EDGE-EFFECT BIAS IN THE SAMPLING OF SUB-CORTICAL INSECTS

Two general models are presented to describe the relations between the average number of insects bisected by sampling unit boundaries, the per cent edge-effect bias of mean-brood-density estimates, the shape and size of the average individual, and the shape and size of the sampling unit. The two general models, when expanded specifically for sampling late-stage mountain pine beetle broods, gave excellent fit to experimental data. The expanded equations are approximations since individual insects were considered as being rectangular in shape and the angles of the long axes of their orientation relative to the sampling unit boundary were considered to have a uniform frequency distribution. Edge-effect bias was a function of the size and shape of the organism and those of the sampling unit. Edge-effect bias resulting from faulty sampling-unit-area delineation is also considered, and suggestions are made for its reduction in sample surveys of sub-cortical insects.