Open Collections

Description

The data contained within describe and support the finding ancient forests project. This research proposed to use light detection and ranging data to quantify 47 forest stand metrics for differentiating old-growth from ancient forests. The study area included 120 plots of 400 square meters located in interior cedar-hemlock forests surrounding Kootenay lake in southeastern British Columbia. Stratification between stands in wet and mesic sites has been hypothesized to allow for more accurate delineation. For this reason, old and ancient forest areas were separated by areas of infrequent and rare stand initiating events. The analysis found that measurable relationships exist between ancient and old-growth forest categories when stratified by wet and mesic environments; however, the relationships vary by category and age. No overarching combination of metrics explained the variation between all categories. Additionally, variation within categories far exceeded that between categories, so a regression equation could not be established. A random forest classification of the data found that the distinction could only be accurately predicted between 25 and 60 percent of the time, and different iterations of the same model exhibited extreme variation. The validity of the results was limited by reliance on estimated attributes from the Vegetation Resources Inventory. If this methodology were to be repeated with field verified measurements as input data, it may be able to mitigate these issues and provide the basis for a reliable predictive classification. This classification could provide an accurate and objective standard that would aid in the identification and conservation of ancient forests.