African Elephant Feeding Preferences in a small South African fenced game reserve

Abstract Recent significant changes to African elephant ( Loxodonta africana ) populations have placed their management at the centre of the global wildlife conservation management debate, with most of the discussion surrounding population declines. However, in small, invariably fenced reserves found over much of South Africa, their numbers have expanded due to elevated levels of protection from poaching, artificial water provision, high reproductive success and lack of dispersal opportunities. Under these conditions, concerns arise surrounding the negative impacts of elephant feeding behaviour and strategy, with calls from some quarters for intervention options to address this ‘elephant problem’. Research in a small (380 km 2 ) fenced reserve in the Limpopo province of South Africa was carried out in a variety of habitat and terrain types to assess type and level of elephant impact in relation to tree architecture. Impact was assessed to be predominantly non-lethal, with only a small proportion of surveyed trees assessed to have been lethally impacted upon. Tree species exhibiting certain architectural characteristics, notably trees with a wider canopy measurement than their overall height measurement, were particularly prone to negative feeding impacts, especially damage to primary and secondary branches. Trees with wide basal stem circumferences were most prone to being killed by elephants, with the overall likelihood of a tree experiencing the most severe impact types increasing with basal stem circumference. In contrast, an increase in height was shown to decrease the likelihood of a tree experiencing severe impact types, indicating that tall trees are less likely to suffer mortality due to elephant impact. We suggest that managers of fenced elephant reserves conduct baseline monitoring of elephant impacts. These studies indicate that width at widest point, and to a lesser extent height at widest point are likely to be useful additions to baseline vegetation monitoring. This will identify the level and type of elephant induced impacts, specifically revealing those locations within fenced reserves with a significant component of target tree species/morphologies. Ultimately this will both initiate and direct preventative management measures, we suggest primarily in the form of exclusion fencing in order to prevent access by elephants.