Elephants, bees and trees

Testing the effectiveness of a novel mitigation method for protecting marula trees against elephant impact

Robin Cook (MSc. candidate)

Supervisors and advisors: Prof Francesca Parrini; Prof Ed Witkowski; Dr Michelle Henley; Dr Lucy King

Human-elephant conflict (HEC) can occur where humans and elephants compete for the same resources, usually along the agricultural interface. In South Africa however, most African elephant (Loxodonta africana) populations are found within enclosed protected areas, thereby preventing them from coming into contact with humans. In this management setup, HEC takes on a new form, where concerns are raised over the impact that high elephant densities have on large trees. Whether the reasons be ecological or aesthetical, extensive elephant impact on large trees is generally not received well by humans. One tree species of particular concern is the marula tree (Sclerocarya birrea subsp. caffra), a large economically and ecologically important tree species in South Africa. It is certainly not uncommon to find marula trees in South Africa’s Greater Kruger National Park (Greater KNP) that have been bark stripped, are missing branches, or have been pushed over. As South Africa’s elephant management policy focuses on managing the effects of elephants, instead of elephant numbers alone, mitigation methods are therefore required to protect important tree species from elephant impact.

A mitigation method that is currently being used in areas of the Greater KNP by Elephants Alive is wire-netting. Wire-netting involves wrapping chicken-mesh around the main stem of a tree in an attempt to prevent elephants from bark-stripping (Derham et al. 2016). Whilst wire-netting has proven to be highly effective at preventing bark-stripping, these trees are still vulnerable to being pushed over or having their main stems snapped by elephants. It is therefore important to test new mitigation methods which can be added to the ‘toolkit’ for protecting trees from elephant impact. One such method used to alleviate HEC is the African honeybee (Apis mellifera subsp. scutellata). Vollrath and Douglas-Hamilton (2002) were the first researchers to demonstrate that African honeybees could be used to protect vegetation from elephant impact. They hypothesised that the sensitive areas of an elephant’s skin around the eyes and ears, as well as the moisture in its trunk, would make it vulnerable to bee-stings. Dr Lucy King then designed beehive fence-lines to keep elephants out of crop fields in Kenya, thereby reducing HEC (King et al. 2017). The aim of our research was to test the effectiveness of African honeybees (and beehives) as a mitigation method for elephant impact on marula trees, and compare its effectiveness to wire-netting.

Our research took place in Jejane Private Nature Reserve (JPNR), a recently adjoined protected area in the Greater KNP. In the 3 years that JPNR has had elephants in the protected area, almost a quarter of the marula tree population have been killed. With such high marula mortality rates, this was the ideal scenario to test the effectiveness of beehives as a mitigation method. Our research was conducted on 150 marula trees in a 30 ha site, where we divided the trees into 50 beehive, 50 wire-netting, and 50 control trees. We built 100 beehives, separated into 50 active and 50 dummy beehives. As previous research has found that single active beehives, or only dummy beehives, are not effective against all forms of elephant impact, our designed aimed to test whether placing 1 active and 1 dummy beehive in a single tree would be effective against elephant impact, lowering the costs of maintaining 2 active beehives.

Having built the beehives, we proceeded to hang nylon ropes from the branches of the 50 beehive trees and attached 1 dummy beehive to each tree’s ropes. Then in an overnight effort, we transferred 50 live honeybee colonies into our active beehives and hung them on the opposite sides of the each beehive tree’s dummy beehive.


For wire-netting, we wrapped chicken-mesh around the main stems of 50 trees in a double-wrapping method, creating a rigid cage-like structure around the tree. The ends of the chicken-mesh were then stapled to the main stem itself.

We monitored elephant activity on all 150 trees, both prior- and post- treatment addition, for 9 months. We recorded any new bark-stripping, primary and secondary branch breakage, main stem snapping, and uprooting on the trees. Furthermore, because of the drought conditions in South Africa, we built and installed feeding stations across the study site, giving the honeybee colonies access to sugar water, nectar, and pollen substitute.

Elephants moved through the study site on numerous occasions, impacting 27 control, 14 wire-netted, and 1 beehive tree. The wire-netting was effective against bark-stripping, but did not stop elephants from breaking branches. The beehives, however, appeared to be highly effective against all forms of impact. Elephants were observed approaching beehive trees, pausing, and then moving off in a new direction. Furthermore, the only beehive tree to receive elephant impact had some secondary branches broken when the dummy beehive was ripped out of the tree, potentially by an aggressive musth bull observed 3 days later in the study site.

Our results suggest that 1 active and 1 dummy beehive in a tree can be highly effective at protecting the tree from elephant impact. However, the usage of beehives may be restricted to small-scale usage on trees of conservation concern due to the high costs and maintenance required for setting up and sustaining the beehives. In comparison to wire-netting, the beehive mitigation method was just over 20-times more expensive in a single year. However if combined with a financial scheme of harvesting and selling honey produced from the beehives (not done in this study), then the beehive mitigation method can be used as both a means of protecting large trees, as well as creating financial revenue for the protected area involved.

We would like to say a special thank you to Woolworths-South Africa, Relate, and Elephants Alive for their support during this study, as well as to numerous private donors who helped sponsor the manufacturing, upkeep costs of the beehives.


Derham, K., Henley, M. D., & Schulte, B. A. (2016). Wire netting reduces African elephant (Loxodonta africana) impact to selected trees in South Africa. Koedoe, 58(1), 1-7.

King, L. E., Lala, F., Nzumu, H., Mwambingu, E., & Douglas‐Hamilton, I. (2017). Beehive fences as a multidimensional conflict‐mitigation tool for farmers coexisting with elephants. Conservation Biology.

Vollrath, F., & Douglas-Hamilton, I. (2002). African bees to control African elephants. Naturwissenschaften, 89(11), 508-511.

Previous Entries African wild ass in Eritrea - a tale from the field Next Entries The future of African ecosystems in the balance