Relationship between diet, body condition, behaviour, and faecal glucocorticoid concentrations in African elephants

Abstract

In contrast to captive elephants, the body condition of wild ranging elephants is highly dependent on the amount and quality of food available, which is in turn dependent on a variety of factors, such as seasonal changes in water availability, or annual fire occurrence. It can also be assumed that distinct changes in body condition might influence the behaviour of the elephants, possibly via physiological changes, like changes in the concentration of circulating glucocorticoids i.e. a decrease in bodyweight due to a low energy diet may represent a form of physiological stress to the animal, which should be reflected in elevated glucocorticoid levels. Although methods already exist to reliably assess adrenocortical activity in African elephants, more information is necessary to illuminate the possible relationship between food availability, circulating glucocorticoids, body condition and behaviour. Therefore, the aims of this study were i) to analyse the excretion pattern of glucocorticoids in the faeces of elephants fed a normal, decreased, or increased diet, ii) to validate different methods for estimating elephant body condition, and iii) to analyse behavioural changes in elephants related to a change in diet. Five male and two female elephants between the ages of 7 and 14 years were studied for a period of 60 days, during which their diet was changed three times. For the first six days they were fed a normal diet, they were then subjected to 20 days of decreased diet, followed by 20 days of increased feeding and, finally, 14 days back on the normal diet. The animals used were housed at an elephant training facility (Elephants For Africa Forever) in Mooketsi, Limpopo province, South Africa. During the period when a normal diet was fed, the elephants spent approximately 5 – 7 hours a day browsing in the bush and were trained for 2 to 3 hours per day. In addition, they were fed 5 -7.5 kg of cubes containing dried Bhana grass, hay, harmony chop (maize product), molasses and salt, and allowed ad libitum Bhana grass and branches at night. During the period of decreased dietary intake, the elephants were restricted by 20% in the time spent browsing in the bush while the time spent walking was increased by 25%. In addition, the amount of cubes fed during training were reduced by 25% and at night the amount of food available was restricted to 35 kg of branches and 15 kg of Bhana grass for the younger elephants, and 50 kg of branches and 25 kg of Bhana grass for the older elephants. During the period of increased diet, the elephants had unlimited access to food in the bush, the time spent walking was decreased by 35% and the cubes fed contained an extra 25% of crushed maize. At night the elephants were fed Bhana grass and branches ad libitum. Faecal samples were collected throughout the 60 days for hormone analyses; every day for the first six days of the experiment and every other day after that. The faecal samples were analysed for 3α,11oxo-cortisol metabolites (3α,11oxo-CM), which have previously been shown to provide a reliable indication of adrenal function in the African elephant. Urine samples were collected on the same day as faeces and analysed for the urea nitrogen:creatinine ratio, which was used as a physiological indicator of elephant body condition. Body condition was also estimated by a variety of other methods, namely; a visual and digital body condition score, girth measurements and weight. Finally, the influence of different diets fed on the behaviour of the elephants was determined, using a behaviour scoring system. A significant difference in 3α,11oxo-CM concentrations was found between the periods of normal diet and increased feeding (One Way Analysis of Variance, n=45, P<0.001, Tukey Test: normal vs. increased P=0.0194), although only a sub-set of faecal samples (3 samples per elephant per period) have so far been analysed. Disappointingly, none of the methods tested in this study for detecting changes in body condition were reliable; moreover, the partly inconsistent results between methods suggest a suboptimal study design regarding duration and/or the quality of changes in diet. Finally, no changes in elephant behaviour as a result of changes in feeding regime were detected in this study. Overall, further research is necessary to better examine the effects of different diets on 3α,11oxo-CM concentrations, body condition and behaviour. Further studies should probably employ more pronounced changes (quantitatively and qualitatively) in diet.