The elephant problem–an alternative hypothesis

Summary Published hypotheses to account for habitat changes wrought by elephants begin from the assumption that elephant‐forest systems possess a stable equilibrium point. The ‘elephant problem’ is conceived as a displacement of this equilibrium by man. Controversy centres around which human activities caused the dislocation of equilibrium and by which mechanisms these activities resulted in local high densities of the elephant Loxodonta africana. A study on elephant‐forest relationships in the Luangwa Valley of Zambia casts doubt upon the basic assumption of these hypotheses and an alternative hypothesis is therefore offered. It begins from the opposite assumption–that there is no attainable natural equilibrium between elephants and forests in eastern and southern Africa. The relationship is viewed instead as a stable limit cycle in which elephants increase while thinning the forest and decline until reaching a low density that allows resurgence of the forest. This in turn triggers an increase of elephants and the cycle repeats. The period of the cycle, if the hypothesis is correct, is in the order of 200 years in the Luangwa Valley. The activities of man can impose an artificial equilibrium on the system such that trees and elephants are trapped at the low density phase of the cycle. When interference is relaxed, as with the conversion of an area to a national park, the cyclic relationship reasserts. The parameters of a system possessing a stable limit cycle need not differ in kind or interrelationship from those of a system with a stable equilibrium. Whether one or other outcome manifests may depend only on the numerical values of the parameters. If the elephant‐forest system is characterized by a stable limit cycle the period and amplitude should change along a climatic gradient and may contract to a stable equilibrium in some climatic zones. A set of predictions is offered to facilitate rejection of the hypothesis.