Makgadikgadi Zebra Migration Project

Region: akgadikgadi Pans National Park

The overall aim is to determine the ecological impact of the Makgadikgadi game fence and provide sound ecological information to the Department of Wildlife that can be used to assist in the long-term management of the region.

The position of the game fence, along the south-western boundary of the Makgadikgadi Pans National Park (MPNP), coincides with the dry-season range of the migratory zebra and wildebeest populations. During the dry season, environmental constraints on foraging behaviour are at their peak and wildlife within the Makgadikgadi is reliant on the waterholes situated near the fence. Therefore, the project is focusing its data collection on the dry season when any impact caused by the Makgadikgadi fence will be at its greatest.

The Makgadikgadi Zebra Migration Research Project sets out to examine and quantify the ecological impact of the 2004 fencing of the Makgadikgadi ecosystem, with a specific focus on the migratory zebra and wildebeest populations. To achieve this, data examining the spatial distribution, foraging behaviour and population dynamics of the zebra and wildebeest herds will be collected, analysed and compared with pre-fence data.

These data will enable the researchers to determine the impact of competition between livestock and wildlife for grazing and water resources. The project also looks to determine the current size of the zebra and wildebeest populations as well as the predation rates of lions upon these herds.

The fencing of the Makgadikgadi system provides a rare opportunity in ecological research to investigate, through partial experimental design, the impact of a large fencing policy on the fluid dynamics of a large mammal community. The results of the project could be used to provide a model and strategy for future fencing policies in conflict situations in other parts of Botswana, and Africa.

The project aims to gather data in the field for two full dry seasons over a three-year period and will form part of a PhD research project awarded by the University of Bristol. After its completion it is hoped that a long-term monitoring project will continue to assist with management objectives for MPNP.

The project is in partnership with the Department of Wildlife and National Parks (DWNP) of Botswana and the University of Botswana’s Harry Oppenheimer Okavango Research Centre (HOORC).

The study is focusing on data collection during the dry season (typically from March/April to October/November) when the migratory zebra population is resident around the waterholes provided along the Boteti River. During this time there is no surface water available within the park and all water requirements of the migratory herds must be satisfied by these waterholes. In the wet season, rainfall initiates the migratory movement of the zebra and wildebeest to the open grasslands of the Makgadikgadi to the East of the Boteti River.

The electrified Makgadikgadi wildlife fence is aligned along the western and southern boundary of the MPNP, following the path of the Boteti River, which forms the traditional boundary of the National Park. The fence now forms an artificial boundary to the dry season home range of the zebra and wildebeest populations and may significantly affect dry season behaviour. The dry season is also the limiting season for the migratory population and the fence may have alleviated or aggravated these limiting factors.
Therefore, the project’s energy and data collection will be focused on the dry season to determine the impact of these factors and provide DWNP with the most pragmatic and cost-efficient management approach to conserve this migratory population.

10 zebra will be fitted with Global Positioning System (GPS) collars in the first year of the project to identify large-scale patterns of zebra movement across the Makgadikgadi and smaller scale patch selection strategies, as well as to identify macro and micro grazing patches, distances covered and speeds travelled by each zebra to find these grazing resources. Using activity sensors and correlating these data with observation data, the collars can also be used to identify the length of time devoted to each major activity (grazing, moving and resting) and how these change through the season. These data will be compared with similar pre-fence data to help determine how movement patterns and foraging strategies have changed.
Ten wildebeest mares will be collared with VHF collars to monitor movement patterns. As zebra are the keystone species within the system there is a greater focus on their ecology. However, VHF collared wildebeest will be tracked on a monthly basis to identify home ranges and patch selection strategies and how these have changed since pre-fence conditions.

All collars will be used to track known individuals during fieldwork, enabling direct behavioural observations to be made and identification of selected grazing sites. Collars will also be used to assess mortality rates within the migratory population.

Water samples will be taken from all pumped waterholes located within the dry season range of the zebra population. Samples will be collected monthly and analysed to assess changing water quality throughout the dry season to determine if this affects patch selection by zebra in surrounding grazing sites.
Aerial censuses of zebra and wildebeest herds are being conducted at the beginning and end of each dry season. This will be combined with ground surveys to accurately determine population size and yearling recruitment.

Community
The project will also assess the response of local communities to the installation of the fence and look at whether the fence is perceived as having had a positive or negative impact on the local area, with specific reference to grazing availability for cattle and livestock predation by predators.

Prior to the erection of the fence, cattle were fitted with GPS collars to determine their foraging patterns and resource use both inside and outside of the National Park. The project will continue with this approach and place collars on cattle to see how the fence has altered foraging patterns around family homesteads. Household questionnaires throughout the Boteti region will provide detailed information on the economic importance of cattle to the rural community and help quantify the impact of the fence on these communities.
The project will work in collaboration with a local safari operator to help diversify local livelihoods. The fence provides the potential for local communities to benefit from the proximity of the migration, without suffering from the consequences of the human-wildlife conflict. Researchers and safari operators will work with these communities to help develop community-run campsites overlooking zebra-filled waterholes, with this project focusing its attention on capacity building and guide training – teaching potential guides about the migration and ecology of the Makgadikgadi.

May 2011
The Makgadikgadi Zebra Migration Research project was initiated in 2001 in response to a plan to fence the Makgadikgadi. After the initial phase of the research was completed in 2005 the project restarted in 2008 in order to assess what impact, if any, the establishment of the Makgadikgadi fence has had on the zebra population within the Makgadikgadi. GPS data available from collared adult zebra mares during 2009 and 2010, as well as the detailed sampling of grazing and water resources and behavioural observations, allows for the analysis of current trends and for a comparison with pre-fence data.

In the current post-fence conditions, zebra return to the Boteti River to drink more frequently than they did prior to the fence (3 days between drinking post-fence, 4 days pre-fence), but they are still walking large distances in order to graze. During 2010 collared zebra have been regularly recorded more than 15km, up to a maximum of 30km, from the Boteti River during their search for grazing. It will be interesting to see how the analysis of pre and post-fence grazing resources compare.

The search for grazing was heavily disrupted in early September when a bush fire burnt across the Makgadikgadi in just over 24 hours, burning nearly all of the available grazing resources within 3 000 km2 of the 5 000km2 Makgadikgadi National Park. Since the fire it has been noticeable that zebra are actively choosing to graze in areas which were burnt, favouring the fresh green shoots which have grown rapidly.

The project will continue with fieldwork until the end of March 2011 when all of the collars will be removed from zebra and the final analysis of data will begin. The initial assessment suggests that the Makgadikgadi fence has had a positive impact on the long-term health of the Makgadikgadi zebra population.

March 2011
After a slow start to the rainy season, the zebra migrated east into the open grasslands during December. Throughout January zebra were to be found on the edge of the pans before moving north into the mixed woodland habitats along the edge of the National Park and into the CT 11 area where they currently remain. While January received some significant rainfall, February was very dry yet the seasonal waterholes remain full. Depending on further rainfall it is expected that the zebra will remain in the eastern Makgadikgadi throughout March and April.

We continue to collect field data from the Makgadikgadi. Grass samples from preferred and ignored grazing sites have been collected to look at the nutritional choices of the zebra. Faecal samples have also been collected to assess moisture content and crude dietary protein. Behavioural observations and population counts are also conducted each month in order to continue with the ongoing data collection in these areas.

The Makgadikgadi Fence
The Makgadikgadi fence between Khumaga and Tsoe is in an appalling condition with parts of it lying flat on the ground in many places and numerous large gaps in the fence which allow for wildlife and livestock to pass unimpeded through the fence in both directions. These gaps have been created by the movement of elephant trying to access the Boteti River which is flowing on the other side of the fence.

While driving along the fence line Elephants were seen passing through the fence heading both towards and away from the river. Fresh lion tracks were also seen passing through more than one gap in the fence. Rhino tracks were seen near a gap although it seems that it not pass through it. Cattle were also seen to be grazing inside the National Park having passed through one of the gaps in the fence.

One of the principal reasons for the establishment of the fence along the western and southern boundaries of the Makgadikgadi National Park was to create a physical barrier between wildlife and livestock populations in order to reduce human-wildlife conflict.

A comparison of pre- and post-fence conditions and movement patterns is one of the aims for the current research. However, one of the preliminary findings of the Makgadikgadi Zebra Migration research, based on observations, is that the exclusion of cattle from the National Park along the western boundary has had a positive effect on the zebra and wildebeest populations.

Prior to the erection of the fence, cattle would dominate the water points within the Boteti riverbed and despite having travelled up to 20km from their grazing areas in order to drink the zebra would not rest in the area. After the completion of the fence, zebra were seen to rest in the riverbed with reduced competition for the limited water resources before heading back out to their grazing areas. The year-round presence of cattle also reduced the grazing resources available for zebra and wildebeest when they migrated back to the Boteti region during the dry season.

As well as removing valuable grazing resources prior to the return of zebra and wildebeest during the dry season, these cattle are also heavily exposed to predation from lion.

In its current condition, the fence can no longer be seen as being an effective barrier and the level of human-wildlife conflict will continue to increase. It is important therefore that a solution to the fence is achieved in the near future so as to prevent the situation deteriorating even further. The return of the Boteti River has complicated this issue with elephant wanting to get to this source.

Work planned for the next reporting period:

• The Removal of all GPS collars in the Makgadikgadi.
• An aerial survey of the Makgadikgadi zebra and wildebeest population will take place in order to accurately assess population sizes.
• Thorough examination of the collected grass and faeces will take place to analyse the movement of zebra in the Makgadikgadi and their nutritional requirements.
• A report will be submitted with the data and relevant research from the first phase of the project highlighted by the conclusions of the current status of the Makgadikgadi zebra population as well as the impact of the Makgadikgadi fence installation.
• Recommendations for the ongoing management of the Makgadikgadi and for future research will be made.