RESULTS & DISCUSSION
Results
When addressing the objective of boreal resistance, it was necessary to separate out native and non-native species. Having the comparison of each will give us an indicator of forest composition, highlighting how varying disturbance regimes might affect the presence of each plant category. In Figure 5, there is a nice clear distinction of how native species are favored across all treatment types. This can be seen as a strong testament to the strength of our boreal forests, and the strong promotion and staying power of these native species. Two separate factorial ANOVAs were used to separately address native and non-native species richness. While pooling all species together, the data satisfies the ANOVA assumptions of normality, and homogeneity, I would not consider native/non-native independent. There is a direct tradeoff between covers, where despite the potential for overlap, it is likely that a native species occupying part of a plot, is necessarily excluding that space from then being occupied by a non-native individual instead. The two separate ANOVAs both yielded a similar result, that only the distinction of in intact forest vs. linear disturbance significantly affected native (F (1,18) = 4.966, p = 0.0388) and non-native (F (1,18) = 7.157, p = 0.0154) species richness. The difference between burned and unburned sites, as well as the interaction between burns and intact forest has no significant affect for either species categorization. |
Figure 5. Average species richness of native and non-native species across the forest/linear + unburned/burned treatment types. All plots along the transect were pooled together for this measure. Average species richness when combining native and non-native counts becomes 29.0, 28.8, 25.4, and 26.7 for intact/unburned, infact/burned, linear/unburned, and linear/burned respectively.
|
The presence of non-native species regardless of disturbance levels indicates that the boreal may not be as resistance as I once thought. Perhaps the larger landscape effects of various disturbances have already begun to provide the opportunity those invasives needed. By seeing how little of an affect burns had on promoting differences in species composition, I could interpret that natural and anthropogenic disturbances enact entirely different pressures on forests. Both native and non-native species alike have evolved with the presence of fires, and have developed in a way that allows them to easily return after a fire. Linear disturbances are instead a very different form of disturbance, and may be changing the environment more dramatically.
With average species richness of all species remaining largely unchanged (Figure 5), it is possible that the boreal has a limited number of niches to be filled, with overall forest composition shifting to invasives rather than non-natives arriving in addition to existing species.
With average species richness of all species remaining largely unchanged (Figure 5), it is possible that the boreal has a limited number of niches to be filled, with overall forest composition shifting to invasives rather than non-natives arriving in addition to existing species.
Figure 6. Percent covers across the transects for each treatment combination. Covers are represented as a proportion of non-native to native species, with a value of 1 showing an equal dominance of species by percent cover. The closer to 0 a value gets, the more heavily it is dominated by native species; and the larger the value beyond 1, the more heavily it is dominated by non-natives.
|
While species richness gives us an impression of the species makeup of a forest, it does not inform the dominance of each species, or how that dominance may vary spatially. In Figure 6, I examine the average percent covers of both native and non-native species through a proportional value, as shown across the gradient of treatment transects. Creating a proportional value to represent percent covers accounts for situations where the total percent cover may have been less than, or greater than 100%. Without this ratio value, the few species rich and species poor sites would artificially skew results, by not accurately representing species proportions. Attempts were made to analyze specific species, emphasizing non-natives, but their distributions are too patchy and in overall low enough frequencies to not bare any substantive results. A standout feature of Figure 6 would be how close to 0 (meaning heavily native-dominated) the forested lines remain. I expected this result, as there is no penetration, or migration that needs to take place for native species, they are already full present and established. When analyzing the data, an ANOVA was once again used with all necessary assumptions being met. Similar to species richness measurements, the only significant affect treatments specifically had on percent cover compositions was the difference in forests vs. linear disturbances (F (1,124) = 4.802, p = .0303), with burns once again having no significant affect. As suggested by the literature, distances of measurement were grouped into either "Near" the
|
beginning of a line (1, 2.5, and 5m) or "Far" along a line (10, 20, and 40m). While individual distances had no significant affect, these categorizations between near and far yielded a significant result between the two (F (1,40) = 6.16, p = 0.0174). A linear disturbance provides opportunity for non-native species, but as it attempts to migrate along a line, it must compete with natives that are able to in-fill from the edges of the remaining, uninvaded forest. So while non-native species are arriving, and can be found further down the transects, the forest resilience is able to largely keep the species composition in favor of natives.
Based on these early findings, a greater emphasis should be placed on the management of our linear disturbances, regardless of the presence of burns. From what can be seen here, a natural disturbance does not affect the forest in the same way, and effects of the two disturbances are largely decoupled, rather than being cumulative. Future studies would benefit from a larger focus on sampling linear disturbances and forests, with more intensive sampling methods rather than splitting the time between burned and unburned sites.
Based on these early findings, a greater emphasis should be placed on the management of our linear disturbances, regardless of the presence of burns. From what can be seen here, a natural disturbance does not affect the forest in the same way, and effects of the two disturbances are largely decoupled, rather than being cumulative. Future studies would benefit from a larger focus on sampling linear disturbances and forests, with more intensive sampling methods rather than splitting the time between burned and unburned sites.
Conclusions
|
A partially burned and deforest area of an Albertan forest (Photo from a Cenovus restoration project)