Selecting models for capturing tree-size effects on growth-resource relationships

• Published in: Canadian journal of forest research Vol. 36; no. 7; pp. 1695 - 1704
• Main Authors: MacFarlane, D.W, Kobe, R.K
• Format: Journal Article
• Language: English
• Published: Ottawa, Canada NRC Research Press 01.07.2006; National Research Council of Canada; Canadian Science Publishing NRC Research Press
• Subjects: Acer rubrum; Acer saccharum; Biological and medical sciences; Comparative studies; equations; Fagus grandifolia; forest trees; forest yields; Forestry; Fundamental and applied biological sciences. Psychology; Growth; Growth (Plants); Growth factors; Growth models; Growth rate; growth-resource relationship; light; mathematical models; Measurement; Methods; nitrogen; Plant growth; Plant species; Quercus rubra; Resource availability; resource availablity; soil water; statistical analysis; Studies; tree and stand measurements; tree growth; tree size; Trees; Canada; Forestry
• ISSN: 0045-5067; 1208-6037
• DOI: 10.1139/x06-054

Subject trees included in growth analyses often vary in their initial size, possibly obscuring the effects of growth factors. We compare methods for incorporating size effects into growth models. For four different tree species, red maple (Acer rubrum L.), sugar maple (Acer saccharum Marsh.), American beech (Fagus grandifolia Ehrh.), and red oak (Quercus rubra L.), we compared models of radial growth rate of saplings as a function of light, water, and nitrogen availability that (i) ignored size effects on absolute growth-resource relationships, (ii) related absolute growth rate (AGR) to size and resource availability, (iii) related relative growth rate (RGR) to resource availability, and (iv) related RGR to tree size and resource availability. Size effects explained 13%-14% of variation in growth rates, and failure to account for these effects resulted in a substantial size bias in growth prediction. Overall, AGR-based models that included size as a predictor variable provided the best predictions and clearest interpretation of growth-resource relationships across all growth model types and species examined. Modeling RGR without including size effects resulted in residual size bias. Including size as a predictor of RGR yielded nearly equivalent results to using size to predict AGR. We suggest that investigators evaluate both AGR- and RGR-based approaches and determine which is most appropriate for their study.