souvenirs, etc.,
or
visa and other travel expenses to and
from the assembly point (e.g. international flights).
Details on how these contributions were
spent are given below.
Income
£
Expedition contributions
1
45,098
Expenditure
Staff
includes local & international salaries, travel and expenses
1
9,235
Research
includes
scientific services,
equipment, animal capture and other research
expenses
1
2,769
Transport
i
ncludes
car hire,
bus transfers,
fuel
& maintenance
5
,535
Bas
e
includes board, lodging and other base camp services
57,435
Administration
includes office costs
,
visa & professional fees and miscellaneous costs
2
,231
Team recruitment
Namibia
as estimated % of PR costs for Biosphere Expeditions
4
,472
Income
Expendit
ure
4
3,421
Total percentage spent directly on project
7
0%
11
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
1.9. Acknowledgements
This study was conducted by Biosphere Expeditions
,
which runs wildlife conservation
expeditions all over the globe. Without our expedition team members (listed above) who
provided an expedition contribution and gave up their spare time to work as research
assistants, none of this research would have been possible. The support team and staff
(also mentioned above) were central to making it all work on the ground. Thank you t
o all
of you and the ones we have not managed to mention by name (you know who you are)
for making it all come true.
Biosphere Expeditions would
also like to thank
the Friends of
Biosphere Expeditions for their sponsorship and/or in
-
kind support.
VM
would
like to thank the Namibian Government, the Namibian Tourism Board and the
Ministry of Environment and Tourism in particular, for giving me the permission to conduct
this study. My thanks also
go
to all expedition team members as well as staff members for
their amazing effort. The
expeditions in 2013
made a major contribution to the
research
on
Okambara.
My thanks also go
to Swaro
v
ski
Optik for providing
binoculars and range
finders
.
I thank the Institute for Zoo and Wildlife Research in Germany for scienti
fic advice
,
help with handling and immobilisation of animals
and analysing blood samples.
Special
thanks go
to
Uschi and Christian Schmitt
, for giving me permission to run the expedition on
their property
and
for their cooperation
and allowing me to live o
n Okambara
,
and t
o the
lodge manager
Bea
.
Also
,
I would like to
thank Alisa Clickenger for running the expedition
on the ground
and her support on numerous occasions. I
thank Jörg Melzheimer, Matthias
Hammer and the other reviewers for their comments on va
rious versions of this
manuscript.
Last but not least, I would like to thank Biosphere Expeditions and every team
member for the contribution that this expedition has made to large
carnivore conservation
in Namibia
.
1.10. Further information & e
nquiries
More background information on Biosphere Expeditions in general and on this expedition
in particular including pictures, diary excerpts and a copy of this report can be found on the
Biosphere Expeditions website
www.biosphere
-
expeditions.org
.
Copies of this and other expedition reports can
be accessed via
www.biosphere
-
expeditions.org/reports
.
Enquires should be addressed to Biosphere Expediti
ons
via
www.biosphere
-
expeditions.org/offices
.
12
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Please note: Each expedition report is written as a stand
-
alone document that can be read
without having to refer back to previous reports. A
s such, much of this section, which
remains valid and relevant, is a repetition from previous reports, copied here to provide the
reader with an uninterrupted flow of argument and rationale.
2. African leopard
ecology
on
a
Namibian game farm
Vera Menges
Biosphere Expedition
s
2.1. Introduction and background
Given the steady decline in biodiversity, it is of increasing importance to connect nature
conservation with science
-
based management (
Lawler et al. 2006, Soulé & Orians 2001
).
Research
not only
serv
es to broaden scientific knowledge, but is
also
essential
for
predicting the success of
management plan
s
for
species.
This is especially true for large
carnivores as they are at the top of the food chain in terrestrial ecosystems
,
but at the
same time repr
esent
its
most vulnerable elements
(
Schipper et al. 2008
). Studies indicate
that ca
rnivores play an essential role
as they structure
as well as preserve the existing
biodiversity through
their prey choice
(
Miller et al. 2001
). The elimination of carnivores
can
lead
to a
chain of negative consequences
, starting with the demographic exp
losion of
herbivores and meso
-
carnivores
and
unsustainable
grazing
pressure, leading to
biodiversity loss at all levels of the food chain
,
and
it
can even
result
in a collapse
of the
ecosystem
(
Estes
& Duggins 1995
, Henke & Bryant 1999
). It is therefore crucial to protect
apex predators
,
in order to
pres
erve biodiversity
as well a
s the ecosystems that host them
.
Research
on
carnivores is usually
not very
practice
-
oriented and m
anagement guidelines
are often intuitive and subject to trial
-
and
-
error methods
,
rather than relying on scientific
facts (
Ray et al. 2005
). However, species
-
specific knowledge of ecology and biology of a
species
is required
for the successful implementatio
n of wildlife
conservation
and
management (
Frankham et al. 2002
).
T
hus the probability of success of the
applied
methods
increases
and important resources
, such as time and finances, are
used more
effectively.
In Africa,
human
wildlife
conflicts
are
among
the three
main
threats to
the
existing biodiversity;
in
particular
for large cats such as
the l
eopard
(
Panthera pardus
)
(
Nowell & Jackson 1996, Ray et al. 2005, Treves & Karanth 2003, Woodroffe 2000
).
Several studies on
leopards (
Panthera pardus
)
exist al
ready
, but
they were usually carried
out
in protected area
s such as Kruger, Serengeti and Etosha National Park
s
(
Bertram
1982, Bailey 1993, Stander 1997
, Durant 1998,
Mizutani 1999, Ray et al. 2005
)
. However
,
the majority of leopards
in
Namibia
occur
on
co
mmercial
farmland.
There, t
he local
farmers are often accused of
persecuting
big c
ats to protect their livestock. Such
behavioural patterns are
primarily due to
the
absence
of
basic strategies to avoid conflicts
with these animals in
the first place
(
Linne
ll et al. 2001, Marker et al. 2003
).
Namibian
farmers
are organis
ed
locally
in
to
so
-
called
conservancies in
which they develop and
agree on
management guidelines. Since there is often a lack of information on the ecology
and biology of the big cats, the
se
management guidelines are
often
neither sustainable
,
nor
do they
solve problems
comprehensively.
M
ost
of the local farmers
are engaged in
breeding
cattle
and also
use the locally abundant
wildlife for their own consumption
as well
13
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
as for
trophy hunting.
L
o
sses
due to predators
are
reported regularly; some
farmers have
particularly
high
losses
,
which
put them
under high economic pressure
,
and
are
therefore
intolerant of b
ig cats and other predators
(
Shwiff & Sterner 2002, Hughey et al. 2003
).
There are no de
tailed studies on the prey preferences
of leopards
outside
protected
areas.
Farmers often assume that leopards
specialise in preying on domestic livestock and take
calves, sheep, goats and poultry as easy prey
.
P
ersecution
of leopards
and their
exterminati
on
on farmland with methods
such as hunting at night with torches, the use of
dogs to chase the cats or shooting them in a box trap is putting the local leopard
population under threat.
To understand the ecological factors that determine demographic trend
s in carnivores, it is
important to study free
-
ranging populations under natural selection pressure. As most
parts of Namibia are under some sort of agricultural management, which very often entails
removal of problem animals, the selection pressures inclu
de human factors. Demographic
parameters such as fecundity, mortality, reproductive success, sex ratio, age structure and
social structure
can
therefore differ from populations in protected areas. These
demographic parameters are key elements to estimate l
ong
-
term viability of populations,
and population viability models need to be fed with high
-
quality data as the output of these
models is extremely sensitive to the input. Information on leopards on commercial
farmlands is scarce and very often preliminary
data are used.
Large carnivores are particularly difficult to study, as they range widely, occur at low
densities, capture probabilities
vary
between different individuals, and they are often
secretive or elusive (
Karanth 1995, Boulanger et al. 2004
).
Le
opards in protected areas,
for example in national parks, are habituated to humans
. Therefore
extended periods of
observation are possible. However, leopards living on commercial farmland generall
y
avoid encounters with humans.
To obtain high
-
quality data,
indirect sampling methods are
required. Fitting individual animals with GPS collars is a su
itable method to study solitary
and
elusive
mammals
in their habitats (
Seidenst
icker et al. 1970, Bailey 1974
) as the data
obtained provide information on home rang
e sizes, movement patterns
and
habitat use.
Information gleaned thus can be incorporated into farm management and may help to
keep financial losses to a minimum, which in turn makes cooperation by stakeholders
more likely.
Also, m
onitoring the abundance a
nd distribution of animals is fundamental to the research,
management and conservation of wildlife populations.
Estimating animal numbers is often
a basic requirement for
determining the status of species. However, this task is deceptively
simple and no si
ngle best approach exists; techniques that work
well in some situ
ations
are useless in others (
Caughley and Sinclair 1994
)
. Many terrestrial
mammals
such as the
leopard
are nocturnal, cryptic in appearance, and
, in the case of leopards on farmland,
general
ly
adept at avoiding being seen, which limits well
-
developed methods
of direct
observation
(
Duckworth 1998, Chiarello 2000, Lopés and Ferrari 2000, Jachmann 2001
)
.
These
challenges leave indirect observation, for example via animal
tracks or remote
photogr
aphy, as often the only realistic option.
Photographic capture of individual leopards,
together with information on date, time and capture location, can provide baseline data for
population density analyses (
Karanth et al. 2004
). Photos obtained can be use
d to identify
individual animals and add valuable information towards population density est
imates and
population dynamics.
In
general, recordings of predator tracks are designed to provide
presence/absence data only, but by following tracks of foraging ca
ts, a wide range of
additional data about behaviour such as prey
-
encounter frequencies, hunting success,
14
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
prey species selection, home range use and social interactions can be gathered (
Stander
et al. 1997
). Scats of predators add another piece of evidence
of predator occurrence. The
hair of prey is relatively
indigestible
and
undamaged
in most carnivore scat and can thus
be used to identify the prey species eaten (
Wachter et al. 2006
). Scat analysis is used to
understand the prey preferences of leopards and
obtain insights into predation habits, thus
showing
if diet overlap and potential competition among carnivores and even smaller prey
occurs.
In addition,
prey preferences
can further be
evaluated
through
GPS
cluster analysis
. This
is a
fairly
new method
to detect
potential kill sites
of carnivores and has been applied only
in a few studies (
Krofel et al. 2012, Pitman et al. 2012, Fröhlich et al. 2012
)
.
Leopards
revisit kill sites for up to several days in order to fully consume their prey, thereby leading
to
a specific pattern of GPS locations ("clusters"
). Remains found at these locations can be
used to identify prey species and therefore provide information on individual prey
preferences of leopards.
Such findings are very important to demonstrate predat
or dietary
preferences and thus enable game ranchers to manage predators on their land.
The abundance and density of prey species are influencing factors on predator
occurrence
,
densities
and prey preferences
and therefore need to be investigated as well.
In addition, t
he management of game
species on game farms is an
important factor in
securing income. Where wild ungulates are utilised by people for either consumptive
purposes
(commercial hunting and game farming) or non
-
consumptive
purposes
(safari
tour
ism), competition and conflict may occur between game ranchers and large predators.
With the advent of game ranching, game prices for most species have increased by more
than 50% over the last 20 years. Many game farms are stocking up with rare and valuabl
e
species such as roan (
Hippotragus equinus
) and sable (
Hippotragus niger
) antelope,
resulting in a large increase in the antelope value over recent years. The typical game
farm is fenced to keep the valuable game species on the property of the owner.
Hist
orically, game migrated perennially from one grazing ground to another. This gave the
grass time to regrow, bloom and reproduce. Fences hinder these dynamics and game
farms run the risk of severe degradation and desertification due to overgrazing.
Manageme
nt therefore becomes crucial in fenced
-
in area
s
and many pieces of information
are needed for successful management,
such as
game density, reproduction rate, primary
production and sustainable stocking rates.
To determine the status of the leopard populat
ion in the study area, the dynamics and
abundance of the leopard population and prey species
need to be ascertained. The
basic
questions that the study focused on were: What is the behaviour and ecology of leopards
living on commercial farmland, particular
ly game farms? Are there any differences to
leopards found in protected areas and national parks?
What is
the local
prey availability
and abundance?
15
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.2. Study site and training of
expedition participants
Okambara
is situated
85
km
south
east
of
Windhoek’
s
Hosea Kutako
international airport
(F
igure 1
.2
a
). The farm is 15
0
km² in size and entirely
surrounded
by a game
fence
(height
4
m)
(F
igure 2.3a
)
. All internal fences have been removed, thus allowing free roaming of
wildlife
(in F
igure 2.3a
turquoise
line
s inside the
study area illustrate
former fence
lines
as
the
study area eight years ago consisted of three
different farms
“Frank” in the south,
“Bildah” in the
centre
and “Okambara” in the northwest)
. The
study site
has
a variety of
landscape
s
(altitude
s
range from 1
,
500
to
2
,
000 m) with many different habitat types
ranging from typica
l
African bushveld to
mountainous
areas
, and it
contain
s
ideal habitats
for all of Namibia’s indigenous mammal species, including elephant and rhino.
Fairly
evenly distribu
ted over the study area
are
nine
dams (man
-
made lakes), which contain
water year
-
round
.
Other dams
are relatively small and only keep water for a few months
after the
rainy season. The area has for many years
not been used for any commercial
farming activi
ty, thus leaving the pasture and bush in
good
condition. The
expedition base
camp site (S
22.44308
,
E
16.96900)
is
situated close to
a man
-
made waterhole called
Gustavposten
.
Okambara
is a
good
ar
e
a
in which
to study
leopard
ecology in a game farm
setting.
Although the
study area
is fenced
in
, the movements of leopards and other felids are not
confined as cats (
as well as other predators
and smaller herbivores) can easily
pass
underneath the fences.
For the first two days
of each
two
-
week
group
,
expeditio
n participants
were given talks and
practical lessons, learning the use of GPS
, compass,
range finder and other research
equipment
and safety techniques, skills and procedures.
F
irst excursions into the field
were under the supervision of Biosphere Expedit
ions staff. After a few days,
participants
were
able to navigate around
the study site, install camera tr
aps,
rec
ord tracks and signs
of mammals and
identify animals. Where necessary
,
research teams were accompanied
by trained local
staff
t
o improve the ac
curacy of data recording
or to provide a safe working
environment. Data entry and p
icture downloads were tasks performed at the expedition
base
.
2.3. Study animal
The leopard
(Panthera pardus)
was the key study species
.
It has the greatest geographic
dis
tribution of all the
big
cats (
Nowell and Jackson 1996
), covering a variety of different
habitats ranging from desert to rainforest. Density varies with habitat, prey availability and
intensity of per
secution, from
below
one
individual
to over 30
per
100 k
m², with
the
highest
densities
recorded
in protected east
ern
and southern African environments (
Hunter 2011
).
Nevertheless, the leopard is listed
o
n Appendix I
of
CITES and
is
classified as Near
Threatened (
IUCN 201
3
)
, with nine genetically distinct subspe
cies. Currently wild cats
such as leopards, cheetahs and caracals are not listed in the Endangered category (
IUCN
20
13
)
al
though excessive trophy hunting
combined with a high “problem predator” take
-
off,
and other factors such as habitat loss, fragmentatio
n and local outbreaks of wildlife
diseases, may potentially put the leopard (and the other predator species) under threat
locally
(
Berry 1990,
Bailey 1993
).
16
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.
3
a
.
Okambara
Game Reserve
consists of three former farms as shown by the red lines
that surround farm
roads (yellow).
The outer red line perimeter is an electrified game fence
;
inner fences (turquoise) have been removed.
17
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Leopards are
solitary, nocturnal carnivores
with spacious home ranges, but only occur in
low densities (
Spong et al.
2000
).
Both sexes are territorial and defensive against
adult
conspecifics of the same sex
;
they mark their territory with scent, faeces and scratch
marks
(
Hamilton 1976, Bailey 1993
)
.
Leopards are very good climbers; they often hide
their prey in trees t
o avoid scavengers.
Their hunting strategy consists of
stalking and
pouncing
;
thus
they do not chase their prey over long distances
(
Bailey 1993, Stander et.
al 1997
)
.
Leopards have a vast range of prey;
Bailey (1993)
noted at least 92 prey species
used by
leopards in sub
-
Saharan Africa,
varying from species as small as the dung beetle
up
to large mammals
such as
adult male eland
antelope
s (
Kingdon 1977
). Yet despite this
apparent ability
to
successfully
exploit prey spanning such an enormous size range,
th
e
leopard
’s
diet is generally dominated by medium
-
sized ungulates (e.g.
Bailey 1993
). A
recent analysis of 33 studies on leopard feeding ecology revealed that leopards
preferentially prey upon species within a weight range of 10
40 kg, even if prey outside
this weight range is more abundant (
Hayward et al. 2006
). The optimum prey weight for
leopards derived from this analysis is 23 kg, based on body mass estimates of significantly
preferred prey species (
Hayward et al. 2006
).
2.4
.
Methods
2.4.1
.
Capturing
and collaring
Box traps were
baited mainly with antelope as well as
zebra meat
and
were
checked
twice
a day (morning and
late
afternoon). Once a target
animal
, i.e.
a
predator,
was captured, it
was darted and
immobilised
. Drug choice, dosages and combina
tions depended on
the
type of species captured and the body weight. Whilst under anaesthesia, animals were
placed in a shaded
location
and a facial cover and eye lubricants were used to prevent
damage to the eyes. Noise levels were kept to a minimum. Vital
parameters were
monitored and an intravenous line was placed to administer fluids if needed and to have
access to the bloodstream should an emergency arise. ID pictures were made from
both
sides of the animal for usa
ge
in
the camera trap survey (F
igure 2.
4.1a
). Various samples
were taken (a range of blood samples, smear of saliva, nasal and conjunctival fluid, faeces
and body measurements). While working in the field, blood samples were stored
,
chilled
and processed later in the lab
oratory
. The animal’s ag
e was determined based on
tooth
wear and general habit. Only f
ully grown animals were fitted with a GPS collar.
Once the anaesthetic was reversed, the animal was
placed at a location in the shade near
the handling si
t
e and observed from a safe distance
to
ensure complete anaesthetic
recovery
.
Both
e
-
o
bs GPS
-
collars
(Figure 2.4.1
b)
and
Vectronic Aerospace
GSM/GPS
collars
were
used.
These
collar types
provide the GPS p
osition (
based on the coordinate system
WGS84
) of the animal, a fine
-
scale ambient temperature and
an activity
measurement;
the Vectronic
collars
also send a
notification in the event of mortality.
Data collected by the
collars were downloaded
at
regular
intervals via
airplane
telemetry.
The weight of
each
co
llar was less than 3% of the animal
’s body weight.
18
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.4.1a
.
ID picture
of
male leopard
L051, left
side
.
Figure 2.4.1
b.
Adult male leopard
L055
with
e
-
o
bs GPS
collar; monitoring vital pa
rameters.
19
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.4.2
.
Monitoring animals
h
ome range
GPS
(Global Positioning System
)
tele
metry
was used
to monitor the animals’ home
ranges
.
Leopards fitted with collars were located by GPS
;
i.e. the transmitter inside the
collar attempts
within defined
intervals
to contact at least three satellites in order to
determine accurately the ani
mal’s position
.
Telemetry data
were uploaded to
movebank.org
and converted into ESRI shape files and
csv
-
files for further a
nalysis
. Afterwards data were entered into
the statistical program R
and the geographical processing program ESRI ArcGIS 10.1 to calculate and display
home range sizes
. The home range size was calculated using two standard methods: the
minimum convex polyg
on (MCP,
Hayne 1949
) and the kernel method (
Worton 1989
).
Data analysis
The MCP method is one of the earliest and still a widely used method for calculating home
ranges (
Harris et al. 1990
). In this method the peripheral locations of a given data set are
connected so that they form a polygon. The MCP method is very simple and the resulting
home ranges are comparable between studies, but it has several disadvantages. For
example, the home range is highly correlated to the number of locations and it does no
t
give any informat
ion on how the area is used
. Studies on habitat utilisation require more
sophisticated analyses such as the kernel method. Currently th
is
method is considered to
be the most suitable one for home range estimation (
Powell 2000, Worton 199
5
). With
it
a
probability density function from the locations is calculated in order to determine a utility
distribution. Home ranges are then defined by drawing contours around areas with equal
intensity of use. The home range looks like a hilly surface.
However, occasional
exploration trips of an animal may lead to overestimated home range sizes. To correct for
this, a certain percentage of the data
set is excluded as outliers (e.g.
5% of the most
remote points being excluded results in the
Kernel 95
). Fr
om a biological point of view, the
kernel method is much m
ore useful than the MCP method,
but for comparison with
previous studies MCP data needs to be considered too.
2.4
.
3
.
Track
counts and scat collection
Twelve
different routes were planned for track
and scat counts (total 70
km)
(Figure
2.6.
4a)
. Each day, a route was selected randomly. Occasionally expedition team members
needed to reschedule for safety reasons because elephants were utilising that particular
area. GPS positions were recorded for all
leopard, cheetah and hyaena tracks found. Data
such as date, number of animals, sex and age class, age of track (very fresh, fresh, old,
not sure) and track size (pad width, pad height, total width, total length), direction of track,
start and end point o
f the track and further comments were recorded.
All leopard, brown
hyaena and cheetah scats found on the transects were collected. Scats were collected
along the same routes as tracks, and date and GPS coordinates were noted. Scats
collected were air
-
dried
and stored. Leopard scats can be discerned from scats left by
other species by their size, shape, consistency (
Stuart and Stuart 2000
), odour and
adjacent tracks visible. In terms of size, hyaena scats are similar to leopard scats, but
they
are
easy to di
stinguish from them, as hyaena scats are much harder and white due to a
high ratio of calcium residue of digested bones (
Walker 1996
). Additionally, in many cases
tracks were found in association with scats, which ma
de identification more precise.
20
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.4.4
.
Camera traps
Results from the capture
recapture methods can be analysed by the program
CAPTURE
(
Otis et al. 1978, Rexstad & Burnham 1991
).
This
program
offers
different
models
to
calcula
te
population size
.
Two different brands of camera traps (
Bushnell
Trophy Cam 2010 & 2011 and
Reconyx
650) were used during the study. Both were equipped with SD memory car
ds up to 8 GB
(yielding up to 8
,
000 pictures at medium resolution settings). Camera traps were
either
positioned in wildlife hotspots close to natural or man
-
made water sources or scattered
over the study site
,
mostly alongside farm tracks.
The minimum dis
tance between stations
was 700 m and the maximum distance was 15 km.
Camera traps were checked
once
a
week to exchange SD cards, make minor adjustments and verify battery status. Leopard,
brown hyaena and cheetah individuals were identified from the pictur
es taken, as well as a
host of other non
-
target animals (primates,
ungulates, etc.). The fur pattern
of each
individual
leopard and cheetah
is unique and individual animals were identified. Brown
hyaenas have stripes on the front legs as well as scars on t
he face or ears, all of which
can be used to identify individuals.
The program
Camera Base
(Version 1.6
,
Tobler 2010
) was used to organise camera trap
pictures and run analyses, for examp
le via the program
CAPTURE
(
Rexstad and Burnham
1991
), which estimates
leopard abundance.
CAPTURE
offers
different
models
and
identifies
which model fits the data set best and then genera
tes capture statistics for all
models (
Jackson et al. 2006
). The most important statistical requirement to calculate
population size based on mark
recapture data is the assumption that the population is
closed (no immigration, no emigration, no mortality a
nd no birth) during the sampling
period.
To meet this requirement, a sampling period between 30 and 90
days should be
considered, so 9
0 days was chosen for this study. If an animal was photographed it was
noted as an event. In order not to overestimate th
e research area, a buffer needed to be
added
.
To estimate the area effectively sampled (A), a convex polygon connecting the
outermost camera traps plus a buffer area, where width (W) is an estimate of half the
home range length for female leopards in the s
ampled area, was computed following
Karanth and Nichols (2002).
Population density was determined by dividing numbers of
identified leopards (by CAPTURE)
by
the sampled area.
2.4.5
.
GPS c
luster
a
nalysis
Based on
temporal high
-
resol
ution
GPS data of colla
red leopards,
a GPS cluster analysis
was performed
(
Pitman et al.
2012
, Fr
ö
hlich et al.
2012
).
Leopards
feed from
their prey
repeatedly, thus returning to the
carcass
(
i.e. hiding place where the prey is located
)
over
a period of
up to
several days. This c
auses a cluster pattern in the data, meaning
numerous GPS positions (of consecutive days) in the same location
(see Figure 2.4.5)
.
Detected clusters were visited in the field and searched for prey remains such as hair and
bones
,
which were then
used to ide
ntify the prey species.
21
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.4.5
a
.
GPS
positions, movements and cluster (yellow circle) of L051
.
2.4.6
.
Game counts
V
ehicle game counts and waterhole observations were conducted over a four
-
month
period. Of primary interest was population demogra
phic data (e.g. male:female ratios, age
composition of herds, number of sexually mat
ure females with calves, etc.).
Distance
sampling is one of the best methods to estimate wildlife populations accurately (
Buckland
et al. 2008
). For this purpose the study
area was divided into line transects following
Buckland et al. (2008). The area was classified into two easily discernable vegetation
types: dense and open.
Vehicle game counts were conducted on farm tracks. The three transects of between 10
and 15 km
(se
e Figure 2.4.6
)
each were driven along at a very low and relatively constant
speed (about 15
20 km/h) and observers on the back of the vehicle counted all animals
they detected on both sides of the road. All game animals within a 1
,
000 m semi
-
circle (the
a
verage viewing distance on foot) in front of the observers were counted
.
Equipment used
included range
finder, binoculars, angle measurer, clipboard, datasheet, pen and different
African mammal identification field guides. Species, number(s), distance
to t
he vehicle
and
angle of the detected animal(s)
from the transect (vehicle midline)
were recorded, as well
as the GPS position of the observer, plus, if possible, any notes about
the
species
age
and sex
.
22
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.4.6
a
.
The three vehicle game count rout
es on Okambara.
VGC 1 = 10.8 km, VGC 2 = 14.9 km and VGC 3 = 12.7km
.
G
ame species were also recorded
at waterholes
. At the beginning of the study, expedition
participants had to construct several hides at each waterhole so that viewing positions
could be
taken up depending on the wind direction. Observations took place during the
day to study animal behaviour such as duration of stay at the waterhole, and whether
animals were drinking or not. Gender, age class and herd composition were also recorded.
23
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.5
.
Results
2.5.1. Capturing/collaring
The capture campaign started
at
the beginning of
August and continued until
the end of
November 2013. During this period three
adult male leopards were captured
(L051, L052,
L055)
; one of them (L051) was captured in a
box trap that had been set at a freshly found
kill
.
L052 and L055 where captured in baited box traps.
Additionally
,
one adult female
brown hyaena as well as four honey badgers
(
Mellivora capensis
)
(one subadult male, one
subadult female, one adult male an
d one adult female) were caught.
All individuals were
immobili
s
ed
. The
honey badgers
were released after samples and measurements were
collected without fitting them
with a GPS device
. The leopard
s were fitted with
e
-
o
bs
GPS
collars and the brown hyaena wa
s equipped
with a
Vectronics GPS/GSM
collar
. All
individuals captured were in
good condition (Table 2.5.1a
).
During the expedition four box traps were set throughout the study site. Each trap that
was
set count
ed
as one trap night. One night with four arm
ed box traps
was
therefore counted
as four trap nights. During the study per
iod box traps were active on 116
days with a total
of
421
trap nights
(T
able 2.5.1b
).
When checking the traps,
86
%
of box traps were found
open, 9% had captured an animal and 5
% of
the traps that had shut were empty (F
igure
2.5.1
b
). Three
leopards, one
brown hyaena
,
four honey badgers
, ten
warthogs,
14
porcupines
and one gos
hawk were captured
. Traps were set in
eight
different locations;
the highest capture success was close to the
Bergposten
waterhole (BT04
) in the
northeast of the study site (trap
position BT04
on Figure 2.5.1b
)
.
Table 2.5.1a
.
Predator
capture
data 2013
.
Capture
d
ate
Species
Animal
ID
Gender
Estimated
a
ge
(years)
Weight
(kg)
Neck
circ.
(cm)
GPS collar
24.08.2013
Leopard
L051
male
5
69
55
yes
24.09.2013
Honey
badger
C054
male
1
11
33
no
29.09.2013
Honey
badger
C055
female
1
6
27
no
29.09.2013
Honey
badger
C057
female
2.5
6.5
28.5
no
29.09.2013
Brown
hyaena
C056
female
2.5
40.5
50
yes
18.10.2013
Leopard
L052
m
ale
6
67.5
56
yes
26.10.2013
Leopard
L055
male
7
68.5
56.5
yes
02.11.2013
Honey
badger
C061
male
2
12.5
32
no
24
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Table 2.5.1b
.
Trap nights (24h) effort and success 2013.
Trap nights
Open
Closed
but
empty
Capture
421
359
23
38
Figure 2.5.1a
.
Capture
success from 421
capture nights (24h) on Okambara in
2013
.
During the capture campaign traps wer
e moved regularly. In Figure 2.5.1b
, the
letters “A”
or “B
” after
labels of box traps (e.g. BT01)
indicates a new po
sition of a box trap (Figure
2.5.1b
; for e
xample
,
box trap 1, BT01, moved to location box trap BT01A).
25
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.5.1b
.
Map of
Okambara with
position of box traps
in
2013
.
26
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.5.
2
.
Home range size
The MCP of L051 covered an area of 410.45 km²; Kernel 95 consisted of 385.48
km²,
Kernel 90 of 347.
63 km² and Kernel 50, the core area of the home range, was 120.44
km².
L052 had an MCP of 272.09 km² whilst Kernel 95 and Kernel 90 covered an area of
161.75 km² and 150.05 km², respectively. The core area of L052
(Kernel 50)
consisted of
50.02
km². The MC
P of L055 was 142.43 km²; Kernel 95 covered
1
11.06 km² and Kernel
90 entailed 103.27 km². The size of the core area of the home range of L055 was 40.74
km² (see Table 2.5.2).
Table 2.5.2
.
Home range size (km
2
; MCP and Kernel) of the leopards collared on O
kambara (L051, L052, L055)
.
Home range size (km
2
)
L051
L052
L055
DATA
256 days
196 days
160 days
Kernel 50
120.44
50.02
40.74
Kernel 90
347.63
150.05
103.27
Kernel 95
385.48
161.75
111.06
MCP 100
410.45
272.09
142.43
27
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.5.2a
.
Locations
of male leopard
s
L051, L052 and L055
on Okambara
and surrounding farms since collaring
.
28
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Figure 2.5.2b
.
MCPs
of male leopard
s
L051, L052 and L055.
29
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2
.5.
3.
Track
counts and scat collection
Track
and scat
routes were
monitored
between
one
and
five
times
e
ach
and a total length
of 360.3
km was
covered
.
Eight
of the routes
were in the plain
s
area
(2, 5, 6, 7, 8, 9, 10 &
11)
and
five
in mountainous areas
(1, 3,
4
, 12 & 13
). The routes
most frequently
monitored
were number
s
1, 3, 9
and
12
(
five
times).
Number
s
4, 5
and
7
were not
monitored
regularly
(two or three times each
)
,
because elephants
were in those areas
frequently
.
Figure 2.5.3a
shows the probability (p %) per kilometre of predator findings (tracks/scats)
for
leopard, brown hy
aena and cheetah
on diff
er
ent routes (T
racks
&S
cats Route No.
1
13
).
Figure 2.5
.3a
.
Probability p (%
)
/km of predator occurrence on the basis of track and scat findings
on particular routes
. LEO = leopard, BH = brown hyaena, CH = cheetah.
Leopard
signs
(76
)
dominate
d
the results
over
cheetah (11)
.
Hyaena signs
(23
)
were
found
all over the study site, sometimes several times in the same location.
Eighty
-
two
percent
of the cheetah signs
(9)
were found in the plains area.
T
h
e largest
number
of tracks and scats from leopard
s
were fo
und close to
the edge of
the
small mountains in the north and
southwest
of the study site,
where
there is also
water
available
(route
s
1 & 3).
Also, on route 6 and 12 several signs of leopards were found;
both routes include waterholes as well. Few signs o
f leopard occurrence were detected on
routes 2, 4, 5, 10 and 13. No signs of leopards were found on routes 7, 8 and 11, all of
w
hich are situated in the plains (see figures
2.5.3a,
2.5.3b and 2.5.3c).
Signs of
hyaena were found mostly on route
s
2
, 4, 6, 7
and 13
; few signs were detected
on routes 1
, 3, 10
and 1
2
.
No signs of hyaena were found on routes 5, 8, 9 and 11.
Only
on routes 3, 5 and 9
were
signs of cheetah occurrence detected
(see figures
2.5.3.a,
2.5.3b and 2.5.3c).
30
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
0
1
2
3
4
5
Amount
1
2
3
4
5
6
7
8
9
10
11
12
13
Route
Predator Scat Frequency
Leo
BH
CH
Figure 2.5
.3
b
.
Amount
of scats from different predators on fixed
survey
routes
, O
kambara
2013
.
L
eo
= leopard, BH = brown hyaena, CH = cheetah
.
0
2
4
6
8
10
12
14
16
Amount
1
2
3
4
5
6
7
8
9
10
11
12
13
Route
Predator Track Frequency
Leo
BH
CH
Figure 2.5
.3c.
Amount
of tracks from different predators on fi
xed survey routes, O
kambara
2013
.
L
eo
= leopard, BH = brown hyaena, CH = cheetah
.
31
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.
5.4. Camera traps
The study period started at
the beginning of August 2013.
F
ifteen
c
amera traps were
placed at strategic points (based on
tracks and scats, near waterholes
) around the study
site. Elephants destroyed
two
camera traps in the plains area during the
study period
. The
study design was adapted accordingly and no further camera traps were placed in areas
that could be reached easily by elephants
.
Due to logistics, all available cam
era traps were
not set on one single day in the field but on several consecutive days.
Overall, between
four
and 15 camera traps were in use
at the same time
throughout the whole expedition
(number of camera traps increasing with each additional day in the
field during the first
slot)
.
Not all leopard
s photographed could
be identified. From
37
events
,
11
% of
photographs were either too poor in quality
(e.g.
blurred or overexposed
) for
the fur pattern
to be sufficiently visible,
or close
-
ups showed only smal
l body sections. In total
33
useable
leopard
events were recorded
throughout the whole expedition
, wh
ere an event is a
picture with as a certain
individual
identifiable
leopard captured by a camera trap. This
equals one leopard capture for every
3.5
night
s
of trapping.
Capture success
A
sampling
period
of
9
0 days was set and conducted
fr
o
m
17
August to 15
November
2013
, yielding
29
events
(as described in the methods section above, an overall sampling
period of 90 days was chosen; not all 33 events happen
ed during this period)
.
Six
individual
adult leopards were identified
by their coat patterns
. Four
adult males
were
recorded
;
one of them had been
captured and
collared in
2012
(LM04
, see
Killian &
et al.
2012) w
ere captured in box traps and equipped with
collars during the study period.
In
addition,
two
mature females were photographed on several occasions
.
One female was
photographed togetherwith L055.
Table 2.5.4
a.
Camera
-
trapping
effort and leopard captures 2013
.
Sampling period
Trap stations
Leopard
s
Identified individuals
17 Aug
15
Nov
1
4
29
6
During the sampling period,
14
camera traps were active
(15 were set out, but one broke
before the beginning of the sampling period)
,
ten
of which recorded leopard
s
.
Seven
cameras
recorded brown hyaena
re
peatedly
; pictures were not clear enough to identify
individuals
.
In two occasions, spotted
hyaenas were
recorded as well.
In addition, pictures
of cheet
ahs
were taken by four
cameras;
five different individuals could be identified
based
on their individua
l fur pattern
(see Table 2.5.4b)
.
Events of pictures taken of other
predator
s
within the study period are listed in Table 2.5.4
c
.
Table 2.5.4
b
.
Camera
-
trapping
effort and
cheetah
captures 2013
.
Sampling period
Trap stations
Cheetahs
Identified individual
s
17 Aug
15
Nov
1
4
13
5
Table 2.5.4
c
.
Number of c
amera trap pictures of different predators.
Year
Sampling
period
Brown
hyaena
Spotted
hyaena
Honey
badger
Caracal
African
wildcat
Jackal
2013
17 Aug
15
Nov
41
2
8
9
12
45
32
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Estimates of leopard captu
re probability, population size and density
The CAPTURE test for closure supported the assumption of population closure (i.e. no
immigration, emigration, births or deaths) during the survey. CAPTURE selected the null
model (M
0
) for the survey. A relativel
y high captu
re probability of 0.4207
was recorded (
the
probability that a leopard in the sampled area is photographed on a single sampling
occasion)
(Table 2.5.4
d
).The sample population was estimated to be
seven leopards (SE
±
0.2138
, 95% CI
6
-
6
). When com
puting the 95% confidence interval, CAPTURE converts
the values to the nearest integer, rather than printing decimals (
Jackson
et al
.
2006
).
For the survey
seven
individual leopards (excluding subadults) were e
stimated to occupy
an area of
93
.42
km
2
. The
buffer width (half of the home range length of a female home
range) was 4.4 km.
The estimated effective area sampled
was 304
.52
km
2
. A density of
1.9
individuals per 100 km
2
was calculated.
Table 2.5.4
d
.
Results of population closure, capture probability,
estimated abundance, standard error and 95%
confidence interval of leopards sampled on
Okambara game farm, Namibia
,
in
2013
.
Null Model (M
0
)
Test for closure
Capture probability
Abundance
(SE)
95% CI
z = 0.
583
P = 0.
614
0.
4207
6
±
0.
2138
6
-
6
2.5.5.
GPS
c
luster
a
nalysis
The GPS c
luster analysis
based on the data of L051, L052 and L055 showed
no
specialisation of individual leopards.
Prey species varied
among
eight different species
(see
F
igure 2.5.5). At 36% of the visited kill sites, remains of gre
ater kudu were found.
Twenty percent
of the kill sites revealed remains of impala
,
and at 16% of the kill sites
remains of oryx were detected. Remains of warthog
or steenbok
were
found at 8% of kill
sites.
Mountain zebra, cattle and porcupine were also fou
nd, each of which constituted 4%
of the kill sites.
Figure 2.5.5
.
Prey range of leopards
(
%
of kill sites)
on Okambara and surrounding farms
.
33
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.
5
.
6
.
Game counts
Vehicle game counts were conducted
during daylight hours
betwee
n August and
Novem
ber 2013
using
three
line transects
(see F
igure 2.
4.6a)
and distance sampling
methods. All routes were driven from south to north and started at the same time in the
morning (
at
sunrise).
There was a tracker or scientist on each game coun
t vehicle to
provide some standardisation of observations and
increase
detection probability.
The three transects lengths were VGC 1 = 10.
8
km, VGC 2 = 14.9
km and VGC 3 = 12.
7
km.
Fourteen
game counts were
conducted, yielding
5
,
041
animals
over
537.6
km
driven
.
Results of counted animals from the three transects were examined over the entire area
(150 km
2
). Livestock animals such as cattle, horses and donkeys were not counted
.
Re
sults are listed in Table 2.5.6a
.
Over
time, the same groups, e.g.
brindled
gnu
(blue wildebeest)
and white
-
tailed gnu
(black wildebeest)
, were observed repeatedly in certain areas of the farm (e.g. on VGC 3).
Animals were
relatively easy
to detect, because they prefer to stay in large groups in open
areas to feed on grass.
Resul
ts of the waterhole observations at
six
different waterholes (F
igure 2.
3
a; Bergposten,
Frankposten, Gustavposten, Sandposten
, Michael’s Dam
and Boma
) produced
1
,
061
animals over
365
sightings. The percentage of juveniles was
18
.6%. The study period was
bet
ween August and November
2013
.
Some
of the time animals noticed the observers in
the hide,
but more than 90
% did not flee.
The total numbers of obser
ved game species are
listed in T
able
2.5.6
b
.
Table 2.5.6a
.
Numbers of
observed
ga
me species
(
total no. of
individuals
seen on vehicle game counts during
expedition period
, mean
no. of individuals observed on vehicle game count during expedition period (total
number/num
b
er of game counts carried out)
,
individuals
/km
2
(mean/150km²
)
)
, O
kambara
2013
.
Species
No.
Individuals
Mean
Individuals
/km²
White
-
tailed Gnu
(Black Wildebeest)
967
69
.
07
0
.
46
Brindled Gnu
(
Blue Wildebeest
)
193
13
.
79
0
.
09
Common Duiker
9
0
.
64
0
.
00
Eland
350
25
.
00
0
.
17
Giraffe
231
16
.
50
0
.
11
Greater Kudu
271
19
.
36
0
.
13
Impala
447
31
.
93
0
.
21
Klipspringer
17
1
.
21
0
.
01
34
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Mountain Zebra
316
22
.
57
0
.
15
Oryx
1
,
504
107
.
43
0
.
72
Plains Zebra
81
5
.
79
0
.
04
Red Hartebeest
88
6
.
29
0
.
04
Sable Antelope
38
2
.
71
0
.
02
Springbuck
233
16
.
64
0
.
11
Steenbuck
106
7
.
57
0
.
05
Warthog
138
9
.
86
0
.
07
Waterbuck
53
3
.
79
0
.
03
Total
5
,
042
360
.
14
2
.
40
Table 2.5.6b
.
Total n
umbers of
observed
ga
me species at waterholes on Okambara in 2013
.
Species
Total
White
-
Tailed Gnu
90
Brindled Gnu
26
Common Duiker
1
Eland
25
Giraffe
43
Impala
288
Greater Kudu
70
Oryx
50
Plains Zebra
16
Red Hartebeest
38
Sable Ante
lope
5
Springbuck
11
Steenbuck
1
Warthog
386
Waterbuck
11
Total
1
,
061
35
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
2.6
.
Discussion and conclusion
s
2.6.1. Capture and collaring
Information on leopards is difficult to obtain by visu
al observations,
so n
on
-
invasive
methods were chosen to gain ecological and biological information
. Also,
these
methods
are cost
-
effe
ctive, objective and repeatable
(
Norton
-
Griffiths 1978
).
With the help of
identifying leopard signs, such as tracks and scats
,
as well as
u
sing
camera traps
,
potential locations for box traps were identified
.
In addition, box traps were placed near
waterholes
close to
mountain
ridges
, since leopards
prefer mountainous habitat but
need
to take in water regularly. Success of trapping the main t
arget species, i.e. leopards, varied
among
box traps.
Only in BT02A and BT04,
set on a track in the mountains and
at
the
Bergposten waterhole respectively,
were
leopards captured. Lack of trapping success at
the other trapping locations
could be due to var
ious reasons. Traps BT01
and
BT01A were
set
near
waterholes
; at
both
locations signs of leopards (tracks or camera trap pictur
es)
were detected. However, GPS
data gained later through the three collared leopards on
Okambara showed that leopards seem to rev
isit the waterholes every 7
14 days
,
depending on the availability of water sources
with
in the home range.
The bait in the trap
s
was replaced every 3
5
days
(depending on resources such as staff and meat)
since
Bailey (1993) showed in his st
udy on leopards
in South Africa
that the trapping success
decreased
significantly with the time bait was left in the trap
.
Therefore it is possible that
o
n various
occasions leopards visited
waterholes when bait
was
older and less attractive
.
In addition
,
very
little rai
nfall was measured in the study area
in 2013
, leading to a long dry
season with very little grass
(
i.e.
little
food for
many
herbivore species
)
. A lot of prey
species were weakened due to the lack of sufficient food supply and therefore
were
easy
target
s
f
or predators such as leopards.
This too would make baited traps less attractive.
Also, some prey animals died due to the drought, so there were also more possibilities
than usual
for leopards to scavenge on these carcasses. This high availability of prey,
which only required little effort
to catch
,
is likely to
have led to a lack of interest
by
the
leopards in the bait.
BT02 was set on a track about 1
km
from a
farm house, near the fence line and close to
various ridges. Although
at least two
leopards
visi
ted
the trap (
as evidenced by
camera
trap pictures), none went in. Leopards are cautious animals, especially in areas where
they
are highl
y persecuted, and
therefore might be hesitant to enter
novel things such as
traps; so
the bait
in BT01 and BT01A
might
have been not interesting/fresh enough for the
leopards to be interested
in going inside the box trap
.
BT03 was set near a farm house
next to a hole in the fence
,
which
camera traps had shown was
used by at least one
leopard (L055
,
which was then captured
in BT02A). This trap
caught a lot of other animals,
m
ainly warthogs and porcupines. C
amera traps set there
at
the beginning of the trapping
season showed an intensive use of the track next to the trap, especially by warthogs.
Trapping a variety of non
-
tar
get animals meant that the
trap was
often
blocked
and this
decreased the
leopard
trapping
likelihood
at this location.
Since there was no
leopard
trapping success, this trap was then relocated to the north of the study area
(BT03A
)
next
to a track
close to
the northern fence where
signs of leopards
had been
detected several
times (tr
acks and camera trap pictures).
N
o leopards were captured
in the new location
either
. Leopards tend to reuse the same tracks/areas for patrolling their territory
,
but
, as
with
w
aterhole revisitations
, there are usually a f
ew days
between
visits
. Again, it is
possible that the leopards had no
interest in the bait due to
the age
of the bait in
co
mbination
with the high prey availability.
36
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
BT03B
was
set late
in
the capture season; le
opard tracks were found in front of the trap,
but no animal was captured. The lack of trapping success might have been
due to the
short amount of time
the trap was set at this location. Setting a box trap
brings with it
a lot
of impact, e.g. cutting shrubs
,
driving to the locations, etc. The trap will also have had an
intense human smell as it was handled by several people during relocation. So
the amount
of time between setting the trap and the end of the trapping season might have been too
short for anima
ls to get
used to the new feature in their surroundings.
L051 was captured on its kill that had been found after following a drag mark and
accompanying tracks. The
probability
of trapping a leopard on a fresh kill is high, due to
the species
feeding beha
vio
u
r. Leopards usually hide their prey and revisit the site for up
to several days until the prey is fully consumed. Thus, a fresh kill, especially if only lightly
consumed, indicates an increased likelihood of the leopard returning to this location withi
n
the next 24 hours.
After the capture of L051, the trap remained
active
at the kill
site
for a
couple of days,
in case of another leopard showing interest in the kill
,
but no animals were
captured. Since the kill was hidden in the middle of the bushes wit
h no clear game tracks
around, it is highly possible that no leopards
passed the location
within these days.
None
of the leopards were recaptured in the box traps
during the trapping season of this
study
. This suggests that leopards have an excellent memo
ry of
trap location and most
likely
develop an aversion to traps (getting trap shy). Interestingly, leopards living in
national parks (Bailey 1993) showed no aversion to box traps; researchers captured some
individuals up to
20
times during a study perio
d.
Several predators were captured and they showed
the presence of
a variety of different
carnivore species on commercial farmland.
An
adult
fe
male
brown
hyaena
C056 was
captured once
at the box trap near the Bergposten waterhole (BT04).
Brown hyaenas are
m
ainly scavengers and are therefore attracted by the bait used in
traps set for leopards
(Estes 2012)
.
Honey badgers were captured
three times. T
hey mainly feed on insects and small
vertebrates but are also known for scavenging
,
which is most likely why th
ey were
attracted to the bait and went into the traps
(Estes 2012)
. A
n adult female
was captured
accompanied by a subadult female, which was most probably her offspring. Several days
later, the same adult female went into the same box trap again; she could
be identified by
an identification chip that was implanted during the previous handling. Since samples had
already been taken from this animal it was re
leased without immobilis
ation. Also,
an
adult
male honey badger
that had been captured at BT03A
went ba
ck into the same box trap in
which it had been trapped several days
before
. Again, the animal could be identified via
chip implant and was released with
out immobilis
ation
. Honey badgers can have vast home
ranges (
males:
541
±
93
km
²; females:
126
±
13 km
²)
but only travel small distances within
a few days (
Begg et al. 200
5
). This could be a reason
why
the honey badgers
were
recaptured within only a few days.
Since they were not captured again it might be that they
either moved
to another area of their home r
ange
or learned from the repeated experience
to avoid the trap.
No cheetahs were captured because the traps were set
specifically
for leopards and
heavily scented bait was used. Cheetah capture requires different trap positions and
settings, e.g. at marki
ng trees or with live bait.
37
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
Blood samples taken during capture and
immobilisation
were sent to the Institute for Zoo
and Wildlife Research (IZW) in Berlin, Germany for further analyses. Results gleaned will
be published elsewhere, probably in late 2014.
2
.6.2. Home range
Home ranges (the area regularly used by an individual) of some carnivore species overlap
considerably among individuals, depending largely on resource density and distribution
and genetic relatedness (
Moyer et al. 2005
).
Leopards, just as
other large carnivores,
cover
home ranges that
have to be of a size
large
enough
to provide sufficient prey
availability throughout the year
.
Where prey distribution is constant
,
these territories are
often stable, but under other circumstances they drift
(e.g. red foxes
Doncaster and
Macdonald 1991
), move with migrating prey (e.g. wolves
Walton et al. 2001
) or are fixed,
but temporarily left by individuals to find prey (e.g. spotted hyaena
Hofer and East 1993
).
Data obtained via GPS collars from the
three male leopards
collar
e
d on Okambara
showed no changes in home r
ange sizes over a period up to nine
months from the end of
the dry season throughout the rainy season.
Since on farmland prey movements are
restricted by fences, especially on game farms
with fences that keep the game
confined
with farm boundaries
, prey availability is stable throughout the year. Therefore leopards
are not required to migrate with prey or leave their home range to find prey.
Typically, adult male leopards require larger h
ome ranges than females
. S
everal studies
revealed similar home range sizes within each sex, with males
varying from 17 to 76 km²
and females ranging from 6 to 18
km²
(
Bailey 1993, Hamilton 1976, Bertrum 1982,
Mizutani
&
Jewell 1998, Stander
et al.
1997, No
rton
&
Lawson 1
985, Norton & Henley
1987,
Seidensticker 1976
).
However, i
n
arid areas
the home range sizes of leopard
s
can
be
much larger
(
male
s
86
km², females 22
29
km²
;
Jenny 1996
)
,
and even for home
ranges of individuals of the
same sex there can be so
me degree of overlap
,
or
some
times
even a complete overlap,
e.g
.
if
the
whole
home range of a female
is
within the home
range of a male
(
Jenny 1996,
Rabinowitz 1989, Grassman 1999
).
Most
leopard home
range data available are from studies that were conducte
d in protected areas (see
Appendix I: Kruger N
ational
P
ark
, Serengeti N
ational
P
ark
). In Namibia the sizes of MCP
(95%) varied
from
108
to
229 km
2
for males and 53
to
179 km
2
for females (see Appendix
I). A study conducted by Marker and Dickman (2005) o
n c
ommercial farmland found 229
km² (MCP95) as home range size
for male leopards and for females a range of up to 179
km
2
.
Stander et al. (1997)
estimated home ranges of male leopards varying from 210 to
1
,
164 km
², whilst
female
home ranges measured 183 to 19
4
km
²
in
north
-
eastern
Namibia
.
In this study,
the home range sizes found for the male leopards correlate mainly with the
study of
Marker and Dickman (2005
)
, since home range sizes
(MCP100)
f
or
L052 and
L055
were
272
km² and
142
km²
respectively
,
but
they
also
correlate
with the findings of
Stander et al. (1997
)
showing
that there can be a high variance between different
indiv
i
duals (L051, 410
km²).
The fact that home range sizes on farmland are larger than in
protected areas, despite the stable prey avail
ability throughout the year, could
be
due to
varying densities and abundance of leopards in protected areas and on farmland.
Stein et
al. (2011)
determined
a
leopard population
density of
3.6
leopards
per 100
km²
on
farmland in central Namibia.
Marker and
Dickman (2005
), on farmland in north
-
central
Namibia, found that the average density of leopards
on farmland
is 2.1 leopards per 100
km
2
.
O
n the current Okambara study site, the estimated leopard density
in
was
1.3
individuals per 100 km
2
and 1.9 individua
ls per 100km²
in 2012 and 2013, respectively
.
38
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
There might be a high variance of leopard density even within an area, depending on the
habitat and management of different farms. Some farmers sh
o
ot leopards regularly on
their premises, thus often removing th
e territor
y holders
and creating
space
for other
,
sometimes neighbouring
leopards to move in. Therefore, leopard densities
outside of
Okambara
,
on which
leopards are not persecuted,
might be lower than expected and
home ranges can be larger than in protect
ed ar
eas where there is less turn
-
over within the
population.
Game
-
proof fences, in general, were not a barrier
and had no influence on the
home range
, as
all three individuals
regu
larly crossed under the fence on
Okambara
.
Furthermore, prey availability m
ight be lower on farmland compared to protected areas.
Although there is game farming in Namibia (including Okambara), the majority of farms
is
engaged in cattle breeding, thus only naturally abundant game species appear on those
premises. Also, the game c
an move between cattle farms as the surrounding fences are
purposely kept low in order to allow migration of game.
Local prey availability is therefore
subject to change and can affect the home range size of leopards
.
L051 had the largest core area (K50,
120
km
²) of the three
male
leopards
in this study
,
whilst L052 and L055 had simi
lar sized core areas (K50, 50 km² and 41
km², respectively).
Although L051 was thought to be the youngest, all of the three males were estimated to be
of similar age (5
7 years
). So, the variances in the home range sizes and the core areas
are most probably not due to age.
It is possible that differences in habitat led to varying
home ranges.
Towards the s
outh, w
h
ere the home range of L051 is situated,
fewer
mountainous areas a
r
e available, whereas towards the north and w
est, the areas of the
home ranges of L052 and L055 respectively, the landscape provides several mountains.
Leopards prefer mountainous habitat and generally try to avoid open areas. Therefore,
most likely there a
re
fewer
neighbouring ter
ritories of adult males in the s
outh, hence
fewer
limitations to the home range of L051. In the north and w
est,
however,
the intraspecific
competition of other
surrounding
territorial males
keeps
the home r
anges of L052 and
L055 sm
aller.
This
notion
is further supported when visualising the GPS data of the three
males; although there is some overlap
along
the edges of the home ranges, the male
leopards usually stay within their individual home ranges and avoid each other.
O
verlap
oc
curs more often between individuals of different sex (
Arthur et al. 1989
). In leopards,
males defend
their
territories against other
sexually mature
males, but tolerate females,
cubs and even dispersing young ma
les within their territories (
Bailey 1993, Ma
rker and
Dickman 2005
).
C
amera trap pictures of female
and
younger leopards taken within the
home ranges of the three males
on Okambara support this finding
.
2.6.3
.
Track
counts and scat collection
The largest number of tracks and scats from leopard
s
wer
e found close to the edge of the
small mountains in the north and southwest of the study site, where there is also water
available (routes 1 & 3).
Also, on route
s
6 and 12 several signs of leopards where found;
both routes include waterholes as well. Few s
igns of leopard occurrence were detected on
routes 2, 4, 5, 10 and 13. No signs of leopards were found on routes 7, 8 and 11, all of
which are situated in the plains.
This correlates
with t
he analysis of the GPS data, whi
ch
showed mountain
o
us
areas to be t
he preferred habitat of leopards.
Results of these non
-
invasive methods
also
show
ed
that
other
predators occur on
Okambara. Cheetahs were detected more frequently in the open areas of the study site,
where they
probably avoid
other predators such as
leopa
rd and brown hyaena and
because their hunting technique is more suited to open areas (
Caro 1994
).
Brown
hyaenas
39
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
occur throughout the study site as
they are looking for carcasses from other predator kills
everywhere and they patrol their home range. Cheetah
s are known to leave most of their
kill after having eaten their fill and do not usually return (
Caro
1994). If leopards do not
take their kill up a tree
or hide it properly
, the probability that
hyaenas
will
scavenge on it is
high.
There were some restri
ctions to these methods
.
Route
1 and
4
included
very hard, stony
surfaces and some of the routes
were in deep sand (2, 5, 6 and 8)
,
which can make the
detection of tracks quite difficult. Also, weather conditions such as wind and rain influence
the success
of finding tracks; there was not much rain
fall during the study period
, but
there
were several days with
strong
wind
s
.
2.6.4. Camera traps
Camera
traps
are
very
useful
tools
in
wildlife research, collecting
a variety of
data
sets
and
allowing for
undist
urbed
observation of species
in their habitats
to explore
their
natural
behaviour patterns
and movements,
and
to determine
population sizes
.
During the four months of the
study period
,
164
pictures of all large predator species
present in the area (leopar
d, cheetah, jackal, caracal, brown hyaena
, African wild cat
and
honey
badger) were recorded. Brown hyaena, honey badger
, African wild cat
and
all of
the
caracal
pictures were
taken
during
or
after sunset.
All of these animals are mainly active
during the n
ight; therefore detecting those species during the day is rather unlikely.
Jackals were recorded mostly during the night, when they are
most active,
but they can
also be seen during the periods of dusk and dawn.
Brown hyaenas and African wild cat
were dete
cted all over the study area,
showing
the vari
ed
habitat selection within these
species. Caracal and honey badger were only detected near
ridges either at
waterholes or
along
fence lines
; apparently they prefer this
type of habitat.
Several studies
sugges
t
that
leopards
are active
between sunrise and sunset
(
Nowell and
Jackson
1996,
Hamilton
1976,
Bailey
1993,
Sunquist
&
Sunquist
2002
)
. This is confirmed
by
pictures taken
of leopards
during this study
period
,
with leopard
s
active just before
sunset as well
as
in the middle of the night
,
and
with
only a
few
records after
sunrise
,
when temperatures
are
still low
. In semi
-
arid
savannah
regions
mammals
are active at
night
to avoid
the
heat of the day
and
associated heat
stress and energy loss
(
Bothma &
Bothma
2
005
)
,
and l
eopards
usually
avoid the
heat and prefer shady
places
to
rest
(
Bothma and
l
e Riche 1984,
Walker
199
6,
Sunquist
and
Sunquist
2002
)
.
This applies to
predators as well as herbivores.
So, the locally abundant prey species are also less active
durin
g the
heat of the day
and
it is self
-
explanatory
that
the
activity patterns
of
leopard
prey
have an influence on
leopard hunting behaviour
. T
his
leads to an increased activity of
leopards between sunset and sunr
ise
(
Bothma and
Bothma
2005,
Sunquist
&
Sunqu
ist
2002,
Jenny and Zuberbühler
2005
)
.
Furthermore, in agricultural areas
,
where there is
also intensive persecution of predators, these species tend
to be
more active during the
night
to avoid humans.
A high percentage of the predator pictures wer
e of br
own hyaenas and jackals, but no
individuals could be
identified. Due to their greater strength, size and stealth, leopards are
the dominant predator
s
in the study area
.
No persecution of predators occurs on
Okambara so predator r
atios
should
be subject mos
tly to
natural selection pressures
. In
general, on Namibian farmland, the leopard is the only competitor for brown hyaena and
40
© Biosphere Expeditions,
an international not
-
for
-
profit
conservation
organisation registered in England, Germany, Fra
nce, Australia and the USA
O
fficially accredited member of the United Nations Environment Programme's Governing Council & Global Ministerial Environment Forum
Officially accredited member of the International Union for the Conservation of Nature
cheetah, and it is likely that the latter will avoid areas where many leopards occur. Brown
hyaena
s
and leopard
s
d
o coexist in spa
ce, but they tend
to avoid each other in time
(Killian
et al.
2012). Generally, leopard
s
and brown hyaena
s
are opportunists and better
adapted to poore
r habitat conditions (
Estes 2012
).
Marker and Dickman (2005), on
farmland in north
-
central Namibia, found
that the average density of leopards in protected
areas is 2.1
individuals
per 100 km
2
.
On this
Okambara study site, the
esti
mated leopard
density was
1.9
individuals per 100 km
2
.
In relation to the study of Marker and Dickman
(2005) and of
Stein et al.
(
2011)
,
who
determ