Wildlife Counts in North West Namibia

Jo Tagg & Greg Stuart-Hill

This presentation highlights the key elements in developing and implementing an appropriate wildlife census

The challenges include meshing a diversity of stakeholders and technologies:

• Stakeholders– Committed local communities– Government departments– Conservation NGOs

• Technologies– Pen and paper– Boot leather– Satellite imagery– GIS

North West Namibia - a little background

Encompasses northern part of Namib Desert

Arid, mountainous and spectacular landscapes

Fifteen registered and soon-to-be registered conservancies and two tourism concession areas

Large, open system with considerable movement of wildlife

Remote areas with often difficult access

The objectives of CBNRM in Namibia

These objectives are achieved through the formation of Conservancies…

The Community-Based Natural Resource Management programme is a joint venture between government, NGOs and communities, which aims to:

promote wise and sustainable use of natural resources and encourage biodiversity conservation

devolve rights over and responsibilities for wildlife to rural communities and create income generation opportunities

encourage and assist communities to acquire skills to manage their area sustainably and actively pilot their own future

Where are conservancies in Namibia?

Between 1995 and 2000, conservancies received income and benefits of N$8.5 million (approx. US$0.9 million)

Over 27,000 rural people live within conservancies

Around 40,000 km2, representing 5 % of Namibia, is now under conservancy management

Through conservancies, local people have now assumed responsibility for managing the wildlife in the region.

Objectives Reasons why information is needed

1. Estimate the Numbers of game

How many?

For:• setting reasonable hunting/capture quotas;• estimating stocking rates to manage competition with livestock

and protect veld; • determining the value of wildlife in the Conservancy.

2. Produce Game Distribution maps.

Where are they?

For land-use planning (Zonation), it is important to identify areas of high game concentrations. Also to see how these distributions change in future years in reaction to rainfall or human factors such as water distribution or settlement .

3. Monitoring Population Change (“Trend”)

Is wildlife increasing or decreasing?

With successive censuses, graphs can be drawn showing population changes of each species (e.g. are springbok increasing or decreasing?). This will tell managers whether or not they are achieving their game management goals and consequently indicate if it is necessary to change management strategies.

One of the most basic requirements is game census tool that will:

The operational challengesDevelop a count method which: is cost effective, sustainable and repeatable

meets the objectives of conservancies and allows individual units to contribute to the important regional picture

balances local ownership and scientific accuracy, and

is appropriate to this extremely remote and rugged area of some 5.2 million hectares

Vehicle-based road count developed• In 2000 a vehicle-based road count was piloted, one by

one, in seven conservancies.

• In 2001 and 2002 the entire area was counted as a single event over 16 days.

• The count involves 18 community organisations, Government, support NGOs and the private sector: a total of about 220 individuals.

• A total of 119 routes, with a combined length of about 6,500 km are counted.

Remember, the method must meet all three objectives:

i.e. Population numbers Wildlife distributions Population trends over time

Methods• Works well for common plains game

but will not give good results for all species (smaller secretive sp, nocturnal sp, and sp in mountains).

• Other monitoring methods (e.g. aerial census, foot patrols, specialist species monitoring) and local knowledge also important.

The road-count was designed so that local people can do this count with minimum outside assistance (e.g. borrowing a few vehicles with drivers once a year).

The road counts will add value rather than replace the other methods

Each route was planned with the local community. These have become permanent routes and will be resurveyed in future years!

FIELD RULES

For determining game NUMBERS1.   Centre line (the road and immediately next to the road) are priority areas for searching.2.   Distance must be to the animal before it runs away3.   Distance must be at right angles to the road4.   Distance is to center of groups of animals (before the group moves away)

5.   Where the route travels next to a fence only the animals inside the fence are counted (the route distance is then halved for that section of the route)

6.   Routes must represent all habitats proportionally (i.e. also count low density areas)7.   Measure strip width per route8.  Only count adults and sub-adults - make a note of numbers of newly born juveniles (or newly

hatched chicks – ostriches)  For TREND analysis, a number of additional rules were added: 9.    Fixed routes will be used for subsequent counts10.  Start time is at sunrise 11.  No binoculars to be used (knowing that leads to underestimation of numbers)12.  Always count from the back of an open bakkie13.  Speed must never exceed 35 km/hr  For Game distributions, an additional rule was added: 14.  Location of each sighting is mapped using the 2km x 2km grid map

Some comment on estimating populations

The road-count data are analyzed in two different ways:

1. As a Strip-Count: provides a quick field-based population estimate - ideal for the conservancies or concession holders.

2. Using Distance ©: more technical but more accurate - appropriate to regional population estimates and is done back in Windhoek.

Both methods use 'Correction-Factors' to estimate the true game populations: i.e. the numbers of animals counted is multiplied by the correction factors.

The next few slides discuss these two approaches in determining game population numbers

ROUTE STATISTICS (Road Count) Route1 Route 2 Route 3 Route 4 Route 5 Route 6 Route 7 Route 8 Route 9 TOTAL Transect length (km) 78 59 54 53.3 51 54.5 350 Transect width (km) 0.8 0.8 0.8 0.8 0.8 0.8 5 Hours & Min 4h15m 2h25m 3h05m 4h15 2h50 3h20m 20h10 Area sampled (km

2) 62.4 47.2 43.2 42.64 40.8 43.6 0 0 0 280

Area represented (km2) 522 458 352 316 396 648 2692

Area excluded (km2) 188 98 444 110 - - 840

% Sample 12% 10% 12% 13% 10% 7% - - - 10% Correction factor 8.4 9.7 8.1 7.4 9.7 14.9 - - - -

Total Area 3532 km2

Each area is divided up into count zones. Certain areas are excluded on the basis that no route

adequately covers such terrain.

RO

AD

Area

Rep

rese

nted

Area

sam

pled

STRIP WIDTH

Zone 1

Correction factor = number of times area sampled fits into the area represented

Strip count methodology to estimate populations

Similar in concept to the transect method

but

It accounts for the differences between species (i.e. it is easier to see an elephant than a springbok!)

alsoIt accounts for the difference in terrain

(i.e. effective strip width changes depending on terrain!)also

It accounts for missing animals in the strip (i.e. not every animal is seen even within the effective strip width!)

Distance based methodology to estimate populations

As sighting profiles are built up for each species, more sophisticated DISTANCE © analysis will be used to determine populations.

RO

AD

Area

Rep

rese

nted

Area

sam

pled

STRIP WIDTH fo

r

Springbok

Zone 1

STRIP WIDTH fo

r

Elephant

BUT Remember the rules!

Key components of the wildlife counts:

Planning

Count zones

Route maps

Zonations derived from

satellite images

Routes planned from orthophotos

GIS generated planning maps

Undertaking

N/W Namibia

5 million ha

Reporting

Gemsbok Distribution

Springbok Population

??

??

??

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Populations estimated

using spatial analyses

Distributions mapped

All activities are underpinned by participation and partnerships

Local mapping of conservancy tracks and features

Accessing and bringing national level data to local communities

Identifying routes – local knowledge is the key

Zonation – satellite images and orthophotos allow insights of habitats in inaccessible areas. Zonation is done manually by combining simple GIS techniques with local knowledge

Production of individual route maps for field work

Developing map reading skills

Planning the count involves:

Undertaking the count:

Logistics of the count are done at conservancy level

Count teams use route maps which have an orthophoto backdrop overlaid with the route as well as other tracks and point features

Clip boards and data sheets used

Reporting involves:

Further scientific and CBNRM programmatic analysis is conducted at central level (national government and NACSO)

1. Conservancy-level analysis

Field analysis and feedback meetings immediately following the count (local communities, NGO and Govt representatives)

2. Regional-level analysis

Regional analysis and reporting workshop held some months following the count (NRWG, regional NGO and regional MET representatives)

• Distribution and numbers sighted are discussed and local insights given

• Population estimates are made using correction factors and local knowledge

• Trend graphs are completed – these are completed manually and remain with the conservancy. Actual numbers sighted per 100km driven are used and thus are uninfluenced by correction factors.

1. Conservancy-level analysis

Meetings are held with conservancy representatives immediately after each count:

Hand drawn trend

charts

Distribution maps

Final Population Estimate Table Population estimates

Species Transect Method

Local knowledge

Aerial Census

Usable Pop estimate Comments and Discussion

Elephant 272 272

150 ** 150 Local intensive survey estimated a pop of 150

Rhino ** 1 ** 1 A known individual in the …….river valley

Springbok 7,756 7,756

4,000 4,415 5,500 Aerial census underestimates springbok

Gemsbok 1,718 1,718

? 1,907 1800 Zebra (Hart.)

261 261

? 530 500 Ground counts do not access the mountain areas

Kudu ** 100 54 100 Whilst ground counts missed kudu there are present

Ostrich 1.143 1,143

? 853 1000

Giraffe 52 52

30 24 30 Ground counts overestimated Giraffe

• Data for common species subjected to DISTANCE analysis

• Technical support people from Government, NGOs and the Windhoek-based technical team discuss population estimates for each conservancy. Individuals with local knowledge in each count area take the lead

• The group agrees upon population estimates for the region and compiles a report in poster format which is completed and printed at the workshop

• Regional population trends produced on the basis of actual number of animals seen per 100km driven

• Distribution maps and distribution maps generated by GIS

2. Regional-level analysis

Ostrich

-

50

100

150

200

250

300

350

2000 2001 2002 2003 2004 2005

Trend charts

Distribution maps

Year 1 Year 10

Distribution change maps

Gemsbok 15,074 35,890 2.38

Giraffe 644 - 4.60

Jackal 840 2,963 3.53

Kudu 1,578 6,787 4.30

Ostrich 1,926 6,262 3.25

Springbok 94,932 237,329 2.50

Steenbok 1,291 19,864 15.38

Zebra 4,087 12,385 3.03

Min

Pop

est

imat

e

Lik

ely

Pop

estim

ate

(exc

l mou

ntai

ns)

Population

estimates

Regional report for 2002:

Regional Database

The data are captured into a database which has been designed to:

produce automated reports for each conservancy. These are circulated to local level stakeholders to aid local level decision making such as setting harvest quotas

produce an automated report for the region. This is used by Government to verify and adjust the quota applications of individual conservancies so that they are compatible with regional game population trends

link to a GIS to produce current game distribution maps and distribution change maps (comparing current to previous years counts)

make data readily accessible for:• sophisticated analysis such as comparing wildlife distributions with ‘rainfall’ distribution (using

NDVI satellite imagery) or measuring the impact of factors such as water distribution, livestock density and changing land use on wildlife distribution

• programmatic M & E

At a central level…

2002 estimates

Population increment

% #s Value N$

Value of increment N$

Gemsbok 35,890 25 8,970 1,000 9,000,000

Kudu 6,790 20 1,360 1,000 1,357,400

Ostrich 6,260 5 310 1,000 313,100

Springbok 120,000 27 32,400 300 9,720,000

Zebra 12,380 7 870 800 693,560

Robust monitoring and management allow conservancies to make land-use decisions based on the real value of wildlife. Estimating populations and potential are the first step in this process.

Approximate value of the annual increase in the game numbers

Additional spinoffs:

Wildlife data support conservancies in attracting donors, investors etc. Caution: it is the activity rather than numbers which are the key as sustained droughts could have big impacts on numbers but as long as conservancies are monitoring it is an indication of good management.

Tourism planning and development, a key activity for the economic viability of a number of conservancies, is dependent upon a temporal and spatial understanding of wildlife.

CONCLUSIONS Equal emphasis has been devoted to the three critical elements of game

population monitoring: (i) participative planning, (ii) undertaking the counts and (iii) analysis and reporting

The technical tools and products derived from remote sensing and GIS have enabled working relationships between stakeholders in north-west Namibia to be built

The use of GIS technologies for the wildlife counts supports not only the management of resources at local level but also management and monitoring across geographic and social areas, over time and with different communities and institutions so the results are continuous, representative and compatible

All stakeholders agree on the derived population estimates even though they challenge long held misconceptions about game numbers in the region

If deemed necessary by key stakeholders, considerably more development work could be done for which GIS/RS would be indispensable tools.