Conclusions

Overall, the final habitat suitability model fulfilled expectations, based on the datasets utilized. We had hoped to include snow cover data in the formulation of the model, however, obtaining appropriate datasets was not possible for the timeframe in which we were working. In future, such a suitability model could be improved by the inclusion of monthly snow depths. Hogg, et al (2008), noted that spring snow cover was particularly important to protect trees’ roots from freeze/thaw cycles, as well as providing moisture during the early growing season. SNOTEL data, available throught the U.S. Department of Agriculture's Natural Resources Conservation Service (http://www.wcc.nrcs.usda.gov/snow/), provides site-specific snow depth information for many locations throughout Colorado. To make use of this resource in a future modelling project, point-specific monthly snow depth data could be extrapolated into a surface (i.e., raster) dataset.

Another consideration for future habitat suitability modelling is the relative importance assigned to each input dataset. For this model, we assumed that all parameters were equally important, and simply summed all layers. It's likely that among the datasets we utilized, some are more important to aspen health than others, and should be weighted appropriately. There is a paucity in the literature for specific controlled studies of soil and moisture conditions on aspen health, though Hogg, et al (2008) found that aspen growth in western Canada was most limited by soil moisture levels.