Application of a Nonlinear Mixed-effects Model to Assess the Effects of Nursery Nitrogen Fertilization on Seedling Height Growth of Taiwan Yellow False Cypress

Abstract

A mixed-effects three-parameter logistic model was used to reevaluate the results from an experiment originally carried out in 1997, which examined the effects of 5 nitrogen fertilization levels (0, 50, 100, 150, and 200 kg ha^(-1)) on nursery seedling height growth of Taiwan yellow false cypress (Chamaecyparis obtusa Sieb. & Zucc. var. formosana Rehder) over a 200-d period. The experiment design was a completely randomized design with 6 replications for each level. For each replication, 6 height measurements were taken at 40-d intervals over the period. The asymptote of the logistic model for each replication was regarded as a random effect parameter that varied according to the level of fertilization, whereas the inflection point and the scale of the model were treated as fixed-effect parameters. A first-order autoregressive process was used to model the correlations between the repeated measurements. Results suggested that the seedling height asymptotically increased as the amount of fertilization increased linearly, with an estimated average rate of increase of about 1.4 cm per 50 kg ha^(-1) of additional nitrogen. The inflection point occurred at about 37 d after the first measurement was taken, and the average seedling height reached about 3/4 of each experimental unit's asymptote about 53 d latter Although the fertilization effects were statistically significant, the biological significance was limited. Both the inflection point and the scale parameter were unaffected by fertilization, implying that factors other than the nitrogen supply might be controlling the seedling height growth pattern. Thus, using the new model indeed provided important new information on the effects of nursery fertilization on the seedling height growth of Taiwan yellow false cypress.