Studying the population dynamics of weeds

Summary Studies of weed population dynamics may be conducted at various levels of complexity. The simplest level involves monitoring a single component of the population throughout several seasons, determining population trends and rates of change. This approach is highly descriptive and, therefore, has a low predictive value. However, it may be useful to evaluate the long‐term effects of various control techniques. In demographic studies, the life‐cycle of the species is divided into various functional stages and the number of individuals present in each stage is monitored throughout the season. From these data, various demographic parameters can be calculated. Population changes through time can be predicted by simulation techniques. This approach offers a practical compromise between the simplicity of long‐term studies and the high analytical and predictive power of more complex analyses. In mechanistic analysis, the population is broken down into a large number of components connected by various physiological und ecological processes. Detailed studies are conducted on each of these individual processes, determining their relationships with environmental factors. The data obtained are integrated into computer models in order to predict the dynamics of the population under various situations. Often the inquirer is confronted by the problem of choice of level of complexity at which it is necessary to study the populations concerned. This choice may depend on the goal of the study, the resources and experience of the people etc. Making the right choice is not always easy, but is essential for the success of any research programme.