Forest ecotone survey by line intersect sampling

What is it about?

This note demonstrates a procedure based upon line intersect sampling on remotely sensed imagery. The estimate of ecotone length per unit area is obtained by visual interpretation of the changes from forest to other land use classes along each sampling line. The experimentation carried out in two test areas within forest landscapes of central Italy proves the operative soundness of the proposed procedure, which is more efficient than the classical approach by forest polygon delineation and perimeter mensuration in a GIS environment.

Featured Image

Why is it important?

The interest of environmenmtal planners and managers in detailed landscape metrics is rapidly increasing. Distinctively, boundaries in the landscape mosaic have direct relevant structural and functional values. The classical ecotone survey approach is by mensuration on maps provided by remote sensing and processed in a GIS. The major argument against sampling is that there is no point in sampling when there is a good map giving the complete picture of the region of interest. Therefore, sampling approaches have their place in situations where no suitable map is available depicting the forest patches, or when such a map cannot be produced readily. However, it must be taken into account that complete mapping (i.e., by polygon delineation) can include errors in image interpretation that tend to be systematic and generally without compensation between omission and commission errors. In contrast, in the procedure proposed here, the estimate of ecotone length is obtained by interpretation of only the changes from one land use class to another along each sampling line, thus drastically reducing the possibility of omission and commission errors. This design strategy may also lend itself nicely to assessing the impact of the errors in image geometry, projection, and interpretation on the estimated ecotone length by quantifying them at the intersection points, whereas to quantify the error of measured GIS boundary lengths is usually a much more complex task.

Perspectives