A spatial analysis of fine-root biomass from stand data in the Pacific Northwest

High spatial variability of fine roots in natural forest stands makes accurate estimates of stand-level fine-root biomass difficult and expensive to obtain by standard coring methods. This study uses aboveground tree metrics and spatial relationships to improve core-based estimates of stand-level fine-root biomass. Using the multiple-tree Ribbens model for pure stands, the approach assumes that the total fine-root biomass at a given point is the additive contribution of the nearest dominant trees and that fine-root biomass for a single tree depends on the distance to the trunk and its size. A Monte Carlo random sampling technique, or sampling on a regular grid, is used to estimate the average fine-root biomass across the stand. We illustrate the applicability of this approach by using it on root-core data from a Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) stand and a western juniper (Juniperus occidentalis Hook.) stand in the Pacific Northwest. We conclude that stand-level fine-root biomass is adequately estimated using the Ribbens model. Unlike the model-based estimate for stand-level fine-root biomass, the accuracy and precision of the arithmetic mean of the coring samples depends on the spatial heterogeneity of root distributions and the representativeness of the root coring samples.