Horticulturists are often interested in evaluating the effect of several treatment factors on plant growth in order to determine optimal growing conditions. Factors could include three or more nutrient elements, or types and rates of irrigation, pesticides or growth regulators, possibly in combination with one another. Two problems with such experiments are how to characterize plant response to treatment combinations and how to design such experiments so that they are manageable. The standard statistical approach is to use linear and quadratic (a.k.a. response surface) regression to characterize treatment effects and to use response surface designs, e.g., central-composite designs. However, these often do a poor job characterizing plant response to treatments. Hence the need for more generally applicable methods. While our goal is to be able to analyze three and higher factor experiments, we started by tweaking two-factor nutrient analysis data. The result was a hybrid model which allows for a given factor to respond linearly or non-linearly. We will show how this was done and our current “in progress” model and analysis for analyzing three quantitative factors.
Ellen T. Paparozzi, Walter W. Stroup, M. Elizabeth Conley and Reid D. Landes
Ellen T. Paparozzi, Melinda McVey McCluskey, M. Elizabeth Conley and Walter W. Stroup
Previous hydroponic studies have shown that nitrogen rates applied to roses can be cut in half as long as known quantities of sulfur are added. A two-year study began in February, 1991, to determine if roses potted in a 2:1:1 mix (soil:peat:perlite) would respond similarly. Six cultivars and three treatments (300 ppm N 20-30-10, N:S at 2:1 and N:S at 4:1 with N being approx. 155 ppm) were replicated three times in a split-plot design. Data included number of flowers and length of stems cut daily. Plants were allowed to grow for 4 months, were cut back, then allowed to grow for 7 months and cut back again. After the second pruning, shoots were harvested for N and S analysis. Soil samples were also taken. Initial data, analyzed through September, indicates that across cultivars the total number of flowers produced was not influenced by the N:S treatments. Certain cultivars, however, were more productive than others. Champagne and Bridal White consistently produced more flowers than Samantha and Amorous, regardless of fertilizer treatment. Certain treatment cultivar combinations were also significant indicating that cultivar response may limit N:S recommendations.
Ellen T. Paparozzi, Jazbaat K. Chahal, Petre Dobrev, Elizabeth A. Claassen, Walter W. Stroup and Radomira Vankova
Nitrogen response was compared in two Plectranthus species, Plectranthus parviflorus and Plectranthus ambiguus, which differ substantially in their phenotypic reaction to nitrogen limitation. The leaves of the former species gradually yellow during the nitrogen stress but are retained on the plant. This species copes with nitrogen deficiency also by gradual hydrolysis of starch grains. The latter species, P. ambiguus, responds by abscission of lower leaves. As plant hormones cytokinins are involved in nitrogen response as well as in regulation of the chlorophyll content, their dynamics were followed to explore the fast responses as well as the impact of nitrogen treatment. Both plant species responded to nitrogen deficiency/supplementation in similar ways, by downregulation and upregulation, respectively, of active cytokinins. However, the different phenotypic reactions imposed by nitrogen limitation were associated with specific regulation of cytokinin pool accompanying the stress response by the two species. The increase of the active cytokinins on nitrogen resupply was faster and stronger in the regreening species, P. parviflorus than in P. ambiguus. However, the P. ambiguus plants maintained higher basal levels of all cytokinin metabolites as well as chlorophyll content when compared with P. parviflorus. Nitrogen deficiency was associated in this species with accumulation of cis-zeatin-type cytokinins, which preceded abscission of lower and later on of middle leaves. The achieved results indicate that phenotypic variations in the response to nitrogen deficiency/supplementation are associated with significant quantitative and qualitative differences in the cytokinin pool.