Detecting response of Douglas plantations to urea fertilizer at three locations in the Oregon Coast Range

Fertilizer trials in coast Douglas-fir (Pseudotsuga menziesii var. menziesi i(Mirb.) France) in the Oregon Coast Range usually indicate small and stati stically nonsignificant response to nitrogen (N) fertilizers. Inherently we ak experimental designs of past trials could make them too insensitive to d etect growth differences that actually exist. Ability to detect real differ ences among treatments should be improved by having more than two replicati ons per treatment and by using covariance analysis to adjust observed treat ment means for unequal starting conditions among experimental treatments. T o demonstrate these assumptions, we used size at fertilization and a prefer tilization (calibration) period of growth as covariates when analyzing data from five coastal plantations. The trials had three to six replications pe r treatment and calibration periods of 6 or 7 years. Nitrogen fertilizer wa s assigned randomly to half the plots at each location when trees were 16 o r 17 years old from seed. Our objectives were to quantify 4- or7-year respo nse to N fertilizer and to demonstrate practical means for detecting respon se. Effects of fertilization on tree diameter and height, and on basal area and volume growth per acre were estimated. Among the five non-thinned plan tations, observed gross basal area growth was changed by -2 to 13 percent i n the 4 or 7 years after fertilization. Observed responses were increased s ubstantially by covariance analyses at some plantations but decreased at ot hers. Random assignment of three to six plots per treatment did not ensure balanced or comparable plots for fertilized and nonfertilized treatments.