Hwange Elephant Movement Study

The overall aim of the project is to gain a better understanding of elephant movements and habitat use in Zimbabwe’s Hwange National Park, where elephant density is particularly high. This will be achieved by studying the relative contribution of resources (water and food), social interactions (dominance hierarchies) and predation risk (by lions) to this key aspect of elephant ecology.

Country: Zimbabwe

It is essential that we learn more about elephant movements and habitat use in order to gain a better understanding of their impact on ecosystems. Previous studies have emphasised the role of water in constraining elephant ecology and habitat preferences are reasonably well understood. However, little is known about the ways in which elephant competition influences use of waterholes and habitat. This is critical in order to better understand and predict elephant dynamics, particularly in areas with high-density populations, such as Hwange National Park.

Results from new studies in Hwange are revealing the constraints imposed by the scarce distribution of waterpans and the likely intense competition that elephants face when choosing where to drink or forage. These choices are likely to be affected by strong inter-group dominance which has previously been reported in elephants. The social influence on elephant movements and habitat use remains relatively unknown, with only Wittemyer et al recently showing that dominant herds in Samburu, Kenya, were able to monopolise the best foraging areas and walked less than subdominant ones. This issue needs to be clarified in order to understand how elephant movements and habitat use change when elephant density increases, and therefore competition increases too.

Although it is likely to have a low impact on elephant demography, regular predation by lion may lead elephant to respond behaviourally (e.g. avoidance) to lion presence. It has been shown in other herbivores that anti-predator spatial responses may override any other drivers of movements or habitat use and this needs to be studied in elephant.

The main objective is to determine how higher elephant densities and individual dominance status may influence how far from water elephants will go to forage, how this influences the quality of the patch in which they forage, and how easily they can access water. The following actions will be taken:

1. Collar 10 elephant cows with GPS collars.
2. Determine the dominance status of the study’s herds using field observation.
3. Determine resource availability and quality by regularly monitoring water presence in pans and quality of foraging patches. Also, determine local elephant densities using censuses.
4. Determine movement parameters and habitat use through elephant GPS-tracking data.
5. Determine predation risk by using field observation of lion presence and possibly lion GPS-tracking data.
6. Build comprehensive models of elephant movement and habitat use which integrate all the above-mentioned aspects (resources, dominance status, and predation risk) to disentangle their relative contributions.

1. Ten elephant cows will be fitted with GPS collars with satellite (Iridium) transmission. Collaring will be conducted in the Hwange area. Elephants collared by the French CNRS team in 2012 will also be studied, particularly those in the Ngweshla area. Acquisition of location data will be done at a frequency of 1 location / 30 min, which has shown to be sufficient to detect visits to water.
2. The dominance ranks of the studied herds will be assessed using field observations of interactions between elephant at waterholes. Wittemyer’s approach, which focuses on physical (with contact) and non-physical (no contact, threat only), agonistic interactions (fight-related), will be followed.
3. Starting from the early dry season (June), all artificial and large natural pans will be monitored every month to track water availability across the study area. Road and waterhole censuses will also be conducted to estimate local elephant densities.
4. GPS data will be analysed to identify drinking events, distance travelled, foraging places (using the residence time method) and their distance to water.
5. Vegetation surveys will be conducted at foraging places identified using the GPS data. Transects will be used to estimate woody cover density and composition, and intensity of elephant browsing. Associated with the literature on elephant diet preferences, this will allow the estimation of the quality of foraging patches. A 30m x 30m vegetation map from the CNRS research group will also be used.
6. The presence of lion in the study area will be recorded by using information from guides, and possibly by liaising with the lion project (WILDCRU team, Oxford University).
7. Finally, all the above-mentioned information will be transferred into statistical models (e.g. resource selection functions) to relate elephant movement and habitat use to water and food availability, local elephant densities, dominance status of the herds, and predation risks.

May 2019
Much work was spent over the year on analysis of data and drafting a way forward with regards to the thesis titled “Revealing Partial Migration in Elephants in Hwange National Park”. We aimed at replacing the old elephant collars towards the end of the 2019 dry season.

The project is currently in its second phase under the title “Assessing the significance of large herbivore migrations in Hwange National Park”. In this broadening of the project, in addition to elephant, wildebeest, zebra, and giraffe have been collared. Using camera traps, we have investigated the drivers (habitat structure and composition, demography and competition) of the animals’ movement strategies and shed more light on wildlife numbers in areas that we have less access to.

Results show that partial migration exists in elephant in Hwange, with three movement strategies (resident, short-distance migrant and long-distance migrants). In the wet season, long-distance migrants have larger home range sizes than short-distance migrants and residents, while there was no difference in home range fidelity or in the distances covered between the three groups. In the dry season, long-distance migrants still had larger home ranges and covered more distance than the other groups; however fidelity for these ranges was low with residents having a higher fidelity for the dry-season home range.

Migration remains a primary driver of population growth and in a changing environment the conservation of migration traits in wildlife are key. The key insights that we have shown include the role of water availability in driving wet and dry season movements of elephants in HNP. As soon as rainfall starts in November elephants begin the movement to the wet season range. This is key in the management of elephant populations in the future landscape changes in HNP. The landscape approach to conservation is also shown in that some of the wet season ranges were in Botswana. This shows the importance of large protected areas in the preservation of a species such as the elephant. This speaks volumes in an area which has established the largest Transfrontier conservation area (KAZA TFCA).

Annual Report, June 2017
The collars worked efficiently through the year with one having problems in November. The maximum distance between the two furthest points in elephant movements was calculated as the basis of the elephant status grouping: 4 residents, 2 intermediate movers and 3 migrants. Elephants moved to their wet season home range in November due to the early rains. As previously seen, some of the elephants went into Botswana and back, while some moved into the central part of the Park.

Home ranges (50 and 95% kernel density estimates) for the wet season (Feb-Apr) and dry season (Aug-Oct) for the three elephant status groups were established while the rest of the months were termed as “transition periods” – i.e. elephants are in transit hence spend less than 50% of their time in these areas. The wet season home range was larger than the dry season home range; this is expected since in the dry season water becomes a limiting factor for elephants and so they use a smaller area (usually containing pumped pans).

There was a distinction noted between the home range of migrants and those of resident elephants with migrants having a larger home range. Most of the migrants use a larger area within each season – hence the barriers of space use in any given landscape and how these shape movement patterns have to be investigated further. It was also noted that elephants have a high fidelity to home ranges used in the previous season. Migrants have a lower overlap than resident elephants between the two seasons.

Annual Report, February 2016

Satellite data collected has undergone analysis to show some key components of elephant movements, which include timing and patterns. Such movement patterns show a wide variation, with some individuals covering long distances in the wet season, remaining for long stretches at the Park boundary, and returning only in the dry season, while others move off rapidly at the start of the wet season and return to dry season patterns quite quickly.

It is also evident that the Wilderness Concession at the south-east of the Park is being utilised heavily by collared elephant during the dry season, with one individual spending 50% of its time at the Mbiza Pan. Many elephants make use of the localised waterholes from which to drink.

In general all elephants seem to show some form of movement after the start of the rains in the wet season. The variations again are wide, with some moving as far away and for as long as possible from the reference waterholes (elephants 2 and 8), while others move off but return regularly (elephants 9 and 10). From the month of July onwards, movements become localised with most of the elephants moving only about 20 km from the reference waterhole. It will be interesting to see if the same pattern is maintained in 2016.

The element of competition has proven to be rather difficult to follow; however we have changed this so as to understand the component of competition at waterholes specifically. At the same time we would like to understand if migration has a demographic impact by studying the age structure of herds in the month of March and April this year and next year in different locations.

November 2015
Two more elephants were collared. The details appeared in a press release by Wilderness Safaris:

Press Release: 10 Elephants Collared in Zimbabwe for Hwange Movement Study

A total of 10 elephants have been collared in Zimbabwe’s Hwange National Park as part of a study that aims to assess the impact social dominance has on elephant movements and habitat use. The study, which is being funded by the Wilderness Wildlife Trust, focuses on areas within Wilderness Safaris’ concessions in the Park, and has been initiated and carried out by Wilderness Safaris Environmental Officer for the Zambezi Region, Arnold Tshipa.

By tracking elephant movements, the study aims to gain a better understanding of the effect elephant dominance has on herds’ behaviour and the impact this has on ecosystems. This has particular relevance in areas such as Hwange, where there are high-density elephant populations as well as high demand for resources such as food and water.

“Recent studies have shown that the scarce distribution of water sources results in a high level of competition between elephant herds when choosing where to drink and forage. What remains unknown is the impact social influence or competition has on elephant movements – an important issue to investigate in ecosystems where elephant densities are increasing, such as Hwange”, says Tshipa.

The first eight elephant were fitted with satellite collars in 2014, while the final two were collared in late September 2015 – both females, one with a herd of five members and the other with a herd of seven.
“We have already noted some interesting findings”, says Tshipa. “Of the initial eight elephant collared, four have shown to be ‘movers’ with one going right down to the border between Botswana and Zimbabwe. The rest have moved only locally, within Wilderness Safaris’ concessions and the park boundary, going towards the communal areas”, he adds.

The next phase of the project is already underway and will include monitoring elephant interactions at waterholes. The study is just one of a number of Wilderness Safaris conservation projects underway in Hwange, ranging from water supply and anti-poaching programmes, to regular game counts and bird species monitoring.

March 2015
Of the eight elephants collared, two have shown substantial movement, with one travelling all the way down to the boundary of Zimbabwe and Botswana. The rest have moved locally within the Wilderness Safaris concession and park boundary, heading east towards the communal areas. Fieldwork will be conducted from the beginning of the dry season, which will include the monitoring of elephant interactions at waterholes. The project aims to collar an additional two elephant early in the dry season in the south-eastern part of Hwange National Park.

November 2014
The project has deployed eight of ten GPS elephant collars. Elephants were found in Airstrip 2, Ngweshla and Scott’s Pan. The remaining two collars will be deployed during February 2015.