Summary :
Project by Cassidy van Rensen for RenR 480
Alberta's boreal forest is a grid of fragmented forest as the oil and gas industry is increasingly clearing linear corridors for seismic exploration. This has important implications for boreal flora and fauna. Threatened woodland caribou are experiencing habitat loss and increased predation from wolves due to forest fragmentation from industrial development. Edge effects and stand reductions can also decrease timber values and impact Alberta's forest industry. While some seismic lines are showing signs of recovery, the majority are not recovering. After disturbance, the forest can remain in a stable state of arrested succession (no recovery). If managers are to restore linear disturbances it is necessary to understand the key factors responsible for arrested succession and how these factors interact. Restoration is costly and time consuming, and priority should be at sites that will not regenerate naturally or fail to regenerate over a long time.
I hypothesized that soil moisture and nutrients, ecosite, and frequency of disturbance will be the greatest determinants of arrested succession and ultimately, successful recovery of seismic lines. To achieve my objectives I plotted seismic lines on a digital elevation model covering 33 townships south of Fort McMurray, Alberta. I randomly sampled points on lines across the landscape and used site and stand attributes of these sample points. I used canopy height online as my continuous responding variable and site and stand attributes as my dependent variables. I analyzed the data using a Regression Tree analysis in R using the rpart package.
Contrary to the hypothesis, time since disturbance was the most important predictor of canopy height online. It accounted for 23.6% of the variation observed in canopy height online. After removing time, the adjacent stand composition and adjacent leading conifer and dominant species, were the most influential predictors of recovery on the seismic lines. These secondary factors did not account for more 5% of the variation seen in canopy height online. In conclusion, it will take upwards of 20 years for most seismic lines to recover. If restoration can reduce time for regeneration, that restoration should be focused on areas that will not regenerate naturally (mixedwood or deciduous forests without ATV or truck disturbance) nor areas that are unlikely to regenerate faster even with costly silvicultural treatments (nutrient poor, wet coniferous forests with continued disturbance from ATVs and trucks).
Alberta's boreal forest is a grid of fragmented forest as the oil and gas industry is increasingly clearing linear corridors for seismic exploration. This has important implications for boreal flora and fauna. Threatened woodland caribou are experiencing habitat loss and increased predation from wolves due to forest fragmentation from industrial development. Edge effects and stand reductions can also decrease timber values and impact Alberta's forest industry. While some seismic lines are showing signs of recovery, the majority are not recovering. After disturbance, the forest can remain in a stable state of arrested succession (no recovery). If managers are to restore linear disturbances it is necessary to understand the key factors responsible for arrested succession and how these factors interact. Restoration is costly and time consuming, and priority should be at sites that will not regenerate naturally or fail to regenerate over a long time.
I hypothesized that soil moisture and nutrients, ecosite, and frequency of disturbance will be the greatest determinants of arrested succession and ultimately, successful recovery of seismic lines. To achieve my objectives I plotted seismic lines on a digital elevation model covering 33 townships south of Fort McMurray, Alberta. I randomly sampled points on lines across the landscape and used site and stand attributes of these sample points. I used canopy height online as my continuous responding variable and site and stand attributes as my dependent variables. I analyzed the data using a Regression Tree analysis in R using the rpart package.
Contrary to the hypothesis, time since disturbance was the most important predictor of canopy height online. It accounted for 23.6% of the variation observed in canopy height online. After removing time, the adjacent stand composition and adjacent leading conifer and dominant species, were the most influential predictors of recovery on the seismic lines. These secondary factors did not account for more 5% of the variation seen in canopy height online. In conclusion, it will take upwards of 20 years for most seismic lines to recover. If restoration can reduce time for regeneration, that restoration should be focused on areas that will not regenerate naturally (mixedwood or deciduous forests without ATV or truck disturbance) nor areas that are unlikely to regenerate faster even with costly silvicultural treatments (nutrient poor, wet coniferous forests with continued disturbance from ATVs and trucks).