Introduction

Factorial experiments are investigations of more than one experimental treatment examined simultaneously. They are widespread in ecology, but probably not widespread enough. Examples are experiments on, say, competition done simultaneously in several places. Observations are available that densities of animals of one species (A) are smaller where another species (B) that feeds on the same seeds is numerous. The model proposed to explain this is that there is competition for food between the two species and, where B are numerous, they decrease the abundance of A. The hypothesis is proposed that removal of B from some areas will cause an increase in abundance of A. The null hypothesis is that there will be no difference in density of A, or A will decrease in abundance where B are removed, compared with plots where B are left intact. The experimental treatments are plots where B are removed by trapping and then continued trapping is done to keep removing any immigrants during the experiment. Procedural control plots are established in which traps are set, but then triggered so that no animals can be caught. This creates the same disturbance to the habitat (e.g. due to people walking around) as in the experimental plots. Finally, there are undisturbed control plots. At the end of the experiment, traps are used in all three sets of plots to estimate the densities of species A.