Differences in recent coarse woody debris between disturbance types were tested by means of single factor analysis of variance by means of mixed linear models. Models were created and compared using Akaike's information criteria to determine the model that most effectively used the predictors, response and random variables. The model with the lowest Akaike's information criteria number was plotted to assess that it met the assumptions for the models.
Recent Coarse Woody Debris Model Selection
Table 3. Created models for recent coarse woody debris (decay classes 1 & 2).
Plot 3. The comparison of recent coarse woody debris between all fire and harvest sites.
Plot 4. Post hoc test of recent coarse woody debris volumes between all fire and harvest sites.
Recent Island Coarse Woody Debris Model Selection
Table 5. Created models for recent coarse woody debris (decay classes 1 & 2) for the island plots only.
Plot 5. Comparison of recent coarse woody debris between fire and harvest islands.
Multivariate Analysis
Plot 12. Principal component analysis displaying the relationship of species to recorded variables.
This principle component analysis was used to see how the species composition is influenced by predictor variables. This plot does not display many distinct patterns, but Populus tremuloides species tend to follow higher counts, total volume and initial stand volume. Being one of the most prominent species this result is not surprising. Salix spp trend to display a negative correlation with count and a positive correlation with organic layer depth. The principle component 1 explains 37.28% of the variation between the data points and principle component 2 explains 28.6% of the variation.
Plot 13. Non-metric Multi-dimensional Scaling of species in relation to recorded variables.
Plot 13 shows similar information as plot 12. It displays the species' relationship to the recorded predictor variables. We can see that Populus tremuloides and Populus spp. follow higher counts, higher mean diameters and higher mean total volumes. Due to our study seeking stands with a majority of deciduous trees in the boreal mixedwood forests it is logical to assume the most common tree species are in the Populus genus. Species such as Betula papyrifera, Betula spp., Salix spp. and Alnus spp. follow lower counts, mean total volumes and mean diameters, but higher organic layer depth. Analyzing these species it is apparent that they are more moisture dependent, which correlates with higher organic layer depths.