Published online by Cambridge University Press: 20 May 2010
Studies on population growth rates (i.e. how fast population size changes) go back several hundred years, at least to the 16th century when the potential for exponential growth of populations was realized (Caswell 2001). Since then, factors affecting variation in population growth rates have been the major focus for human demography and a major part of population ecology. Although ecologists have dealt with this problem for a long time, surprisingly few generalizations have appeared that allow us to predict variation in population growth rates within and among natural populations. A major reason for this may be difficulties in separating out the relative contribution of density-dependent and density-independent factors on the population growth rate (see reviews in Sinclair 1989; Caughley 1994; Turchin 1995).
The purpose of the present study is to summarize how stochastic effects affect the long-term growth rate of populations with no age structure. We will then extend some recent work (Sæther et al. 2002), using data on fluctuations of bird populations, where stochastic factors as well as parameters characterizing the expected dynamics are being separately estimated. We suggest many characteristics of avian population dynamics are closely associated to variation in the specific population growth rate because of the presence of covariation between differences in the expected dynamics and the stochastic component of the fluctuations in population size. Finally, we strongly emphasize that reliable population projections, for example for use in population viability analysis, will not only require estimates and modelling of the expected dynamics as well as the stochastic components, but also assessment of the effects on the predictions of uncertainties in parameter estimates.