Application of the Conservation Planning Tool Zonation to Inform Retention Planning in the Boreal Forest of Western Canada

Emulation of natural disturbance (END) has become a dominant paradigm for sustainable forest management in Canada. A popular approach to END is retention harvesting, in which live trees and other components of forest structure are retained within harvested areas. A primary objective of retention forestry is to maintain biodiversity and to hasten post-harvest recovery of forest structure and function. In forests with a natural disturbance regime dominated by wildfire, approaches to retention harvesting are based on the premise that wildfires create a mosaic of varying fire severity, including unburned patches. Locating where retention should be placed to best achieve management objectives is a challenging problem, and evidence-based approaches to operational applications are rare. We suggest here that harvest planners could benefit from the use of systematic conservation planning principles and methods to inform retention design. Specifically, we used a conservation planning—or prioritization—tool, Zonation, to create spatially-explicit scenarios of retention harvesting in a boreal mixedwood forest in north-western Alberta, Canada. Scenarios were informed by several environmental variables related to site productivity; in particular, we used a metric of topographic wetness that is based on airborne lidar data and was previously shown to correlate with patterns in post-harvest forest regeneration and biodiversity. The nine retention scenarios examined here related to the placement of retention focused to drier, mesic, or wetter sites in combination with other prioritization constraints. Results were compared with an existing harvest plan to assess differences in the spatial pattern of retention (e.g., percent overlapping area, number of patches, size of the patches). We also tested for the homogeneity of forest attributes (e.g., tree species, deciduous density) between scenarios and the existing harvest plan using multivariate dispersion analysis. Our results showed limited overlap among scenarios compared to the existing harvest plan; they were also characterized as having more and smaller patches with the use of a timber-cost constraint further affecting retention patterns. While modelling results significantly differed from current retention practices, the approach presented here offers more flexibility in testing different scenarios and assessing trade-offs between timber production and conservation goals using a standardized conservation planning toolkit.