Fire islands have higher mortality than harvest islands
With respect to mortality of all species of trees in all types of plots, fire island edge plots have the highest mortality means and spreads.
Islands created by fires had different amounts of mortality than harvest created islands based on an ANOVA of a fixed effects linear model. Interestingly, in both reference plot types the harvest disturbance plots had slightly higher mortality, although not significantly so. |
This barplot shows the visual evidence for higher mortality in fire island plots across the study area. The interaction between disturbance (fire or harvest) and location (reference or island) was significant in a fixed effects linear model.
There is a slightly higher mortality seen in harvest reference plots, but the differences between these means are not statistically significant. |
No observed edge effects
In both edge and interior plots, fire disturbance shows higher mortalities but this difference is not large enough to be statistically significant based on an ANOVA of a linear mixed effects model.
No patterns in species mortality by treatment
This PCA is looking at mortality by species. This ordination shows close to 45% of the variance between observations but there are some interesting relationship to explore. The aspen (Populus tremuloides) and poplar (Populus balsamifera) mortality rates are driving the ordination because our site selection was based on deciduous dominated forest stands, which were either aspen or poplar dominated. We also see that birch (Betula spp.) and fir (Abies balsamea) mortality are correlated--often these two species occurred in the same plots, so this makes sense that we would see their mortalities correlate as well. Both fire/harvest and island/reference don't seem to drive any of the mortality that we see in this ordination, other than possibly the birch and fir (which were present in a handful of plots in the Flattop fire area). Other than that, no one treatment type resulted in higher a mortality of any one species.
Island size doesn't correspond to mortality
We see here that island area doesn't strongly correspond with mortality rates. There is a very weak negative relationship between larger island size and smaller mortality amounts in fire Interior plots, and harvest edge and interior plots. Fire edge plots show a very weak positive relationship between larger island size and higher mortality amounts. An ANOVA of a linear fixed effects model shows that fire island mortalities are still significantly different than harvest islands, this fits with the results from an earlier model. However, the model did not show a significant effect of island size on observed mortality amounts.