The temporal variability of animal abundances: measures, methods and patterns

Abstract

From first principles, the temporal variability of a time series of abundances can be defined as the average deviation of values from a mean value on a proportional scale. In this paper we review: (i) the different kinds of temporal variability; (ii) the different ways in which it can be measured; (iii) the design of appropriate sampling schemes; (iv) methods of analysing variability; and (v) patterns in temporal variability. We emphasize that some commonly applied measures are not appropriate, that several do not measure the desired feature of time series, and the importance of considerations of trend and sampling error. A number of suggestions are made for the improvement of the basis for comparative analyses of levels of variability, and some of the potential pitfalls are identified. Given the serious faults in many previous analyses of ecological patterns in the temporal variability of animal abundances, emphasis is laid on the theoretical basis for different patterns, and hence a set of hypotheses for testing is generated.