Structural equation modelling: an alternative for assessing causal relationships in threatened plant populations Original Research Article

Structural equation modelling (SEM) is a powerful tool to explore and contrast hypotheses on causal relationships among variables using observational data. It constitutes an alternative to experimental approaches that is especially useful in the conservation of small populations where the implementation of treatments may have a negative effect on population viability. We are presently applying SEM to study the factors that condition reproductive success, seed emergence and plantlet survival in several plant species. We are also using model comparisons through multi-sample analysis to assess the implications of different microhabitats on the viability of a population. The most outstanding advantages of this tool are the global perspective used in the study of complex problems, the ability to discern the essential from the accessory, and the possibility of evaluating oneʹs own hypotheses. The basic procedure, the limitations of this method and further applications in conservation and management are also discussed.