Abstract
Path analysis was used to study shoot elongation of Scots pine seedlings grown under six relative mineral nutrient addition rates (Ra). The shoot length is composed of two growth components: number of stem units (NSU) and mean stem unit length (MSUL). Under suboptimal conditions (Ra = 1% and 2%) most variation in shoot length (LEN) was due to variation in MSUL, and the correlation between the two components was negative. Under optimal conditions (Ra = 3% and 4%), NSU and MSUL contributed equally to the total length variation and the correlation between them was positive. Under supraoptimal conditions (Ra = 5% and 6%), the number of stem units had the greatest effect on shoot length variation and the correlation between growth components was negative. A simple allocation model was used to help interpret the changes in sign of correlation coefficient along the nutrient gradient. The model correctly predicted the sign of the correlation but not its magnitude.
Notes
Present address: University of Wisconsin‐Madison, Department of Forestry, 1630 Linden Drive Madison WI 53706, USA.