Performance of a canopy light interception model for conifer shoots, trees and stands.

A hierarchical approach was used to evaluate the behavior of a canopy light interception model using submodels for interception at levels of the shoot, crown and stand. Mean silhouette areas of individual shoots of lodgepole pine (Pinus contorta var. latifolia Engelm.) calculated assuming spherical shoot orientation agreed well with silhouette areas measured with a video projection system. A theoretical mean silhouette to total leaf area ratio (STAR) calculated using measured shoot characteristics and a model of shoot geometry agreed reasonably well with STAR values from measured shoot silhouette and leaf areas, although the range of calculated STAR values was slightly lower than that of the measured values. The shadow areas of tree crowns were underestimated for lodgepole pine; however, a model using regular shoot dispersion in the crown gave better agreement than a model based on random shoot dispersion. The underestimates may have been caused by underestimation either of leaf or branch area for the crowns or of the extinction coefficient. Two equations used to calculate leaf area from sapwood area measurements both underestimated leaf area. Accordingly, there are too many uncertainties to determine whether crown interception is best represented by a random or regular distribution of shoots. Mean canopy transmission in the Engelmann spruce (Picea engelmannii Parry ex Engelm.) stand was estimated very well by a random (Poisson) model of the spatial distribution of trees, whereas, in the lodgepole pine stand, a regular (hexagonal) model gave the best results. The Engelmann spruce stand had less regular tree spacing than the lodgepole pine stand. Taken as a whole, the weakest submodel for calculating canopy interception was the model for individual crown transmission, which underestimated crown shadow areas. However, the problem may be in leaf area estimates rather than in model performance.