Objective 1
Several multifactor ANOVAs were conducted to look at the interaction effects and main effects of water table depth, region (AB or NWT), plot type (Hummock or Hollow), and fire year on depth of burn values. All interactions between these variables were insignificant or had a low variance component <15%. Plot type, which was a proxy for water table depth, did not have a significant effect on depth of burn values for both regions (Figure 12). Region had a significant main effect (P < 0.001) on depth of burn values and accounted for 10% of the variance observed in the depth of burn values. As seen in Figure 12, the Alberta plots had more variable and overall higher depth of burn values than NWT. This was expected since the sites from the two regions are subject to different local weather conditions and vegetation characteristics (Flannigan et al., 2009). There was a significant main effect of fire year (P < 0.001) on depth of burn, which accounted for 20% of variance in depth of burn values. Similar to region, the weather conditions present at the time of the different fire events likely had a large influence on fire behaviour (Flannigan et al., 2009). Regarding the main predictor variable of this study, water table depth had a significant main effect on depth of burn (P < 0.001), but only accounted for 7% of depth of burn variance. Through conducting pairwise comparisons (Table 3), water table depths ranging from 0 to -60 cm were determined to not be significantly different in terms of their impact on depth of burn. However, water table depths above the surface, greater than 0 cm, had significantly higher depth of burn values than the other water table classes (Table 3, Figure 12). This result was the opposite of our expectation and the results of previous studies (Rowe et al., 2017; Nelson et al., 2021), and may be due to sampling errors from measuring depth of burn on Carex aquatilis roots submerged in high water tables. Alternatively, more severe fires with deeper burns may have removed a significant layer of peat; thus, the water table would be relatively closer to the surface (Lukenbach et al., 2016).
Objective 2
To analyze bryophyte regeneration, multifactor ANOVAs were conducted to look at the interaction effects and main effects of depth of burn, water table depth, region, plot type, and fire year on bryophyte cover. All interactions between these variables were insignificant or had a low variance component <15%. Depth of burn, our main predictor variable, did not have any significant main effects on bryophyte regeneration for both regions (Figure 13). This result is contrary to our prediction that more severe fires, as measured by greater depth of burns, would hinder regeneration through significant removal of organic matter. Bryophyte regeneration may have been more closely related to water availability and chemistry after fire (Malson & Rydin, 2007). In addition, bryophyte species have different tolerances and response mechanisms to disturbances which may explain why our results differ from previous studies (Benscoter et al, 2006). Region had a significant main effect (P < 0.01) on bryophyte cover values, but only accounted for 1% of the variance observed in bryophyte values. Plot type had no significant main effects on bryophyte cover suggesting that microsite position had a minimal influence on recovery.
There was a significant main effect of fire year (P < 0.001) on bryophyte cover, which accounted for 50% of variance in bryophyte values. In Figure 14, bryophyte cover increases as years post fire event increases as expected, and pairwise comparisons showed significant differences in bryophyte cover for each fire year. In the first year post-fire, regeneration was likely limited to pioneer species, but around 5 years post-fire more late successional species would have been established (Magnan et al., 2012). Water table depth had a significant main effect on depth of burn (P < 0.05), but only accounted for 1% of bryophyte cover variance. Pairwise comparisons did not show significant differences between the water table classes.
Conclusions
For objective one, our prediction that deeper water table levels would be positively related to deeper burn depths, our proxy measure of fire severity, was not met. Although our ANOVA showed that water table depth did have a minimal effect on depth of burn variances, water depth classes ranging from 0 to -60 cm did not have significantly different effects on burn severity. One possible explanation is that these layers of peat maintain high moisture content from other water sources independent of the water table, which acts as a buffer to fires. This can be inferred due to the high moisture retention abilities of peat and bryophytes (Huang et al., 2015) and the ability of peat fibers to draw from deep water sources through capillary action (Argo & Biernbaum, 1994). Regarding the management applications of these results, the extraction of groundwater by industries near these peatlands is not likely to significantly increase local fire risks. For objective two, our measure of fire severity, depth of burn, in addition to water table depth did not have any consistent significant effect on bryophyte cover. This lack of relationships suggests that, although higher fire severities can increase a loss of organic matter, they may not have lasting effects on regeneration. Although water availability in peatlands can varying burn severities (Lukenbach et al. 2016), our study shows that water table levels were not a critical component of bryophyte recovery. As this study implies, the choice of fire managers in these regions to let severe peatland fires burn would likely not have long-term impacts on bryophyte recovery (Benscoter et al, 2006). However, our sites are biased towards small areas in the NWT and northern Alberta and sample sizes for the different fire years are small, so these results are not representative of all northern fens. Before fire managers dismiss the potential impacts to peatlands, more exploration should be done to confirm the presence or absence of relationships between water table levels and fire severity as well as fire severity and bryophyte regeneration.