Disturbance And Recovery Trajectory

High density of seismic lines in northern Alberta is contributing to a shift in predator-prey interactions between wolves and caribou and negatively impacting caribou populations. Regeneration of seismic lines may help counter this. To better understand relative importance of (1) site-specific disturbance history, (2) stand characteristics, (3) terrain features and (4) light and moisture on the vegetation recovery of seismic lines to an average height of 3 m, we used remotely-derived site history, line, stand, human activity and terrain variables to construct a series of competing models. Study area was 180,603 ha of Central Mixedwood and Boreal Highlands in Stoney Mountain 800 area, which includes over 12,000 km of linear disturbances and important caribou range. We used LIDAR to construct Canopy Height Model (2 m res), created 1369 2 x 50 m plots on 2D seismic lines using ArcGIS 10.1 and calculated average height of vegetation, then used generalized linear models (GLMs) with logit link to model recovery to 3 m with explanatory variables derived from Alberta Vegetation Inventory (AVI), Wet Areas Mapping (Environment and Sustainable Resource Development-ESRD), lineal inventory from Greenlink Forestry, and Natural Resources Canada Fire Database.

Results: The most supported model is the full global model. When we compared site-specific disturbance history, stand characteristics, and terrain features, we found moisture and light most important in explaining 2D seismic line recovery to 3 m height of vegetation. Within the global model, time since disturbance dependent on depth-to-water (DTW,) fire presence, DTW, dominant species in adjacent stand and distance to roads are the most influential variables. Implications: Regeneration of seismic lines is complex; multiple factors need to be considered. Important to assess factors affecting light and moisture at a site to determine likelihood of natural regeneration. Environmental conditions at a site and regeneration time likely to play greater role than the disturbance history. Very wet sites on wider seismic lines adjacent to black spruce dominated stands and in close proximity to roads may show slow to no recovery, even after 30 years. Previous occurrence of a fire in a stand adjacent to a seismic line may promote vegetation regeneration. Future Work: Use top GLM to predict recovery success over landscape and develop spatial optimization model that priorities conservation and restoration of linear disturbances that most benefits woodland caribou and minimizes cost, and create a decision-support GIS tool to assist restoration planning by land managers.