An improved method of DRIS for determining the nutrient status of pecan (Carya illinoensis (Wang) K. Koch) with foliar diagnosis /
| Main Author: | |
|---|---|
| Other Authors: | , , |
| Format: | Thesis Book |
| Language: | English |
| Published: |
1989.
|
| Subjects: | |
| Online Access: | Link to OAKTrust copy |
| Abstract: | A system for calculating DRIS (Diagnosis and Recommendation Integrated System) indices was developed for pecan (Carya illinoensis (Wang) K. Koch). The data consisted of 1805 foliar analyses and associated marketable weight of nuts per tree from Tifton, Georgia Experiment Station for 1961 to 1984. Records were organized into a format to be used by a computer program by individual sample number, sample type, elemental foliar analysis, and tree yield. Lab analyses were summed and the mean yield was calculated for the entire database. An F probability test was performed on the nutrient ratio variances of high and low yielding subpopulations sorted by the mean yield. When ratio variances were equal, ratio means were tested by Student's t. Significant mean and variance ratios at the 0.05 level were included in DRIS indices. A printout indicated a positive or negative total DRIS index. Comment statements explained the significance of the relative numbers. Pecan analyses were compared in four ways: by including the 0 yielding records in the mean and population ratios; by excluding the 0 yielding records from the mean and population ratios; by using the current year's analyses with next year's yield and excluding 0 yielding records; and by using log transformation of the means and SDs and excluding 0 yielding records. Log transformation generated insensitive indices. The first two methods incorrectly diagnosed the extent of sufficiencies or insufficiencies much of the time. Using current year's analyses with next year's yield apparently gave the best prediction of hidden deficiencies from test data. PARADES had several advantages over previous DRIS programs. PARADES used the population mean yield as the cutoff value which provides a uniform value for an equally distributed population. PARADES used both the F and t tests to select norms, unlike other methods. After selection of significant high mean ratios and CVs, all 156 ratios for 13 elements were used and not 78 with the reverse numeric sign being used for the reciprocal ratio. Erroneous results may have obscured previous uses of DRIS. |
|---|---|
| Item Description: | Typescript (photocopy). Vita. "Major subject: Horticulture." |
| Physical Description: | x, 156 leaves : illustrations ; 29 cm |
| Bibliography: | Includes bibliographical references. |