Wednesday, 10 July 2013

Summary of results

Following the conclusion of fieldwork there were lots of data to analyse which took some time. However, the analysis is now complete and a summary of the results is posted below. If anyone would like to receive a full pdf of the final thesis then please email me at and I will happily send this to you. This thesis was submitted for a PhD at the University of Bristol and the PhD was awarded to me in May 2013.I am currently working on a few papers emanating from the research for publication in peer-reviewed scientific journals.

I have recently returned to Botswana to work with the CKGR research team studying predator-prey interactions within the Central Kalahari Game Reserve and the Makgadikgadi Pans National Park. Once things have settled down I will tell you a little bit more about the work that we are doing and what our current priorities are. If you have any questions or would like to contact me then please email me at

Thank you to everyone who helped out in anyway during the research, your help was greatly appreciated.

Dr James Bradley

This research was initiated as a follow-up study to the research of Dr Chris Brooks conducted between 2001 and 2005. The research examined the foraging behaviour, spatial distribution and adaptability to environmental changes of plains zebra (Equus quagga) in the Makgadikgadi, Botswana, following the construction of an electrified fence in 2004. Seasonal changes in resource availability were documented and GPS collars were used to record detailed movement data. The ongoing population dynamics of the zebra population were recorded throughout the study and the impact of the fence on the migratory zebra population was assessed where possible.

The results of this study show that zebra are highly adaptable and have the flexibility to respond to significant environmental changes, yet they need to continue to be able to move freely within the Makgadikgadi. Initial results suggest that the Makgadikgadi fence has had a positive impact on the zebra population but further monitoring is needed to assess the long-term impacts of fencing on the Makgadikgadi ecosystem and the resident wildlife. These findings add to those of Brooks (2005) and further our understanding of the resource requirements, spatial distribution and foraging behaviour of zebra in the Makgadikgadi.

During the research period, the Makgadikgadi was subject to significant environmental variability which influenced the spatial distribution and foraging behaviour of zebra. The Makgadikgadi experienced significant unseasonal rainfall in June 2009 which had a major effect on the location of the zebra herds within the Makgadikgadi. In addition, the Boteti River started to flow again in 2009 after a 20 year hiatus, significantly changing water availability for wildlife and livestock. Finally, a large bush fire in September 2010 removed nearly all of the available forage biomass, significantly influencing movement patterns and foraging behaviour. In addition to these one-off events, the Makgadikgadi experienced above average rainfall from 2008 to 2011 which led to increased forage growth across the Makgadikgadi when compared to the pre-fence study period.

Seasonal water availability determined the spatial distribution of zebra in the Makgadikgadi while forage quality and quantity was both spatially and temporally variable. The movement patterns of zebra reflected resource availability at multiple spatial scales and showed that zebra follow an area restricted search strategy. By adapting fine-scale foraging patterns to the quality and quantity of resources available, zebra were able to improve foraging efficiency. Zebra adopted an unselective foraging strategy at the feeding site, maximising intake rate and reducing the temporal and energetic costs of foraging during both a typical wet season and an atypical dry season. However, sites with increased forage dry matter were preferred in two of the three available habitats during the typical wet season. Pan grassland provided the highest quality forage with zebra taking advantage of fresh rains to move further into the pan grassland to forage. The mixed woodland and pan grassland habitats in CT/11 were used extensively throughout the wet season and were particularly important towards the end of the wet season as waterholes dry up.

When a large bush fire passed through the Makgadikgadi in early September 2010, environmental conditions changed overnight, yet zebra were able to adapt their behaviour to minimise the impact of the fire. They were required to extend their drinking interval, travel further and work harder to meet nutritional requirements. All of the collared zebra followed a similar post-fire strategy suggesting that zebra have learnt to be highly adaptive to the challenges faced within the semi-arid Makgadikgadi. Yet, despite the foraging restrictions caused by the fire, zebra were not required to push their physiological limits as much as was necessary during the long dry seasons of 2002 and 2003 (Brooks 2005). However, the impact of the fire may continue to be felt in the longer term as it may take 2-3 years for phytomass to return to pre-fire levels. It is also unknown how the increased demands placed on zebra after the fire will have affected survival rates of both adults and foals.

Continuous GPS data collected from collars fitted to zebra made it possible to determine how environmental characteristics influence the foraging behaviour of a large herbivore. First-passage time analysis (FPT) of recorded GPS data made it possible to identify discrete movement paths and identify foraging patches. This analysis showed how the foraging behaviour of zebra within feeding patches reflected seasonal differences in resource availability and quality.

Comparison between pre-fence study and current study
During the pre-fence study, zebra were recorded pushing their physiological limits; drinking on average every four days and foraging up to 35km from the Boteti riverbed (Brooks 2005). It was hypothesised that the construction of the Makgadikgadi fence and removal of livestock from the MPNP would influence zebra movement and foraging behaviour.

The current study found that zebra decreased their drinking interval to every 2-3 days but with occasional longer intervals of 4-5 days. However, while zebra foraged within 5km of the riverbed, something they did not do prior to the fence being constructed, they continued to select forage areas that were 15-20km from the Boteti River.

Before the fence was built, zebra were reliant on natural water seeps and two pumped waterholes in the riverbed for drinking water. However, these waterholes were also used by livestock and so competition for water was high. Following the erection of the fence and the consequent exclusion of farmers and livestock, zebra freely enter the riverbed throughout the day, even stopping to rest in the riverbed; something that did not happen pre-fence. The return of the Boteti River has provided abundant fresh water, resulting in zebra spending less than 0.5% of their time within 100m of the riverbed compared with nearly 5% of their time during the pre-fence study (Brooks 2005). Furthermore, while zebra are still vulnerable to predation around the riverbed, surplus killing is no longer evident. However, the observed changes cannot be attributed solely to the fence as forage and water availability were significantly different between the pre- and post-fence studies.

Ground surveys were conducted during both the pre-fence study and the current study to record the population dynamics of zebra in the Makgadikgadi. These surveys focused on the recruitment of yearling zebra (1-2 years old) to the population as an indicator of population health. In 2003, there was a yearling recruitment rate of 16±3 (SD) yearlings per hundred adult females, while the current study recorded 23±4 yearlings. The yearling recruitment in the stable zebra population in Kruger National Park, South Africa was 17 yearlings per hundred adult females, suggesting that the Makgadikgadi population is currently in a healthy state and may even be growing. However, further monitoring is required to determine whether this high recruitment rate is maintained.

The Makgadikgadi fence was designed to be a physical barrier to separate wildlife and livestock yet, in its current state, it is highly permeable. Over time, multiple crossing points through the fence have been established by elephants which have allowed unrestricted access to the Boteti River. These crossing points have subsequently allowed cattle and donkeys to enter the MPNP once more. The fence appears to have had a positive effect on the behaviour of zebra however. Restricting livestock access and the resulting removal of inter-specific competition for grazing resources may even be contributing to improved yearling recruitment. With this in mind, it is to be hoped that the fence can be modified and rebuilt in accordance with the Makgadikgadi Management Plan recommendations.

Brooks, C.J. (2005) The foraging behaviour of Burchell's zebra (Equus burchelli antiquorum). PhD Thesis, University of Bristol.