Print PageIn an attempt to identify the characteristics that make forests more resilient to disturbance, it has been suggested that there may be a positive correlation between resilience and structural complexity. However, to test this hypothesis on a large scale, we must be able to measure structural complexity in a relatively efficient way. Mean Information Gain (MIG) may be an efficient measure of forest structural complexity. Inspired by Kolmogorov’s measures of information complexity, MIG quantifies the diversity of associations between pixels in photographs taken within forest communities. In this study, we compared MIG to traditional measures of forest structure (canopy closure, tree density, and structural diversity). We then examined the extent to which MIG was reflective of forest structure, as quantified by these traditional measures. As part of a large scale experiment in sustainable forest management in Quebec, Canada, we collected mapping data and took photographs in 15 forest sites with different disturbance histories (from natural to partially harvested forests). Specifically, at each site, we mapped the living trees and dead wood pieces in a 1600m² plot and we photographed the vertical structure at 40 different locations.
Results/Conclusions
MIG effectively captured variations in vertical structure (canopy closure, tree density, and structural diversity). Like these traditional measures of structure, MIG allowed us to differentiate between different intensities of partial cutting, indicating a sensitivity to the intensity of disturbances. Considering the information provided by MIG and the ease of the sampling method, MIG may be a useful index for measuring the complexity of forest structure under operational conditions. Future research should examine its sensitivity to different types of disturbances.
