The Growth, Ecology and Population Structure of Giant Tortoises on Aldabra

P. Grubb

Abstract

The giant tortoise of Aldabra, Geochelone gigantea, shows quite marked changes in proportions with age, although during growth the relations between the length of the carapace and various measurements of the plastron and scutes involve not only strong but also weak allometry. Certain scutes show a predisposition to split during growth. Accidental damage to the carapace is frequent. Males reach over 100 cm in carapace length but females are smaller, up to 80 cm. There is no segregation between the sexes in any single measurement investigated, except among the very largest animals. A general appraisal of carapace and tail shape is sufficient to sex only animals above 60 cm in carapace length. The number of annuli on each scute corresponds to the number of years of age at least up to the formation of the tenth to fifteenth annulus. A general assessment of the pattern of growth is made by plotting body measurements against number of annuli. Growth curves of individual tortoises are reconstructed by relating measurements of successively formed annuli to age. Growth rate is recorded by plotting the difference between successive pairs of annulus measurements against age. The growth rate of ageable tortoises varies between local populations on South Island and between populations of South and Middle Island. Growth rate declines with age, reaching asymptotes at mean values of between 20 and 30 years. Some individuals exhibit sudden increases in growth rate after several years of very slow growth. There is a well-marked daily cycle of activity, feeding being limited to the early morning and late evening. Agonistic behaviour is virtually absent. Breeding is seasonal and the males select partners from within a limited size range of tortoises. Most mating attempts are unsuccessful. On Aldabra, tortoises occur in a wide variety of habitats, in each of which they depend on a different plant species or vegetational association for food. On coastal plains the chief source of food is Sporobolus virginicus. A variety of small herbs is consumed on the barren stretches of coastal champignon. Distribution in these areas is profoundly affected by the availability of shade. Further inland, the tortoises browse heavily on Guettarda speciosa in woods dominated by this tree. They take advantage of seasonal successions in the vegetation associated with freshwater pools, feeding on each community as it develops. Most of the woody plants near the pools are ignored. On the platin, browsing is selective and the regeneration of some trees is held in check. A very important food source here is the 'tortoise turf' (a sward in which Panicum sp. is often dominant) developed under conditions of heavy grazing and susceptible to erosion by wind and the tortoises themselves. On Middle Island, where the population is small, the tortoises exert very little effect on the vegetation. Associations with other animals are mostly casual, but along the south coast dunes Coenobita rugosus is dependent on tortoise faeces for food. Fossilized tortoise bones have been discovered at many points on Aldabra, deposited in brown limestone. They probably date from before the interstadial of about 30 000 years ago. Some adult tortoises range over 7 km or more, across a variety of habitats, but many individuals appear to be sedentary. The population of South Island is enormous-of the order of 100 000 animals-with a density of about 30 hm<latex>$^{-2}$</latex> on the platin. Higher densities are reached in Guettarda woodland. Local variation in numbers, size range and age structure depend on habitat preferences, differential movement of age classes and regional differences in growth rate. Attempts at assessing age class distribution are affected particularly by undersampling of the younger age classes, and the difficulty of counting the worn growth rings in animals with more than about 14. In the census sample, which may itself be an imperfect sample of the whole South Island population, at least 35% of the animals are below 20 years of age and only about 20% can have reached sexual maturity. More than 50 age classes may be present, but this and similar deductions are still speculative.