The BTM land use layer was a very large file, so it was difficult to manage. Many dissolves had to be executed in ArcMap before it could be imported into IDRISI. I tried to use FME, but it would not pick up the coordinate system, and it created several records for a file of very few records. The data conversion process therefore took several hours and days.

There were also problems with the BTM map sheets. The polygons of adjacent map sheets did not seem to line up, which is likely due to digitizing errors. In addition, although the map sheets were merged, ArcMap did not recognize them as one object. This created a problem when the biogeoclimatic map was clipped to the BTM map and converted into raster. A line of empty data was created across the middle of the image, which corresponded to the line separating the top map sheets from the bottom. As mentioned earlier, this problem was overcome by clipping the biogeoclimatic map to a digitized polygon instead of the BTM map. However, the problem of the unaligned polygons of the adjacent map sheets was accepted, which will have created more error in the model.

Another problem that I came across was that the Decision Wizard would not let me retrieve the AHP weights once the pairwise comparison file was created. An hour of frustration later, I tried the MCE outside of the wizard and it worked fine. Also, the FUZZY module would not work inside the macro modeler. This, however, was not a big problem since it worked fine in the Decision Wizard and on its own.

This analysis does not seem to represent reality very well, since many more Spotted Owls could be supported in the suitable habitat, especially with a threshold of 40%, than are supported in reality. This was unexpected because I tried to standardize the factors in a way that would limit the suitability a great deal, which I thought was appropriate. However, I did notice that there was much more old forest than I expected there to be. The problem may be the definition of old-forest. The old-growth forests required may be older than the 140 plus years definition used by the BTM. In addition, the study area for this project extended much further east and further north than the actual range, which would help explain the larger amount of suitable habitat. Old-growth forest and the suitable biogeoclimatic zones exist in these regions, but there may be other factors, such as temperature, which inhibit the survival of the spotted owl there. Therefore, the study area should have been further restricted. There are, of course, other factors related to the spotted owl population besides physical habitat, such as competition and the abundance of prey. I would not like to put much emphasis on this, however, because logging companies and the government may say that there is lots of good habitat for the owls, and so loss of habitat is not the problem, which would strengthen their argument for logging the old-growth forests.

Another limitation of this study is that it only analizes the spotted owl habitat in British Columbia. Since borders do not affect owls, it is important that management be done at an international level.