Several experiments were conducted to determine the influence of modifying the grapevine microclimate on the incidence of fruit rot and selected rot metabolites of Virginia grown White Riesling (<i>Vitis vinifera</i> L.) grapes. Due to the relatively high incidence of rot observed in these studies, an evaluation of the effects of rot on the sensory evaluation of juice and on selected aroma components was also undertaken. The effect of shoot topping to 10 or 20 nodes or the application of ethephon on the incidence of grape rot was measured for three seasons. Canopy modification by both topping and ethephon treatment increased sunlight penetration into the fruiting zone. Fruit rot incidence and the concentrations of rot-associated metabolites were significantly and mutually correlated. Fruit rots and rot metabolites were greatest with control and ethephon-treated vines and were least with vines whose shoots were topped to 10 nodes. In a separate three-year study, two to four leaves were removed from the fruiting zone of grapes grown on two training systems. Selective leaf removal generally increased sunlight penetration into the grapevine canopy. The incidence of Botrytis and sour rot as well as the concentrations of rot organism metabolites in harvested fruit were reduced by leaf removal. The effect of grape sour rot on general fruit chemistry, free volatile terpenes, potentially free volatile terpenes, selected polyols and rot metabolites was undertaken. Grape sour rot increased °Brix, glucose to fructose ratio, titratable acidity, tartaric acid, glycerol and gluconic acid. Grape sour rot reduced the concentrations of geraniol, nerol and linalool, while increasing oxidation products, aromatic alcohols and long chained alcohols. Grape juice sensory analysis appeared to be influenced by the severe fruit rot which occurred in both the control and to a lesser extent in canopy manipulated vines in both studies. The importance of fruit rot was demonstrated by both the production of odors, metabolites, and the destruction of native grape aroma constituents. Reductions in the incidence of fruit rot by selective leaf removal and shoot topping were demonstrated and were probably due to several microclimate factors that collectively increased evaporative potential and spray penetration into the fruit zone. The results illustrate that remedial canopy management practices that improve fruit zone porosity can mitigate the severe disease pressure that exists when rot-prone cultivars are grown in humid grape growing regions. The inability to reduce fruit rots further than noted in these studies may be due to the complex nature of fruit rots observed here, compared to those (principally Botrytis rot) examined in the majority of other canopy management studies. / Master of Science

Complete vineyard floor cover cropping and root pruning (RP) were evaluated for their ability to regulate excessive vegetative growth and improve berry and wine composition of ‘Cabernet Sauvignon’ (Vitis vinifera L.). Treatments were: tall fescue (Festuca arundinacea Shreb.) ‘KY-31’ and ‘Elite II’, hard fescue (Festuca ovina L.) ‘Aurora Gold’, perennial ryegrass (Lolium perenne L.), orchardgrass (Dactylis glomerata L.), and an under-trellis herbicide strip combined with KY-31fescue interrows. Compared to herbicide strip/non-root pruned (NRP), Elite II fescue reduced vine pruning weight (kg/vine) 28%, individual cane weight (g) 20%, and canopy leaf layer number 25%. KY-31 fescue/RP lowered vine pruning weights 29% compared to an 8% reduction in pruning weights of vines grown in herbicide strip/NRP plots from 2005 to 2010. KY-31 fescue produced the greatest biomass and stand density. With the exception of a yield reduction in vines grown with KY-31 fescue in 2006, cover crops minimally decreased grape yield. Yearly climatic variation had a greater effect on berry weight and composition (pH, TSS, TA) than did treatments. Limited treatment differences detected in chemical compounds by gas chromatography–mass spectrometry (GC-MS) analysis in wines made from treatment vines in 2010 were correlated to descriptive sensory terms. Cover crop water use, as evapotranspiration, determined by mini-lysimeter (ML), ranged from 3.28 mm/d for KY-31 fescue to 1.52 mm/d for herbicide-treated plots. In 2008, root biomass of vines grown on KY-31 fescue/RP was increased at the 60 to 80 and 80 to 100 cm soil depths compared to root biomass of KY-31 fescue/NRP vines at those depths. Cover crops minimally impacted vine water potential (ΨPD, Ψmd, Ψstem) and grapevine nitrogen levels relative to the herbicide strip, indicating that the grasses were not overly competitive with grapevines. Root pruning and complete vineyard floor cover crops favorably reduced grapevine vegetative growth, although treatment effects diminished over time, possibly in response to redistribution of grapevines’ roots and climatic variation at the site. / Ph. D.

Loblolly pine is a commonly planted pine species in the Southern US which is intensively managed as well as a major contributor to the timber industry. Various silvicultural treatments are commonly applied to pine plantations including thinning and sometimes pruning. Tree crowns contain the active photosynthesis region and play a vital role in tree growth. Among various tree crown measurements, live crown ratio (LCR) is derived from height to live crown base (HLCB) and total tree height. Accurate measurement of HLCB is basis for live crown ratio prediction. Due to numerous definitions and practical considerations, HLCB and crown structure are difficult and slow to measure accurately. Despite this, LCR is a useful predictor in various growth and yield models. Due to the challenges in measuring tree crowns, accurate live crown ratio prediction models are useful. The LCR model of (Dyer and Burkhart, 1987) was refit with intensively managed plantation (IMP) data. The parameters were significant, and the residual plots showed no concerning patterns but the prediction of height to live crown base for pruned trees was not logical as it sometimes predicted HLCB lower than pruning height. To address this, the base model was modified to accommodate the pruning effect and provide logical predictions. LCR is in range of 0 - 1 and HLCB is greater or equal to pruning height. If trees are not pruned, it reverts to the original model. The models were validated with a dataset of IMP measurements not used in fitting. Validation statistics suggest the model performs nearly as well as the original, unconstrained base model. It is expected that the new model will be useful for forest managers to predict LCR of both pruned and unpruned trees. The second part of the study is to understand the importance of common tree variables in predicting the lumber recovery in planted loblolly pine. A random forest model was used to determine the variable importance of DBH, total tree height and live crown ratio for total board ft., high-grade, and high-grade lumber proportion compared to total board ft. DBH ranked at the first position followed by total tree height and live crown ratio similarly ranked for volume and high-grade lumber volume. For proportions of high-grade lumber, tree height was at top rank followed by LCR, and DBH. However, the effect of these variables for lumber recovery was not explored. It is suggested that future work can explore parametric model forms for accurately predicting lumber recovery using simple, easy to measure tree variables. / Master of Science / Loblolly pine plantations are found in large areas of southern United States. Significant investments are made for maximizing wood production and economic benefits. The growth of trees is largely governed by its crown which is the green foliage found in upper parts of tree. Planted trees are often thinned to provide crown sufficient light for improved growth. Tree crowns are sometimes pruned to reduce knot size and improve lumber quality. If we can understand how these activities affected the crown length of trees, we can manage our plantation accordingly. With this motivation, the live crown ratio which is the ratio of length of live crown and total tree height was modeled. An existing live crown ratio model for planted loblolly pine trees was modified to provide logical predictions for pruned trees and was updated with newer data. Simple tree measurements like tree diameter, height and age were used in the model. Common tree variables were used because LCR is derived from height to live crown base which is difficult to measure in field. The new model will be useful to forest managers for predicting the growth of loblolly pine plantations subjected to pruning. Similarly, an attempt is made to explore the various standing tree characteristics affecting the amount of lumber that can be obtained. This will assist in understanding the lumber grade by using common tree measurements and value of a stand from lumber production side can be known.

Tendo em vista a importância de se conhecer as exigências climáticas da videira nas condições de cultivo, o objetivo deste trabalho foi determinar os índices biometeorológicos: graus-dia (GD), índice biometeorológico de Primault (IB), índice heliotérmico de Huglin (IH) e índice heliotérmico de Geslin (IHG) para a cultivar Niagara Rosada podada em diferentes épocas e fases do ciclo vegetativo. A poda seca foi realizada no final do inverno (fase de mobilização de reservas) e a poda verde foi realizada no verão (fase de acúmulo de reservas). O delineamento experimental utilizado foi o de blocos ao acaso, constituídos os tratamentos de 9 datas de poda: 20/07, 04/08, 19/08, 03/09, 18/09, 01/02, 15/02, 02/03 e 16/03 em dois anos agrícolas (2005/2006 e 2006/2007). A parcela experimental foi composta de 12 plantas sendo duas bordaduras (dez plantas úteis) com treze anos de idade, no espaçamento de 2,0 m x 1,0 m, conduzidas no sistema de espaldeira sob irrigação. As variáveis analisadas foram os períodos e subperíodos, em dias, poda-colheita, brotação-colheita, podabrotação, brotação-florescimento, florescimento-"veraison" e "veraison"-colheita. Para a análise de variância dos índices biometeorológicos foi utilizado o teste de Tukey para comparar as médias das datas de poda, duas a duas, e o teste F para contrastes ortogonais comparando a média da poda seca versus poda verde. Na análise do índice GD e IH verificou-se que plantas podadas em épocas distintas apresentaram valores diferentes para completar o período poda-colheita. Para IB e IHG verificou-se a mesma tendência em plantas podadas em épocas distintas, apresentando valores diferentes para completar o período brotação-colheita. Os subperíodos que não diferiram estatisticamente entre as épocas de poda foram respectivamente: florescimento- "veraison" para o IB e brotação-florescimento para o IH e IHG. Os demais subperíodos apresentaram diferenças pelo teste F para contraste ortogonal para as épocas de poda. / The goal of this work was to determine four biometeorological indexes: degreedays (DD), Primault's biometeorological index (BI), Huglin's heliothermal index (HI) and Geslin's heliothermal index (GHI) for grape 'Niagara Rosada' (Vitis labrusca L.) pruned at different seasons and physiological phases. The dry pruning was done at the end of winter (carbohydrate mobilization phase) and green pruning (carbohydrate stock phase) was done during summer. The experiment was carried out in a randomized block design consisting of 9 treatments, pruning dates 20th July, 04th August, 19th August, 03rd September, 18th September, 01st February, 15th February, 02nd March and 16th March during two years (2005/2006 and 2006/2007). Twelve plants were used as experimental plot, being two border plants. The grapevine was 13 years old grown under trellis system with a 2.0 x 1.0 meter spacing under irrigation. The grape phenology was evaluated considering numbers of days for each one of the following period and sub-periods: pruning - harvest, sprout - harvest, pruning - sprout, sprout - flowering, flowering - early ripening and early ripening - harvest. To analyze those indexes, the Tukey test was used comparing pruning dates and F test for orthogonal contrast comparing plants pruned at different seasons (dry pruning versus green pruning). It was observed that plants pruned at different seasons do not require the same amount of energy to complete pruning - harvest period when using DD and HI. The same tendency was observed when using BI and GHI for those plants to complete its period sprout - harvest. Some sub-periods that did not differ statistically between dry pruning and green pruning: flowering - early ripening for BI and sprout - flowering for HI and GHI. Other sub-periods differ statistically when using F test for orthogonal contrast.

A uva cv Niágara Rosada tem uma grande importância econômica na viticultura do estado de São Paulo, principalmente na região de Campinas, sendo consumida como fruta fresca no mercado interno. Essa cultivar pode produzir de 10 a 15 toneladas por hectare, cachos com massa de 100 a 300 g, com produção concentrada nos meses de dezembro a fevereiro; sendo que atualmente é comum se obter uma segunda safra proveniente da poda de verão, com colheita entre maio a julho. Porém a obtenção de duas safras todo ano pode reduzir a vida útil e apresentar efeito depressivo à videira. O objetivo deste trabalho foi avaliar a influência de diferentes freqüências de poda de produção e a proteção química de doenças na produção e qualidade dos frutos. Foram avaliados três tratamentos na condução de uva no sistema de espaldeira: T1: cinco podas em dois anos e meio (duas podas por ano), com tratamento fitossanitário para controle do míldio da videira (Plasmopora viticola), T2: quatro podas em dois anos e meio (1,5 podas por ano), com tratamento fitossanitário para controle do míldio da videira, T3: cinco podas em dois anos e meio (duas podas por ano), sem tratamento fitossanitário para controle do míldio. Em cada ciclo de produção foram avaliados: a) número de cachos por planta, b) massa dos cachos por planta em quilogramas , c) rendimento em quilogramas por hectare; d) SST (° Brix) dos frutos, e) porcentagem de severidade de míldio nas folhas, f) porcentagem de desfolha após a colheita. A partir da terceira safra observaram-se diferenças significativas entre os tratamentos, quanto ao número de cachos por planta, massa dos cachos, rendimento e sólidos solúveis totais, mostrando uma resposta estatisticamente positiva no tratamento T2 (com n - 1 podas realizadas, onde n: número de podas de verão realizadas e com tratamento fitossanitário). O tratamento fitossanitário mostrou resultados significativos quanto a redução na severidade de doença nas folhas e redução da desfolha após a colheita. / The cultivation of grapes \'hp Niagara Rosada\' has a great economic importance in viticulture in the state of Sao Paulo, mainly in the region of Campinas, being consumed as fresh fruit in the domestic market. This cultivar can produce 10-15 tons per hectare, clusters with mass 100-300 g, with manufacturing concentrated in the months from December to February, and currently it is common to get a second harvest from the summer pruning, to yield between May-July. But to obtain two harvests every year can reduce the life and present depressive effect on the vine. The objective of this study was to evaluate the influence of different pruning frequencies of production and chemical protection of disease in production and fruit quality. Three treatments were evaluated in the conduct of the grape trellis system: T1: pruning five in two and a half years (two cuttings per year), with treatment plant for control of grape downy mildew (Plasmopora viticola), T2: four pruning in two years and a half (1.5 cuttings per year), with treatment to control downy mildew of the vine, T3: pruning five in two and a half years (two cuttings per year), without a treatment to control downy mildew. In each production cycle were evaluated: a) number of clusters per plant, b) mass of clusters per plant in kilograms per plant, c) yield in kilograms per hectare; d) SST (° Brix) of fruits and) percentage of severity of mildew on the leaves, f) percentage of defoliation after harvest. From the third harvest there were significant differences between treatments regarding the number of bunches per plant, cluster weight, yield and total soluble solids, demonstrating a statistically positive response in treatment T2 (with n - 1 pruning, where n : number of summer pruning performed and treatment plant). The treatment plant showed significant results regarding the reduction in disease severity on leaves and reduction in defoliation after harvest.

A uva cv Niágara Rosada tem uma grande importância econômica na viticultura do estado de São Paulo, principalmente na região de Campinas, sendo consumida como fruta fresca no mercado interno. Essa cultivar pode produzir de 10 a 15 toneladas por hectare, cachos com massa de 100 a 300 g, com produção concentrada nos meses de dezembro a fevereiro; sendo que atualmente é comum se obter uma segunda safra proveniente da poda de verão, com colheita entre maio a julho. Porém a obtenção de duas safras todo ano pode reduzir a vida útil e apresentar efeito depressivo à videira. O objetivo deste trabalho foi avaliar a influência de diferentes freqüências de poda de produção e a proteção química de doenças na produção e qualidade dos frutos. Foram avaliados três tratamentos na condução de uva no sistema de espaldeira: T1: cinco podas em dois anos e meio (duas podas por ano), com tratamento fitossanitário para controle do míldio da videira (Plasmopora viticola), T2: quatro podas em dois anos e meio (1,5 podas por ano), com tratamento fitossanitário para controle do míldio da videira, T3: cinco podas em dois anos e meio (duas podas por ano), sem tratamento fitossanitário para controle do míldio. Em cada ciclo de produção foram avaliados: a) número de cachos por planta, b) massa dos cachos por planta em quilogramas , c) rendimento em quilogramas por hectare; d) SST (° Brix) dos frutos, e) porcentagem de severidade de míldio nas folhas, f) porcentagem de desfolha após a colheita. A partir da terceira safra observaram-se diferenças significativas entre os tratamentos, quanto ao número de cachos por planta, massa dos cachos, rendimento e sólidos solúveis totais, mostrando uma resposta estatisticamente positiva no tratamento T2 (com n - 1 podas realizadas, onde n: número de podas de verão realizadas e com tratamento fitossanitário). O tratamento fitossanitário mostrou resultados significativos quanto a redução na severidade de doença nas folhas e redução da desfolha após a colheita. / The cultivation of grapes \'hp Niagara Rosada\' has a great economic importance in viticulture in the state of Sao Paulo, mainly in the region of Campinas, being consumed as fresh fruit in the domestic market. This cultivar can produce 10-15 tons per hectare, clusters with mass 100-300 g, with manufacturing concentrated in the months from December to February, and currently it is common to get a second harvest from the summer pruning, to yield between May-July. But to obtain two harvests every year can reduce the life and present depressive effect on the vine. The objective of this study was to evaluate the influence of different pruning frequencies of production and chemical protection of disease in production and fruit quality. Three treatments were evaluated in the conduct of the grape trellis system: T1: pruning five in two and a half years (two cuttings per year), with treatment plant for control of grape downy mildew (Plasmopora viticola), T2: four pruning in two years and a half (1.5 cuttings per year), with treatment to control downy mildew of the vine, T3: pruning five in two and a half years (two cuttings per year), without a treatment to control downy mildew. In each production cycle were evaluated: a) number of clusters per plant, b) mass of clusters per plant in kilograms per plant, c) yield in kilograms per hectare; d) SST (° Brix) of fruits and) percentage of severity of mildew on the leaves, f) percentage of defoliation after harvest. From the third harvest there were significant differences between treatments regarding the number of bunches per plant, cluster weight, yield and total soluble solids, demonstrating a statistically positive response in treatment T2 (with n - 1 pruning, where n : number of summer pruning performed and treatment plant). The treatment plant showed significant results regarding the reduction in disease severity on leaves and reduction in defoliation after harvest.

Tendo em vista a importância de se conhecer as exigências climáticas da videira nas condições de cultivo, o objetivo deste trabalho foi determinar os índices biometeorológicos: graus-dia (GD), índice biometeorológico de Primault (IB), índice heliotérmico de Huglin (IH) e índice heliotérmico de Geslin (IHG) para a cultivar Niagara Rosada podada em diferentes épocas e fases do ciclo vegetativo. A poda seca foi realizada no final do inverno (fase de mobilização de reservas) e a poda verde foi realizada no verão (fase de acúmulo de reservas). O delineamento experimental utilizado foi o de blocos ao acaso, constituídos os tratamentos de 9 datas de poda: 20/07, 04/08, 19/08, 03/09, 18/09, 01/02, 15/02, 02/03 e 16/03 em dois anos agrícolas (2005/2006 e 2006/2007). A parcela experimental foi composta de 12 plantas sendo duas bordaduras (dez plantas úteis) com treze anos de idade, no espaçamento de 2,0 m x 1,0 m, conduzidas no sistema de espaldeira sob irrigação. As variáveis analisadas foram os períodos e subperíodos, em dias, poda-colheita, brotação-colheita, podabrotação, brotação-florescimento, florescimento-"veraison" e "veraison"-colheita. Para a análise de variância dos índices biometeorológicos foi utilizado o teste de Tukey para comparar as médias das datas de poda, duas a duas, e o teste F para contrastes ortogonais comparando a média da poda seca versus poda verde. Na análise do índice GD e IH verificou-se que plantas podadas em épocas distintas apresentaram valores diferentes para completar o período poda-colheita. Para IB e IHG verificou-se a mesma tendência em plantas podadas em épocas distintas, apresentando valores diferentes para completar o período brotação-colheita. Os subperíodos que não diferiram estatisticamente entre as épocas de poda foram respectivamente: florescimento- "veraison" para o IB e brotação-florescimento para o IH e IHG. Os demais subperíodos apresentaram diferenças pelo teste F para contraste ortogonal para as épocas de poda. / The goal of this work was to determine four biometeorological indexes: degreedays (DD), Primault's biometeorological index (BI), Huglin's heliothermal index (HI) and Geslin's heliothermal index (GHI) for grape 'Niagara Rosada' (Vitis labrusca L.) pruned at different seasons and physiological phases. The dry pruning was done at the end of winter (carbohydrate mobilization phase) and green pruning (carbohydrate stock phase) was done during summer. The experiment was carried out in a randomized block design consisting of 9 treatments, pruning dates 20th July, 04th August, 19th August, 03rd September, 18th September, 01st February, 15th February, 02nd March and 16th March during two years (2005/2006 and 2006/2007). Twelve plants were used as experimental plot, being two border plants. The grapevine was 13 years old grown under trellis system with a 2.0 x 1.0 meter spacing under irrigation. The grape phenology was evaluated considering numbers of days for each one of the following period and sub-periods: pruning - harvest, sprout - harvest, pruning - sprout, sprout - flowering, flowering - early ripening and early ripening - harvest. To analyze those indexes, the Tukey test was used comparing pruning dates and F test for orthogonal contrast comparing plants pruned at different seasons (dry pruning versus green pruning). It was observed that plants pruned at different seasons do not require the same amount of energy to complete pruning - harvest period when using DD and HI. The same tendency was observed when using BI and GHI for those plants to complete its period sprout - harvest. Some sub-periods that did not differ statistically between dry pruning and green pruning: flowering - early ripening for BI and sprout - flowering for HI and GHI. Other sub-periods differ statistically when using F test for orthogonal contrast.

Made available in DSpace on 2016-12-08T16:44:52Z (GMT). No. of bitstreams: 1 PGPV05MA007.pdf: 712076 bytes, checksum: ae4a73c50f7dfa4757824c3d125bd9a0 (MD5) Previous issue date: 2005-12-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This work was carried out to study the effects of roots pruning with cupper sulphate, cupper oxychlorate and ethephon (2-cloroethtlphosphonic acid) on quality and growth of Pinus taeda L. seedlings. The experiment was conducted in a commercial nursery in Lages, SC. The treatments with cupper sulphate or cupper oxychlorate were applied by immersing the containers in a mixture of equal parts of water and latex paint, at cupper doses of 0, 12, 24, 36, and 48g L-1 of solution (paint + water). The ethefon was sprayed to the aerial part of seedlings (10-15cm height) at 0, 50, 75, 100, and 125 mg (a.i.) L-1. The experiment followed a completely randomized block design with five replicates. Eight months old seedlings were assessed for height, aerial dry matter, root dry matter (principal and lateral roots), collar diameter, number of lateral roots at three root portions (superior, medium, and lower parts), Dickson Quality Index (DQI), height/collar diameter ratio, height/aerial dry matter ratio, and aerial dry matter/root dry matter ratio. Considering the increase of product dose cupper sources, only cupper sulphate reduced (P < 0.01) the attributes of root dry matter (total, principal and lateral roots), collar diameter, and seedlings height. The increase of cupper sulphate dose reduced the total root dry matter mainly by inhibiting the development of lateral roots. The increment of cupper sulphate doses resulted in increase of copper concentration in the substrate and in the root tissues of Pinus seedlings than the increase of cupper oxychlorate doses. As a result seedlings treated with increasing doses of cupper sulphate showed significant reduction of DQI and aerial part growth, indicating cupper toxicity. The increment of cupper oxychlorate doses did not reduce the total root dry matter, but increased the number of lateral roots at the median and higher portion of the root system. The cupper oxychlorate, even at the highest dose, did not reduce the DQI and the aerial part growth, indicating the absence of cupper toxicity. Then, despite of its root pruning effect, cupper sulphate significantly reduced the emission of lateral roots, as opposed to the effect of cupper oxychlorate. Therefore, the increase of cupper oxychlorate doses seems to show a more positive effect than the increase of cupper sulphate doses on root pruning, since a seedling with a large number of lateral roots of small length is desirable aiming the seedling establishment after transplantation. The increase of ethefon dose sprayed at seedlings aerial part, to release ethylene, was not effective for root pruning. The increase of ethefon dose only increased the principal root dry matter. There was no effect of ethefon on the other seedling growth attributes / O presente trabalho teve como objetivo avaliar os efeitos da poda química das raízes com sulfato de cobre, oxicloreto de cobre e ethefon (ácido 2-cloroetilfosfônico) sobre o crescimento e qualidade de mudas de Pinus taeda L. produzidas em tubetes. O experimento foi conduzido em viveiro comercial de produção de mudas no município de Lages, SC. Os tratamentos a base de sulfato de cobre ou oxicloreto de cobre foram feitos através da imersão dos tubetes em uma mistura de partes iguais de água e tinta látex comum, contendo 0, 12, 24, 36 e 48g de cobre L-1 de solução (tinta + água). O ethefon foi aplicado via aérea em mudas com 10-15cm de altura nas concentrações de 0, 50, 75, 100 e 125 mg (i.a.) L-1. Foi utilizado o delineamento em blocos ao acaso com cinco repetições. As variáveis analisadas em mudas com oito meses de idade foram alturas de parte aérea, massa seca da parte aérea e de raízes (total, de raiz primária e de raízes secundárias), diâmetro do coleto, número de raízes secundárias em três porções do sistema radicular (superior média e inferior), Índice de Qualidade de Dickson (IQD), e as relações altura/diâmetro do coleto, altura/massa seca de parte aérea e massa seca de parte aérea/massa seca do sistema radicular. Considerando o aumento da dose aplicada das fontes de cobre testadas, apenas o sulfato de cobre ocasionou redução significativa (P < 0,01) na massa seca de raízes, no diâmetro do coleto, na altura e na massa seca de raízes secundárias. O incremento nas doses de sulfato de cobre promoveu redução na massa seca total do sistema radicular, especialmente através da inibição do desenvolvimento de raízes secundárias. O incremento nas doses de sulfato de cobre ocasionou aumento na concentração de cobre presente no substrato e nos tecidos do sistema radicular das mudas em relação ao oxicloreto de cobre. Desta forma, as mudas que receberam tratamento com as doses mais altas de sulfato de cobre apresentaram redução no IQD e no crescimento da parte aérea, demonstrando claramente sintoma de toxidez por cobre. O incremento nas doses de oxicloreto de cobre não causou redução na massa seca total do sistema radicular, mas promoveu aumento no número de raízes secundárias, concentradas especialmente nas porções mediana e superior do sistema radicular. O oxicloreto de cobre, mesmo nas doses mais elevadas, não ocasionou redução no IQD e no crescimento da parte aérea, indicando que não houve toxidez pelo cobre. Portanto, apesar do efeito promotor do sulfato de cobre na poda química do sistema radicular, o produto reduziu significativamente a emissão de raízes laterais, o mesmo não ocorrendo com o oxicloreto de cobre. Desta forma, o incremento nas doses de oxicloreto de cobre parece apresentar um efeito positivo maior em relação ao sulfato de cobre, já que um sistema radicular com maior número de raízes secundária curtas é desejável visando estabelecimento das mudas a campo. O incremento nas doses de ethefon aplicados na parte aérea das mudas, visando a liberação de etileno, não foi efetivo na poda química de raízes. O incremento nas doses de ethefon apenas ocasionou aumento significativo no acúmulo de matéria seca de raiz primária. Não houve efeito de dose do ethefon sobre os demais atributos de crescimento analisados

This final thesis covers the basics of artificial neural networks, with focus on supervised learning, pruning and the problem of achieving good generalization ability. An empirical investigation is conducted on twelve dierent problems originating from the Proben1 benchmark collection.The results indicate that pruning is more likely to improve generalization if the data is sensitive to overtting or if the networks are likely to be trapped in local minima.

Please read the abstract in the dissertation / Dissertation (MSc (Agric))--University of Pretoria, 2001. / Plant Production and Soil Science / unrestricted