The use of controls is fundamental to experimental design. Individuals and populations will change during the course of an experiment even in the absence of treatments. Untreated controls allow this background change to be measured and separated from the effects of treatments.

The location of controls is important. Controls should encompass the same range of environmental variation as the treatment sites. Obviously, we should not select low weed density sites to receive the agents and high density sites to be left as controls - we should have some low and high density treatment plots, and some low and high density controls. Furthermore, control and treatment sites should be well interspersed, although care is needed to prevent drift (insecticides) or dispersal (biological control agents) of applied treatments into controls.

[ Back ] [ Next ]