Submitted by Jorge Silva (jorgelmsilva@ufrrj.br) on 2016-10-20T19:09:28Z No. of bitstreams: 1 2016 - Carlos Magno Moreira de Oliveira.pdf: 3071191 bytes, checksum: 379c33cebd9108d71d2092ecebb6104d (MD5) / Made available in DSpace on 2016-10-20T19:09:28Z (GMT). No. of bitstreams: 1 2016 - Carlos Magno Moreira de Oliveira.pdf: 3071191 bytes, checksum: 379c33cebd9108d71d2092ecebb6104d (MD5) Previous issue date: 2016-02-26 / CAPES / Commercial plantations of Eucalyptus has increased in Brazil, mainly in the state of Minas Gerais. The assessment of the effects of soil and climatic variables on the growth of trees becomes important tool to optimize management activities. In this perspective, this study aims to adjust and parameterize the 3-PG model for the North of Minas Gerais, using data from permanent plots of forest inventory, climate data from weather stations and data from remote sensing. This study was conducted in Eucalyptus of commercial plantations in the city of Bocaiuva (Minas Gerais). To parameterize the constant parameters of the relationship between DAP and height (H) relationship of power between DAP and height (NHB), constant relationship between volume and DAP (aV) and relationship of power between volume and DAP (NVB) were used in 2031 comments. The PG-3 model was run with parameters suggested in literature for the region and then executed with changes in the parameters mentioned above. To evaluate the effect of climate variability on planting were assessed estimates of 3-PG model fed with climatological normal and monthly averages and then analyzed the effects of environmental modulators fans and fDPV on the growth of planting. The database of climatological normals were obtained from INMET and global solar radiation was obtained from the GL 1.2 model for South America. The database with monthly average was acquired by INMET, and the precipitation was obtained from tropical TRMM satellite and solar radiation was estimated by the model proposed by FAO 56. Subsequently, we evaluated the efficiency of the 3-PG model in estimating the growth and production of settlement regarding estimates of empirical models, considering the FR parameter (Fertility Rating) as a site quality indicator. Through significant improvement was observed statistical analysis of procedural modeling with the modification of the suggested parameters, indicating that the values of this parameter can be used for other municipalities that region. Statistical analyzes showed that there was a reduction of variation of the data through the standard error of estimate (Syx%) and model efficiency (EF) greater than 0.80 for DBH, height and volume, indicating significant improvement in the model estimates . Comparing the estimates of dendrometric variables with climatological normal and monthly averages, it is observed, the graphics, the estimates were similar, with the monthly averages showed higher estimates. Statistical analysis demonstrated that the monthly averages were closer to those observed data, Syx 28% and EF 0.82 to volume while normal climatic Syx% showed almost 60% and EF 0.22. fAS modulator showed greater restriction on the development of trees, especially in the months of July, August and September. The procedural modeling showed average yield estimates in respect of empirical modeling, and the site estimated by 3-PG showed values close to the average estimated by site traditional model. The FR parameter on soil fertility, showed good relationship with the classification of production sites, and further studies to improve the accuracy in the classification of sites indexes. The 3-PG model can make accurate estimates of dendrometric variables, showing an important tool for evaluating areas with fitness for forest production, considering the effect of environmental and climate variables in the growth of trees. The procedural modeling can be used together with empirical tools to assess the growth and production of forest plantations on different management activities and in the future producing areas. / Plantios comerciais de Eucalyptus tem aumentado no Brasil, principalmente no estado de Minas Gerais. A avalia??o dos efeitos das vari?veis ed?ficas e clim?ticas sobre o crescimento das ?rvores torna-se ferramenta importante para otimizar as atividades de manejo. Nessa perspectiva, esse trabalho tem por objetivo ajustar e parametrizar o modelo 3-PG para regi?o Norte de Minas Gerais, utilizando dados de parcelas permanentes de invent?rio florestal, dados clim?ticos de esta??es climatol?gicas e dados oriundos de sensoriamento remoto. Esse estudo foi realizado em plantios comerciais do g?nero Eucalyptus no munic?pio de Bocai?va (MG). Para parametrizar os par?metros constante de rela??o entre DAP e altura (aH), pot?ncia de relacionamento entre DAP e altura (nHB), constante de relacionamento entre volume e DAP (aV) e pot?ncia de relacionamento entre volume e DAP (nVB) foram utilizados 2031 observa??es. O modelo 3-PG foi executado com par?metros sugeridos na bibliografia para regi?o e posteriormente executada com modifica??es nos par?metros mencionados anteriormente. Para avaliar o efeito da variabilidade clim?tica sobre o plantio, foram avaliadas as estimativas do modelo 3-PG alimentado com normais climatol?gicas e m?dias mensais e posteriormente analisado os efeitos dos moduladores ambientais fAS e fDPV sobre o crescimento do plantio. O banco de dados das normais climatol?gicas foram obtidas do INMET e a radia??o solar global foi obtido do modelo GL 1.2 para am?rica do sul. O banco de dados com m?dias mensais foi adquirido pelo INMET, sendo que a precipita??o foi obtida do sat?lite tropical TRMM e a radia??o solar global foi estimado pelo modelo proposto pela FAO 56. Posteriormente, foi avaliado a efici?ncia do modelo 3-PG em estimar o crescimento e produ??o do povoamento em rela??o as estimativas dos modelos emp?ricos, considerando o par?metro FR (Fertility Rating) como indicador da qualidade de s?tio. Por meio de an?lises estat?sticas foi observado melhora significativa da modelagem processual com a modifica??o dos par?metros sugeridos, indicando que os valores dessa parametriza??o podem ser utilizados para outros munic?pios dessa regi?o. As an?lises estat?sticas demonstram que houve redu??o da varia??o dos dados por meio do erro padr?o da estimativa (Syx%) e efici?ncia do modelo (EF) maior que 0,80 para as vari?veis DAP, Altura e Volume, indicando melhoria significativa nas estimativas do modelo. Comparando-se as estimavas das vari?veis dendrom?tricas com normais climatol?gicas e medias mensais, observa-se, pelos gr?ficos, que as estimativas foram semelhantes, sendo que as medias mensais apresentaram estimativas superiores. A an?lise estat?stica demonstra que as medias mensais foram mais pr?ximas aos dados observados, com Syx% 28 e EF de 0,82 para volume, enquanto que as normais climatol?gicas apresentaram Syx% de quase 60% e EF de 0,22. O modulador fAS apresentou maior restri??o no desenvolvimento das ?rvores, principalmente nos meses de julho, agosto e setembro. A modelagem processual apresentou estimativas de produ??o m?dia em rela??o a modelagem emp?rica, sendo que o s?tio estimado pelo 3-PG apresentou valores pr?ximos ao s?tio m?dio estimado pela modelam tradicional. O par?metro FR, relativo a fertilidade do solo, apresentou boa rela??o com a classifica??o de s?tios de produ??o, sendo necess?rios outros estudos para melhorar a acur?cia na classifica??o dos ?ndices de s?tios. O modelo 3-PG pode realizar estimativas precisas de vari?veis dendrom?tricas, mostrando-se ferramenta importante para avalia??o de ?reas com aptid?o para produ??o florestal, por considerar o efeito de vari?veis ambientais e clim?ticas no crescimento das ?rvores. A modelagem processual pode ser utilizada juntamente com ferramentas emp?ricas para avaliar o crescimento e produ??o de plantios florestais sobre diferentes atividades de manejo e tamb?m em ?reas futuramente produtoras.

Dans de nombreuses régions tropicales dont certaines zones du Brésil, les plantations industrielles d'Eucalyptus connaissent une expansion rapide. L'impact environnemental et la durabilité de ces plantations soulèvent de nombreuses questions, notamment en termes d'évolution du bilan carboné et de la fertilité des sols, et d'impact sur les ressources en eau. Cette thèse contribue à apporter des réponses à ces questions, en étudiant les bilans de carbone, d'eau et d'azote des plantations d'Eucalyptus de l'état de São Paulo. L'approche privilégiée est celle de la modélisation écophysiologique, et l'échelle visée in fine est celle de grandes plantations. Cette thèse présente une première approche empirique d'estimation spatialisée de la production des plantations, basée sur la construction et l'interprétation de séries temporelles d'indice de végétation du satellite MODIS sur un ensemble de parcelles d'âge et de productivité contrastés. Ensuite une approche plus complète de modélisation écophysiologique est développée, basée sur le modèle C-H2O-N « G'DAY » qui a été adapté au cas des plantations brésiliennes. La simulation par le modèle des flux de C, H2O et N sur l'ensemble de parcelles évoqué précédemment, à l'échelle de la rotation, a été réalisée avec et sans la contribution de données d'indice foliaire, obtenues à partir des séries temporelles de réflectance MODIS. Ces séries temporelles de données satellitaires sont riches en informations sur la variabilité spatio-temporelle de la productivité, qui est bien reproduite par le modèle écophysiologique proposé / Industrial Eucalyptus plantations are expanding rapidly in many tropical regions, and in particular in south-eastern Brazil. Their sustainability and environmental impact are an issue of concern, and questions have been raised regarding their carbon balance and effects on soil fertility and water ressources. This thesis addresses some of these questions, by studying the carbon, water and nitrogen balance of Eucalyptus plantations in São Paulo State.A process-based modelling approach was adopted, with the ultimate aim of application at the scale of large plantations. A first empirical approach was developed to estimate wood production on a spatial basis, involving the construction and interpretation of MODIS vegetation index time series on a set of Eucalyptus stands of contrasted age and productivity levels. The complete ecophysiological C-H2O-N cycling model "G'DAY" was then adapted for Brazilian plantations. Rotation-length simulations of C, H2O and N fluxes were carried out on the afore-mentioned set of plantation stands, either with or without the integration of leaf area index data obtained from MODIS reflectance time-series. The satellite time series carry useful information about the spatio-temporal variability of productivity, which is also well reproduced by the proposed ecophysiological model

Le colza est une oléoprotéagineuse exigeante en soufre (S) mais caractérisée par une faible efficience d’usage du S (EUS). La baisse des retombés atmosphériques soufrées, l'existence de fortes interactions entre les métabolismes soufrés et azotés et l’augmentation prédite des températures terrestres peuvent conduire à une altération des rendements et de la qualité des graines de colza. Dans ce contexte, en s’appuyant sur des approches in planta (conditions contrôlées et de plein champ) et in silico (expérimentations numériques via le modèle écophysiologique SuMoToRI "Sulphur Model Towards Rapeseed Improvement"), les principaux objectifs de cette thèse étaient d’étudier l’impact (i) de différentes stratégies de fertilisation S et N, (ii) du changement climatique et (iii) de la variabilité des paramètres « plante » du modèle sur la croissance ainsi que sur les composantes du rendement et la qualité des graines de colza.Cette étude a permis de confirmer les effets synergiques des apports de S et de N ainsi que leurs effets antagonistes lors d’apport excessif d’un des deux éléments sur l’EUS et l’EUN, soulignant l'importance d'équilibrer les apports S/N. Décaler l’apport en S a permis d’améliorer la qualité protéique des graines en augmentant l’abondance relative en napines (protéines de réserve des graines riches en cystéine). Deux indices de la qualité protéique des graines ont pu être proposés : (i) la teneur en S des graines, fortement corrélée avec l'abondance relative en napines et (ii) le ratio napine:cruciférine-30kDa (cruciférines : protéines de réserve pauvres en S), permettant d’apprécier l’équilibre des apports S/N. Les simulations réalisées avec le modèle ont montré que des ajustements de la fertilisation S devront s’opérer dans un contexte d’augmentation des températures et/ou de diminution du rayonnement incident conduisant à une baisse de la biomasse et à une augmentation du S stocké dans les feuilles. Ces résultats requestionnent les schémas conventionnels de fertilisation et l’utilisation d’idéotypes variétaux et culturaux adaptés aux schémas de fertilisation S et N ainsi qu’au dérèglement climatique. / Oilseed rape is an oleoproteaginous crop with high sulfur (S) demanding and characterized by a low S use efficiency (SUE). The decline in atmospheric S deposition, the existence of strong interactions between S and nitrogen (N) metabolism and the predicted increase in terrestrial temperatures can lead to an alteration in seeds yields and quality. In this context, using in planta (controlled conditions and fields experiments) and in silico (numerical experiments through an agro-ecophysiological model SuMoToRI, “Sulfur Model Towards Rapeseed Improvement”) approaches, the main objectives of this thesis were to study the impact of (i) different S and N fertilization strategies, (ii) climate change and (iii) the variability of the model’s “plant” parameters on growth as well as the yield components and the seeds quality of oilseed rape.This study confirmed the synergistic effects of S and N inputs and their antagonistic effects when of one of the two elements was in excessive rate on EUS and EUN, highlighting the importance of balancing S/N inputs. Delaying S input has improved seed protein quality by increasing the relative abundance of napins (cysteine-rich seed storage proteins). Two seed protein quality indices could be proposed (i) seeds S content, strongly correlated with relative napine abundance and (ii) the ratio napins:cruciferins-30kDa (cruciferins : S-poor seed storage proteins). S), making it possible to assess the S/N balance inputs. The simulations carried-out with the model showed that the adjustment of S fertilization must be performed in a context of increasing temperature and/or a reduction of incident radiation which lead to a decrease of biomass and an increase of S stored in leaves. Overall, these results questioned conventional fertilization strategies and the use of varietal and crop ideotypes adapted to S and N fertilization strategies as well as to climate change.