Foliage-sapwood area relationships for Abies balsamea in central Maine, U.S.A

Studies of forest productivity commonly invoke the pipe-model theory, which implies that leaf area or leaf mass in the tree crown is proportional to cross-sectional xylem (sapwood) area below the crown, to estimate leaf area or leaf mass from cross-sectional sapwood area. Prior ecophysiological studies have suggested that models to predict projected leaf area (PLA) from sapwood area for Abies balsamea (L.) Mill. are valid across a broad geographical region in northeastern North America. However, no single study has explicitly tested the applicability of different model forms to predict PLA from sapwood area. The objectives of this study were to (1) test the consistency of leaf area/sapwood area ratios at the base of the live crown among four canopy positions, (2) select the best sapwood-based model out of several published model forms to predict projected leaf area, (3) explore the ability of nonsapwood-based models to predict projected leaf area, and (4) test the validity of a common model to predict projected leaf area for A. balsamea in central Maine, U.S.A. We detected no strong statistical differences in the leaf area/sapwood area ratio at the base of the live crown among the open grown, codominant, intermediate, and suppressed canopy positions. Using a modified likelihood criterion to compare the ability of various model forms to predict PLA, we found that a log-linear model incorporating sapwood area at breast height (BH) and crown length (CL) performed the best, but a log-linear model with CL as the sole predictor variable also performed well. We concluded that a single logarithmically transformed model form using sapwood area at BH and CL to predict PLA is valid among canopy positions, but a single model to predict LA from sapwood area is not valid across a broad geographical region in northeastern North America.