Environmental stochasticity and African elephant population dynamics : investigating limitation through juvenile mortality.

Abstract

The successful conservation management of African elephants depends largely on understanding the fundamental processes driving the population regulation of this species. Southern Africa’s increasing populations have raised concern over the impact of high elephant densities on the system, in stark contrast against the elephant’s more precarious position in other parts of Africa. As we search for solutions from the processes of historical elephant regulation, we realise that there is a decided lack of empirical evidence to explicitly direct our efforts. In this PhD, I attempt to investigate the application of the classic pattern of large herbivore population limitation, which mainly involves high juvenile mortality in response to stochastic environmental events, to African elephant population dynamics. Firstly, I evaluated the magnitude and frequency of mortality events that would be required to prevent elephant population growth. The death of 85 % of infants and weaned calves would need to occur twice a generation, while a single severe mortality event (causing the death of all infants and weaned calves and 10 % of the rest of the population) once a generation would be sufficient. However, the severity of these events is not matched in natural occurrence in Africa today and only a single recorded event in Tsavo National Park, Kenya, in the 1970’s has come close when more than 7 000 died during a very severe drought. Secondly, I evaluated the potential role of fire as a stochastic, massmortality event limiting elephant populations. I found that fire functions in a similar manner to other environmental catastrophes and primarily causes high juvenile mortality. However, this catastrophic event also highlighted the extreme behavioural and physiological impacts experienced by the elephant population involved. The potential role of these types of events on long-term female fecundity needs further investigation. In isolation, this type of mortality event would need to occur with high frequency to prevent population growth. However, in combination with a decrease in female fecundity, these stochastic events may have a much greater impact on population demography than first thought. Thirdly, I investigated a potential mechanistic link between stochastic mortality events and juvenile susceptibility to resource limitation. Allometric relationships dictate that juveniles select a diet of higher quality than adult elephants. We found that this was achieved by weaned calf selection of higher quality plant parts, although use of plant types and plant species was similar to that of adult females, who they move across the landscape with. The strong sexual dimorphism exhibited by this species was reflected in adult male use of lower quality forage than adult females (or juveniles) in both dry and wet seasons. Diet quality scaled negatively with body size, but adult females consistently selected a higher quality diet than adult males, irrespective of body size. The nutritional and reproductive demands placed on an individual during different life-history stages therefore influence foraging strategies, together with nutrient requirements, e.g. phosphorus for pregnancy/lactation selected consistently by females when unrestricted in the wet season, protein for growth selected consistently by weaned calves. Competitive displacement of adult females to feed at higher levels in the canopy by calves also influenced feeding behaviour. Therefore intraspecific body size, nutritional requirements (in terms of nutrients and energy) and competition had a strong influence on foraging strategy employed by age-sex classes of elephants in response to seasonal environmental change. More selective juvenile foraging requirements means that juveniles are most susceptible to resource limitation, for example during stochastic environmental events such as droughts. In small, closed systems, juvenile mortality is likely to have a strong influence on elephant population regulation, with a slight, temporary decrease in female fecundity possibly acting in conjunction with juvenile mortality effects. Therefore, stochastic environmental events such as drought and fire may be the only natural incidence of population regulation to occur in these systems, where populations continue to grow exponentially and there is no evidence of density-dependence (as in the case of many small, fenced reserves in South Africa). In large, open, high-density systems in other parts of southern Africa, density dependence acts strongly on female fecundity and causes low levels of juvenile mortality in areas of local population aggregation. Therefore, in isolation, natural juvenile mortality is unlikely to regulate African elephant populations, but in conjunction with decreased female fecundity in response to density-dependent feedbacks and stochastic environmental events, population regulation may occur. The management of long-lived megaherbivore species with similar demographic drivers must include an appreciation of the complexity of population response to manipulation of mortality or fecundity effects. Small changes can potentially result in large shifts in population dynamics. Further insight into the mechanisms driving these processes will allow sound scientific support of megaherbivore management decisions to be made throughout Africa.