ABSTRACT
Transect-sampling, or more precisely line intersect-sampling, is a relatively simple and widely used technique in ecological work and forest inventories. To get its flavor let us consider two standard forestry examples. Suppose that one needs a quick estimate of the volume of windthrown wood or of logging residue on a clear-felled area. The idea is to walk along a straight line, the transect, and to determine the volume of all the trees or logs on the forest floor that intersect this straight line. The procedure certainly selects a sample of the trees on the ground. Depending on the choice of transect and sampling scheme, our problem is to identify the inclusion probabilities and then derive an estimate for the entire population of downed logs. In the second example let us assume that one wants to estimate the percentage of the forest area where regeneration is occurring. We are confident that, by traversing a straight line, we will monitor the length of the transect that intersects with the regeneration patches. Intuitively speaking, one may argue that the proportion of the transect length that crosses those patches will somehow reflect the proportion of the surface area covered with new vegetation.
