Okavango Large Herbivore Ecology Project

Region: Okavango Delta, Botswana

The Large Herbivore Ecology Project focuses on gaining detailed information of present seasonal population densities, demographics and distribution of the key large-bodied herbivores – Cape buffalo, African elephant, blue wildebeest, giraffe, greater kudu, impala, Burchell’s zebra, red lechwe and tsessebe – across the southern Okavango Delta. This is being complimented by gathering detailed season, habitat and area specific data on the available resource characteristics, local flooding regime and fire incidence. It is hoped that these data will increase our understanding of the current state of the Okavango Delta’s herbivore population, their limiting factors and how they utilise the Okavango system; allowing ecosystem managers to identify and thus protect critical factors necessary for their survival.

The Okavango Delta is a large land-locked alluvial fan located in the Kalahari Basin, in north-east Botswana. It is made up of a rich mosaic of habitats and covers approximately 22,000km2, making it the largest inland delta in Africa. The ecosystem is driven by an annual flood that brings water and initiates forage growth during the dry season, allowing the area to support a wide variety of animals at densities only surpassed within southern Africa by Kruger National Park. High levels of tourism within the Delta make the area vital for Botswana’s economy, bringing in revenue and providing employment.

Despite the policies that aim to protect the Okavango Delta from detrimental human interference, the ecosystem remains under threat, internationally from increased water extraction in Angola and Namibia (even an abstraction of 1% would alter the habitat mosaics due to the shallow nature of flooding), as well as increased human pressure from communities and tourists (seasonal burns to increase game viewing success modifies habitats whilst human disturbance can reduce reproductive success of fauna, as already demonstrated for the wattled crane, one of the region’s rarest birds).

These threats are compounded by a lack of knowledge about the community ecology of many of the ecosystem’s large-bodied mammals, especially herbivores. A more accurate assessment of the population dynamics of the dominant medium-to-large sized mammalian herbivores needs to be gathered, along with improved information about their seasonal habitat preferences, resource selection strategies and movement patterns, if effective long-term management strategies are to be implemented. It is crucial to ascertain how potential changes in flood dynamics will affect the spatial distribution, habitat selection and population structure of the Delta’s herbivore populations.

The primary goal of the project is thus to provide scientific and holistic ecological data that can be used by ecosystem managers in long-term active management strategies and to produce an interpretive assessment of the potential environmental impacts of the Delta’s principal conservation issues. The information provided by the project on the present population demographics, assembly patterns, preferred resources and habitats, movement patterns and spatial distribution patterns of medium-sized herbivores within the southern Okavango Delta will dramatically increase our understanding of what factors are of primary importance in regulating the system’s herbivore population.

Through the study, population demographic data will allow us to identify herbivores with abnormal population structure, a good indication of population instability. By identifying the herbivores at most risk, more detailed research into causes and possible mitigation strategies can be initiated at an earlier stage and species specific management plans developed.

The information that this project will allow Department of Wildlife and National Parks (DWNP) and NGO ecosystem managers to understand what within habitat characteristics large-bodied herbivores are selecting for and how they move between these resources. Such understanding is critical to ensure that areas or habitats determined as of critical importance to the long-term stability of the system’s herbivore population are protected. This information will also be used to create a fauna layer for a multi-disciplinary dynamic systems model that is being developed at the Harry Oppenheimer Okavango Research Centre (HOORC). The herbivore layer will allow us to model how populations vary with flooding and vegetation and will therefore be vital in enabling predictions to be made on how changing flood regimes, due to either natural or human consequences, will affect mammalian populations.

During the course of this study, a long-term population demographics monitoring programme will be developed. This will be used to complement the annual DWNP aerial surveys, allowing population trends to be accurately established.

Increasing understanding of the Okavango Delta ecosystem and how vulnerable it is to disturbance is an additional target of this project. Education is vital, with content and delivery method being modified according to the target audience. Formal presentations are made twice a year at the HOORC, which are open to the public and interested parties, including DWNP officials, academics from University of Botswana, employees of NGOs and other stakeholders are personally invited. Less formal workshops are also held twice a year for students attending the Botswana Wildlife Training Institute in Maun and are also held for groups of school and university students visiting from abroad. When requested, informal talks and workshops are also given to individuals working within the Delta ecosystem and to visitors to the area.

Project Findings (interim) – December 2008
I have now finished fieldwork for this particular project and am presently in the UK writing up the research for a thesis and publishing the results. The project was very successful and I feel we collected data that has greatly increased our understanding of herbivore distribution in this complex system (such as the critical resource species/characteristics for herbivore distribution, effect of spatial memory, impact of seasonal variation of home range location in general herbivore density) – however like most research projects I have also uncovered some other questions that need answering.

The most unexpected outcome of the work on zebra movement was the discovery of an intact zebra migration between the peripheral Delta (i.e. southern Moremi, Santawani) to the Makgadikgadi Pan grasslands. This migration was made by a significant number of my collared animals and at an approximately 580km round trip, it is the second longest zebra migration in existence. It is made even more exciting by the fact that the zebra have only recently been able to restart this migration due to the removal of the Nxai Pan veterinary fence – it therefore has implications for other ecosystems where wildlife corridors are being erected as it suggests that terrestrial mammals are capable of recommencing historical movements if routes are reopened.

I am presently publishing this data in an academic journal. However, we still have a lot more to find out about the migration – how many animals do it, does the route vary each year, is it a distinct population, the driving forces, and many more! In the immediate term I am looking to re-collar migratory individuals (and ideally increase the number of collared animals) later this year to ensure that we keep track of the movement through known individuals and to conduct a population estimate at the commencement of the movement next November.

Quarterly Report – May 2008
The last three months of the project have been successful, with field research undertaken in both the Self-drive Moremi and Mombo/Chief’s Island study areas. Significant rain fell in January and February, initiating wet season vegetation growth and filling of seasonal pans across the study areas. Wet season data collection was completed by March.

As was expected, habitat usage by the zebra varied significantly once enough rain had fallen to fill seasonal pans and to initiate vegetation growth. In both areas, utilisation of grassland and open acacia habitats increased during the wet season. My first year results show that during the wet season the annual grass growth is maximal in these habitats; zebra may therefore have moved into these areas to select these nutrient-rich annual grasses.

Although both areas showed similar patterns, floodplain usage in Mombo remained significantly higher than in self-drive. Data collected in Year One demonstrated that the grass composition and quality differed significantly between the areas. On floodplains, the Mombo study area floodplains had higher amounts of C. dactylon and P. repens and higher ground cover. These floodplains were also flooded for more sustained periods of time and as such have greater re-growth when the flood resided. These factors may have explained to some extent the differing habitat usage between study areas.

The differing wet and dry season home range was especially distinct in the self-drive study area. Cross-over between seasonal home ranges was minimal; especially in terms of zebra visiting the core wet season home range in the dry season. As discussed before, this may be due to forage availability. However, studying the geography of the wet season home range the role of water availability also becomes obvious; when seasonal pans and water holes are dry zebra would have to walk many kilometres between foraging and drinking areas, making such areas poor energetic choices in the dry season.

Only two of the Mombo collared zebra (z3668 and z3685) showed distinct seasonal differences in home range area, although as discussed habitat utilisation did differ. These zebra moved further to the south during the wet season, utilising the more expansive open acacia woodland and grassland areas found to the south of Mombo Island and on Chief’s Island. The remaining collared zebra (e.g. z3747) remained in the north of the study area, utilising much the same area as during the dry season. These differences may be due to differences in quality and quantity of preferred forage compared to the self-drive study area, meaning that zebra (and other grazers) can be supported here throughout the year.