On the analysis of monitoring data: Testing for no trend in population size

The goal of a monitoring program is to find out whether the size of a population is stable, declining, or increasing over time. Whereas statistical tests for temporal trends are straightforward, there is no test for no change or stability. We propose that the combination of traditional trend analyses such as linear regression and tests for density dependence may be used for population stability analyses. We illustrate the approach by analyzing a data set from an orchid monitoring program. Combining statistical tests for temporal trends with tests for statistically significant density dependence is a useful and simple tool in the analysis of monitoring data.