Foliage-sapwood area relationships for balsam fir in north-central Minnesota

Leaf area is an important determinant of plant productivity. Because foliage is difficult to measure directly, allometric relationships often are used to predict branch-level and tree-level attributes such as total mass, needle mass, crown biomass, and projected leaf area. Two modeling approaches (i.e., the "crown weighing" and the "branch summation" approach) and two model forms (i.e., a nonlinear model form and a linear logarithmic model form) have generally been used to develop allometric relationships. It is unclear, however, whether these approaches result in similar predictions, which of these model forms provides us with the better predictions, and whether allometric equations developed for the same tree species in one region can be safely applied in another region. To investigate these questions, allometric equations to predict branch-level attributes (mass and leaf area) were constructed for balsam fir (Abies balsamea (L.) Mill.) in Minnesota. The "branch summation" approach was used to construct equations to predict crown biomass and projected leaf area from sapwood area, dbh, and crown length. These predictions were compared to those from the crown weighing approach. No statistically significant differences were detected. Results also indicated that the logarithmic form of the allometric model was the model of choice over the nonlinear form. Furthermore, branch-level and tree-level equations developed for balsam fir from north-central Minnesota differed from those of other regions within the range of data for the trees sampled. This reinforces the need for developing localized allometric equations.