Evidence for transient dynamics in plant populations based on long‐term demographic data

Abstract

Summary Recent interest in the importance of short‐term, transient responses in population dynamics has prompted the development of many new methods for describing transient responses in population models. Many of these methods focus on the largest possible responses for a model, thus providing a description of the ‘transient potential’ of a model that does not depend on knowledge of initial conditions in the population. This study uses long‐term demographic data from five perennial plant species to evaluate how well measures of transient potential reflect actual transient responses in the populations. I used previously published matrix population models to create a set of initial conditions derived from empirically‐based simulations for each population. Maximum transient responses within this set of initial conditions were then compared with responses used in describing transient potential. Transient responses based on empirical modeling results were smaller than responses used for measures of transient potential. Realized responses were strongly correlated with the distance between the current demographic structure in the population and the asymptotically stable stage structure of the model (SSD). Theoretical transient potential was predictive of the amount of variation in distance from SSD observed in each population in the simulations. Synthesis. Methods based on largest possible responses tend to overemphasize the role of transient dynamics. These results suggest that traditional, asymptotic analyses may be appropriate in many cases. Measures of transient potential can be helpful for identifying species and situations that may be prone to larger transient responses, but do not necessarily indicate that transient dynamics are more important in those systems.