Search Results
You are looking at 1 - 4 of 4 items for
- Author or Editor: Michele Choma Marini x
Data generated from horticulture experiments often are expressed as percentages. Since these data may not approximate normality, we often automatically transform the data before analysis. Is this automatic transformation always correct? Non-transformed data may already approximate a normal distribution. Should these data be transformed? What does transforming a normal data set do to it? To address these questions, a real data set was used to check the assumption of normality and examine regression models before and after transformation of normal and non-normal percentage data. Transformations include both standard arcsin of the square root of the proportion and a log (log) transformation. We will show that transforming normal data can make it non-normal and alter conclusions. Testing assumptions before and after transformation will be emphasized. The SAS procedures GLM and UNIVARIATE will be used to generate output and demonstrate results of normality tests and different conclusions resulting from unnecessary transformations.
Abstract
Diffuse photon flux density of photosynthetically active radiation (PAR) was measured throughout the season in peach trees [Prunus persica (L.) Batsch. cv. Harken] trained to an open center. Penetration of PAR generally decreased as shoot extension continued through the season. PAR was greater at the tree periphery and center than midway between the 2 points. PAR also decreased from the upper levels to the lower levels of the tree canopy. Specific leaf weight (SLW) at all positions increased from early June until August. SLW generally followed the same pattern as PAR, with lowest SLW developing midway between the tree periphery and center. Net photosynthesis (Pn) was greatest for peripheral leaves, lowest for leaves midway into the tree, and intermediate for leaves from the tree center. Pn and SLW were linearly correlated. Leaves developing at different distances from the tree center had similar stomatal resistances. There was no consistent influence of canopy position on chlorophyll content, but interior leaves tended to have greater chlorophyll content than peripheral leaves.
`Sweet Sue' peach (Prunus persica L. Batsch) trees were subjected to a factorial arrangement of treatments. At planting, trees were headed at 10 or 70 cm above the bud union and trees were trained to an open-vase or central-leader form. For the first 4 years, high-headed trees were larger than low-headed trees. After 7 years, open-vase trees had larger trunk cross-sectional area, tree spread, and canopy volume than central-leader trees. Open-vase trees had higher yield and crop value per tree, but lower yield and crop value per unit of land area or unit of canopy volume than central-leader trees. Crop density and yield efficiency were similar for all treatments.
Girdled or nongirdled `Biscoe' peach (Prunus persica [L.] Batsch) secondary scaffold branches were covered with shade fabric to provide a range of photosynthetic photon flux densities (PPFD) from 44 to 20 days before harvest (DBH), from 20 to 0 DBH or 44 to 0 DBH. Fruit quality was affected differently by the various periods of shade during the final swell of fruit development. Shading 40 to 20 DBH did not affect fruit weight or quality, whereas shading 44 to 0 DBH had the greatest effect on fruit weight and quality. Fruit quality was generally similar on branches exposed to 100% and 45% incident PPFD (IPPFD). Fruit on” girdled branches generally responded to shade more than fruit on nongirdled branches. Fruit weight was positively related to percent IPPFD for girdfed but not nongirdled branches shaded 20 to 0 DBH and 44 to DBH. On nongirdled branches, fruit exposed to 45% IPPFD for 44 to 0 DBH had 14% less red color and 21% lower soluble solids content (SSC) than nonshaded fruit. Harvest was delayed >10 days and preharvest fruit drop was increased by shading to <23% IPPFD. Shading branches for 20 to 0 or 44 to 0 DBH altered the relationship between flesh firmness and ground color: Firmness declined as ground color changed from green to yellow for fruit shaded 44 to 20 DBH, but firmness declined with little change in ground color for fruit shaded 20 to 0 or 44 to 0 DBH. Girdling results indicated that fruit weight and SSC partially depended on photosynthate from nonshaded portions of the canopy, whereas fruit redness, days from bloom to harvest, and ground color depended on PPFD in the vicinity of the fruit.