Document: TAR-3-65-13

A spatially realistic distributed model of fragmentation resulting from human disturbance in the mixedwood boreal forest of Saskatchewan.

TCHIR, T.L.* and E.A.JOHNSON

University of Calgary, Calgary, AB, T2N 1N4, CANADA. ¹

Abstract:
A decision-based model of the fragmentation process using empirical land-use maps from 1900-1963 was developed to determine which parts of the landscape were more likely to experience human disturbance and which were more likely to be left as forested remnants. Although socioeconomic and biophysical attributes (i.e. nearness to neighbours, transportation networks, moisture and nutrient gradients) strongly influence the quality of land, reliable measures of these attributes were not all available to settlers. Instead, settlers relied on surrogate measures of moisture and nutrient status, such as vegetation, soil texture and stoniness. Using these surrogate measures as decision rules, the model determined that soil texture, soil formation and the amount of stoniness govern the fragmentation process with 71%-94% accuracy. To predict the distribution of the measures that underlie the settlement process, a spatially realistic slope wetness model (TOPMODEL) was used. Incorporating easily measured physical variables, TOPMODEL captures the variability of soil properties by redistributing the soil water and nutrients according to hillslope position, thereby predicting variations in surface soil water. The distribution of the soil moisture and nutrients on the landscape matched the distribution of fragmentation on the empirical settlement maps. Therefore, the slope wetness model could be used as a spatially realistic general model of the decision-based fragmentation process.

Keywords:

This abstract is being presented at: 10:30 AM in session:
Poster Session #5: Landscape Ecology.