Made available in DSpace on 2014-07-29T16:24:21Z (GMT). No. of bitstreams: 1 Aurelio L A Martinez.pdf: 175334 bytes, checksum: 963d9bfcab566da7ecf7bfad5d5e4878 (MD5) Previous issue date: 2007-08-31 / The banana is one of the most important fruit in the human diet being and, important source of income for small producers, mostly responsible for its production. The explored´s cultivars present diseases problems that hinder its steady production to long it time. With intention to evaluate the agronomics characteristics of six cultivars of banana tree under irrigation, in Goiânia-GO, was carried through an assay in the period of August of 2005 the March of 2007, understanding the first cycle of production. The experimental design was randomized block, with six spot. They had been used dumb piece of corm, planted in 3x2 m spacing. It was evaluated the characteristics vegetative cycle, productive cycle, total cycle, leaf number in the bloom and the harvest, height of plant, circumference of pseudsterm, number of hands, number of fruits, length and diameter of the central fruit of second hand, number of fruits of second hand, mass of second hand and mass of hand. The average total cycles had been of 479 days for BRS FHIA 18, 504 days for Caipira , 481 days for the Nanicão IAC 2001, 531 days in Prata Zulu, 502 days in Thap Maeo, 516 days in Tropical. The plants had not presented problems with leves illnesses, despite Yelow-Sigatoka being seen in leves of BRS FHIA 18 and Thap Maeo . Already Tropical and Zulu Silver had had some plants infecteds for Panama-Disease. The average mass of hands kg in BRS FHIA had been of 12,3 18, 18,2 kg in Caipira, 24,8 kg in IAC 2001, 22,8 kg in Zulu Silver, 21,3 kg in Thap Maeo and 19,0 kg in the Tropical one. The gotten results had led the conclusion that all cultivars present good characteristics for agricultural exploration, especially in domestic orchards. The cultivars present good production in first cycle, with exception of BRS FHIA 18. Tropical and Prata Zulu are sensible to the Fusarium oxysporum f. sp. cubense. / A banana é uma das frutas mais importantes na dieta humana e importante fonte de renda para pequenos produtores, responsáveis pela maior parte da sua produção. As cultivares mais difundidas apresentam problemas fitossanitários que impedem sua produção estável ao longo do tempo. Com intuito de avaliar as características agronômicas de seis cultivares de bananeira sob irrigação, em Goiânia-GO, foi realizado um ensaio no período de agosto de 2005 a março de 2007, compreendendo o primeiro ciclo de produção. O delineamento experimental foi em blocos casualizados, com seis repetições. Foram utilizadas mudas pedaço de rizoma, plantadas em espaçamento 3x2 m. Foram avaliadas as características ciclo vegetativo, ciclo produtivo, ciclo total, número de folhas no florescimento e na colheita, altura da planta, circunferência do pseudocaule, número de pencas, número de frutos, comprimento e diâmetro do fruto central da segunda penca, número de frutos da segunda penca, massa da segunda penca e massa de pencas. Os ciclos totais médios foram de 479 dias para BRS FHIA 18 , 504 dias para a Caipira , 481 dias para a Nanicão IAC 2001 , 531 dias na Prata Zulu , 502 dias na Thap Maeo , 516 dias na Tropical . As plantas não apresentaram problemas com doenças foliares, apesar da Sigatoka-Amarela ser vista em folhas de BRS FHIA 18 e Thap Maeo . Já a Tropical e Prata Zulu tiveram várias plantas infectadas pelo Mal-do-Panamá. As massas médias das pencas foram de 12,3 kg na BRS FHIA 18 , 18,2 kg na Caipira , 24,8 kg na IAC 2001 , 22,8 kg na Prata Zulu , 21,3 kg na Thap Maeo e 19,0 kg na Tropical. Os resultados obtidos levaram à conclusão que todas as cultivares apresentam boas características para exploração agrícola, especialmente em pomares domésticos. As cultivares apresentam boa produção já no primeiro ciclo, com exceção da BRS FHIA 18 . As cultivares Tropical e Prata Zulu são sensíveis ao Fusarium oxysporum f. sp. cubense.

Made available in DSpace on 2014-06-11T19:31:35Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-12-10Bitstream added on 2014-06-13T19:02:00Z : No. of bitstreams: 1 silva_shmg_dr_botfca.pdf: 1183066 bytes, checksum: 90313e74dffe24f7d376280e7532d49f (MD5) / Outros / O presente trabalho investigou a utilização do processamento digital de imagens conjuntamente com a análise de componentes principais e redes neurais artificiais como ferramentas de apoio para uma melhor identificação dos estádios iniciais do desenvolvimento da Sigatoka Negra, em nível de campo, para que medidas de controle sejam adotadas mais rapidamente e, assim, reduzir danos e prejuízos causados pela doença na bananicultura. Foram coletadas imagens digitais de folhas de bananeiras infectadas com a Sigatoka Negra nos estádios 1, 2 e 3, sadia e com fitotoxidez por óleo. A seguir, extraíram-se histogramas dos componentes de imagens no sistema RGB (Red, Green eBlue) para 256 intensidades de cinza das amostras, totalizando 768 variáveis para cada amostra. Com a aplicação de uma técnica de seleção de atributos, a análise de componentes principais, conseguiu-se reduzir as variáveis de entrada de 768 para 11 variáveis canônicas, representado uma redução de 98,6%. Em seguida, considerando-se as variáveis canônicas, realizou-se a fase de classificação com o uso de redes neurais artificiais. De maneira geral, as maiores freqüências de acertos do modelo foram para as classes que mais interessam ao monitoramento da enfermidade, mostrando a robustez do classificador gerado, evidenciada pela baixa probabilidade de classificação incorreta (19%). / This study investigated the application, specifically the digital processing of images, with main components analysis and artificial neural networks as tools to support for a better identification of the primaries stages of the Black Sigatoka, in field level, so that control measures are adopted more quickly and consequently it reduces injuries and damages caused by the disease in the banana crops. It were collected digital images of banana leaves infected with Black Sigatoka in stages 1, 2, and 3, healthy and with oil fitotoxity. To proceed, histograms of the components of images were extracted in the system RGB (Red, Green and Blue) for 256 intensities shades of gray of the samples, totaling 768 variables for each sample. With the application of a technique of selection of attributes, the main components analysis, it was possible to reduce the variables of entrance of 768 for 11 canonical variables, represented a reduction of 98,6%. Therefore, being considered the canonical variables, it was accomplished the classification phase with the use of artificial neural networks. In a general way, the largest frequencies of successes of the model went to the classes that more they interest to the control of the diseases, showing the robustness of the generated classifier, evidenced by the low probability of wrong classification (19%).

Made available in DSpace on 2015-03-26T13:51:59Z (GMT). No. of bitstreams: 1 texto completo.pdf: 536332 bytes, checksum: 5951891794c17e210fb52fda2d6fa297 (MD5) Previous issue date: 2013-03-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Devido à ausência de estudo de caracterização da variabilidade genética das populações de Pseudocercospora fijiensis recentemente introduzidas no Brasil, o objetivo desse trabalho foi avaliar a adequabilidade do marcador IRAP para estudar variações genéticas entre indivíduos, bem como determinar a estrutura genética da população brasileira de P. fijiensis com base no fingerprinting gerado por amplificação de polimorfismo entre retrotransposons (IRAP). Um total de 22 locos foi amplificado, sendo 77.3 % polimórficos. A análise de agrupamento revelou dois principais grupos no Brasil. A diversidade gênica (HE) foi de 0.22 e pela análise de variância molecular verificou-se que a maior variabilidade genética está dentro das populações. A Análise Discriminante de Componente Principal (DAPC) revelou que não há nenhuma estruturação relacionada com as origens geográficas e cultiva hospedeiro. O sistema de marcador baseado em retrotransposon IRAP é ferramenta apropriada para estudar a variabilidade genética em P. fijiensis. / Due to the lack of characterization study of genetic variability in populations of Pseudocercospora fijiensis recently introduced in Brazil, the objective of this study was to evaluate the suitability of IRAP marker for studying genetic variations between individuals, and to determine the genetic structure of the population of P. fijiensis based on fingerprinting generated by inter-retrotransposons amplified polymorphism (IRAP). A total of 22 loci were amplified and 77.3% showed a polymorphism. Cluster analysis revealed two major groups in Brazil. The observed genetic diversity (HE) was 0.22, and through molecular analysis of variance, it was determined that the greatest genetic variability occurs within populations. The Discriminant Analysis of Principal Components (DAPC) revealed no structuring related to the geographical origin of culture of the host. The IRAP-based marker system is a suitable tool for the study of genetic variability in P. fijiensis.

Submitted by Marco Antônio de Ramos Chagas (mchagas@ufv.br) on 2017-04-24T12:31:48Z No. of bitstreams: 1 texto completo.pdf: 486412 bytes, checksum: ea7846af4d8a5809764bfecb46fb6ed2 (MD5) / Made available in DSpace on 2017-04-24T12:31:48Z (GMT). No. of bitstreams: 1 texto completo.pdf: 486412 bytes, checksum: ea7846af4d8a5809764bfecb46fb6ed2 (MD5) Previous issue date: 2003-12-17 / A sigatoka negra, causada pelo fungo Mycosphaerella fijiensis, ocasiona perdas que oscilam entre 50 a 100% na produção de bananeira e bananeira-da-terra. Visando obter informação sobre epidemias da sigatoka negra em ambas as musáceas, realizaram-se dois experimentos. No primeiro, avaliaram-se os componentes epidemiológicos: períodos de incubação e latente médio; severidade aos 24 e 40 dias; freqüência de infecção; intervalo, em dias, para atingir a severidade máxima; dias transcorridos do aparecimento de sintomas até atingir a severidade máxima; e as áreas abaixo da curva da severidade e da porção necrosada. Verificou-se que M. fijiensis não foi específico para bananeira ou bananeira-da-terra e que as populações variam quanto à agressividade. A inoculação artificial em mudas pode ser usada para avaliar componentes de resistência de diferentes genótipos à doença. No segundo experimento, estudou-se o progresso da doença em plantações comerciais de bananeira 'Williams' e bananeira-da-terra 'Barraganete', durante as épocas seca e chuvosa. Independentemente da época avaliada, os valores de área abaixo da curva da severidade estimados para bananeira foram maiores que para bananeira-da-terra. Na época chuvosa, o progresso da doença foi maior que na época seca. Em bananeira-da-terra, detectou-se correlação significativa da severidade e o número de horas semanais com temperatura entre 24 e 28°C e umidade relativa maior de 90%, quando se consideraram os valores das variáveis climatológicas registrados quatro e três semanas antes da severidade, nas épocas seca e chuvosa, respectivamente. Não se detectou correlação significativa de severidade e intensidade de precipitação pluviométrica, nas duas épocas avaliadas. Este é o primeiro relato de estudos de sigatoka negra em bananeira-da-terra no Equador, nas épocas seca e chuvosa. É, também, o primeiro estudo de componentes epidemiológicos de isolados equatorianos de M. fijiensis. Os resultados obtidos serão importantes para subsidiar o manejo da sigatoka negra, bem como programas de melhoramento que visem obter resistência a M. fijiensis. / Black sigatoka, caused by the fungus Mycosphaerella fijiensis, is the most important leaf disease of banana and plantain crops, causing 50 to 100% production losses. To generate information regarding black sigatoka epidemics on both Musacea, two experiments were conducted. In the first, isolates of M. Fijiensis obtained from diseased leaves of banana and plantain from several regions of Ecuador, were inoculated in banana and plantain plantlets. The following epidemic components were evaluated: mean incubation and latent periods, initial severity, infection frequency, number of lesions/leaf area, days to reach maximum severity, area under disease progress curve for severity, and area under disease progress curve of leaf necrosis. Regarding all components, it was found that M. fijiensis was not specific for either banana or plantain, and that there is variability in fungal aggressiveness. It was concluded that artificial inoculation of M. fijiensis in plantlets can be used in evaluating disease resistance components of different genotypes. In the second experiment, disease progress was studied in commercial crops of banana 'Williams' and plantain 'Barraganete', during dry and rainy seasons. In both seasons, area under disease progress curve (AUDPC) on banana was higher than AUDPC for iplantain. Values of AUDPC in plantain were higher in the rainy season than in the dry season. Correlation analysis was done between weather variables and disease severity in plantains. Disease severity was not correlated with temperature, relative humidity, or precipitation. Significant correlation was detected between severity and both number of hours with temperature ranging from 24 to 28° C and relative humidity higher than 90%, when these variables were registered four or three weeks before severity assessment, in either dry or rainy seasons, respectively. This is the first report of epidemiological studies of Black Sigatoka in plantains in the dry and rainy seasons, in Ecuador. It is also the first study of aggressiveness components of M. fijiensis, which may become important to assist future breeding programs to obtain resistance against the pathogen. / Dissertação importada do Alexandria

Submitted by (lucia.rodrigues@ufrpe.br) on 2017-03-22T16:19:02Z No. of bitstreams: 1 Otacilio Monteiro da Rocha Junior.pdf: 493982 bytes, checksum: 6eac9d2040ce61a3f47221fcd1d41acc (MD5) / Made available in DSpace on 2017-03-22T16:19:02Z (GMT). No. of bitstreams: 1 Otacilio Monteiro da Rocha Junior.pdf: 493982 bytes, checksum: 6eac9d2040ce61a3f47221fcd1d41acc (MD5) Previous issue date: 2007-07-27 / Conselho Nacional de Pesquisa e Desenvolvimento Científico e Tecnológico - CNPq / Yellow Sigatoka disease or leaf spot Banana caused by fungus Mycosphaerella musicola (anamorph Pseudocercospora musae), it is an important disease of banana disseminated throughout in Brazilian territory. This work aimed to determine the spatial pattern of the Sigatoka disease in the field and appropriate sample’s size for quantification of the disease in different planting areas and severity levels. The spatial pattern of the disease was investigated in three areas of banana’s planting, being delimited in each planting a portion of thirty contiguous lines, with 30 covas/line. Yellow Sigatoka disease was evaluated in each plant-mother and, considering the respective location, the analysis of space spatial’s autocorrelation was made. The three areas presented aggregate pattern of diseased plants, with predominance of the three areas presented pattern of diseased plants, with predominance of the aggregation inside of the lines aggregation inside of the lines. To determinate the appropriate sizes of the samples, it was made pilot’s sampling in 30 plantings, being established in each planting a sub-area of 2 ha, where 50 plants were selected by systematic sampling and evaluated by disease’severity. The severity of the Sigatoka yellow in the 30 plantingareas submitted to pilot’s sampling varied from 5,3% to 46,7%, not being verified significant correlations between the severity levels and the plantings’ age, as well as of these last ones with the sizes of the samples. The samples’ size for quantification of Sigatoka yellow banana’s severity were determinated considering the aggregate pattern of diseased plants. Obtained data were analyzed by method that specifies the acceptable mistake, with a aggregate pattern of diseased plants. The sample’s size were correlated negatively with the levels of disease’ severity (r = -0,60). Considering the planting area’ average, in futures studies of the severity of the Sigatoka yellow, at which excellent precision’ level is demanded (error = 5%), the sampling of 66 plants is recommended for each 2 ha cultivated, with the evaluation of the nine leaves more young/plant. The results obtained in this study serve as base for futures epidemic studies of the Sigatoka yellow banana, as data were originated from fields under different conditions and estimated considering growing needs of precision.Yellow Sigatoka disease or leaf spot Banana caused by fungus Mycosphaerella musicola (anamorph Pseudocercospora musae), it is an important disease of banana disseminated throughout in Brazilian territory. This work aimed to determine the spatial pattern of the Sigatoka disease in the field and appropriate sample’s size for quantification of the disease in different planting areas and severity levels. The spatial pattern of the disease was investigated in three areas of banana’s planting, being delimited in each planting a portion of thirty contiguous lines, with 30 covas/line. Yellow Sigatoka disease was evaluated in each plant-mother and, considering the respective location, the analysis of space spatial’s autocorrelation was made. The three areas presented aggregate pattern of diseased plants, with predominance of the three areas presented pattern of diseased plants, with predominance of the aggregation inside of the lines aggregation inside of the lines. To determinate the appropriate sizes of the samples, it was made pilot’s sampling in 30 plantings, being established in each planting a sub-area of 2 ha, where 50 plants were selected by systematic sampling and evaluated by disease’severity. The severity of the Sigatoka yellow in the 30 plantingareas submitted to pilot’s sampling varied from 5,3% to 46,7%, not being verified significant correlations between the severity levels and the plantings’ age, as well as of these last ones with the sizes of the samples. The samples’ size for quantification of Sigatoka yellow banana’s severity were determinated considering the aggregate pattern of diseased plants. Obtained data were analyzed by method that specifies the acceptable mistake, with a aggregate pattern of diseased plants. The sample’s size were correlated negatively with the levels of disease’ severity (r = -0,60). Considering the planting area’ average, in futures studies of the severity of the Sigatoka yellow, at which excellent precision’ level is demanded (error = 5%), the sampling of 66 plants is recommended for each 2 ha cultivated, with the evaluation of the nine leaves more young/plant. The results obtained in this study serve as base for futures epidemic studies of the Sigatoka yellow banana, as data were originated from fields under different conditions and estimated considering growing needs of precision. / A Sigatoka-amarela, causada pelo fungo Mycosphaerella musicola (anamorfo Pseudocercospora musae, é uma importante doença da bananeira e encontra-se disseminada em todo o território brasileiro. Este trabalho teve como objetivos determinar o padrão espacial da Sigatoka-amarela no campo e os tamanhos ideais das amostras para quantificação da doença em diferentes áreas de plantio e níveis de severidade. O padrão espacial da doença foi investigado em três áreas de plantio de bananeira, sendo em cada plantio delimitada uma parcela de 30 linhas contíguas, com 30 covas/linha. A severidade da Sigatoka-amarela foi avaliada em cada planta-mãe e, considerando a respectiva localização, foi efetuada a análise de autocorrelação espacial. As três áreas apresentaram padrão agregado de plantas doentes, com predominância da agregação dentro das linhas. Na determinação dos tamanhos ideais das amostras, foram efetuadas amostragens-piloto em 30 plantios, sendo em cada plantio estabelecida uma sub-área de 2 ha, onde 50 plantas foram selecionadas por amostragem sistemática e avaliadas quanto à severidade da doença. A severidade da Sigatoka-amarela nas 30áreas de plantio submetidas às amostragens-piloto variou de 5,3% a 46,7%, não sendo constatadas correlações significativas entre os níveis de severidade e as idades dos plantios, bem como destas últimas com os tamanhos das amostras. Os tamanhos das amostras para quantificação da severidade da Sigatoka-amarela da bananeira foram determinados considerando o padrão agregado de plantas doentes. Os dados obtidos foram analisados pelo método que especifica o erro aceitável, com um padrão agregado de plantas doentes. Os tamanhos das amostras correlacionaram-se negativamente com os níveis de severidade da doença (r = -0,60). Considerando a média das áreas de plantio, em futuros levantamentos da severidade da Sigatoka-amarela, nos quais seja exigido excelente nível de precisão (erro = 5%), recomenda-se a amostragem de 66 plantas para cada 2 ha cultivados, com a avaliação das nove folhas mais jovens/planta. Os resultados obtidos nesse estudo servem como base para futuros levantamentos epidemiológicos da Sigatoka-amarela da bananeira, uma vez que os dados foram originados de campos sob diferentes condições e estimados considerando necessidades crescentes de precisão.

Made available in DSpace on 2016-06-02T19:06:02Z (GMT). No. of bitstreams: 1 4859.pdf: 3995047 bytes, checksum: d6fe8fefac57632109ba02702cb2a16a (MD5) Previous issue date: 2012-06-11 / Financiadora de Estudos e Projetos / The digital image processing has been used to solve a significant number of problems in the agricultural sector, especially with the evolution of remote sensing systems. This work presents a computational model based on digital image processing and remote sensing to infer about the risk in the agricultural environment. To validate the method, a experimental study was conducted about the risk of fungal disease in banana plantations. Temporal series of meteorological and monitoring data of the disease, organized in classes for the definition of probability distribution models, based on polynomial functions were used to validate results as well as satellite images integrated by fusion techniques, geometric corrections and re-sampling with interpolators based on kriging techniques. Fusion methods by IHS (Intensity, Hue, and Saturation) and PCA (Principal Component Analysis) and the Gaussian models, exponential and cylindrical for the ordinary kriging were tested. The IHS fusion technique demonstrated to be more interesting in relation the PCA technique, with correlation coefficients between bands 2, 3 and 4 originals and hybrids, of 0.2318, 0.0304 and 0.1800, respectively. The method of ordinary kriging for re-sampling of the images showed better results when adjusted by the Gaussian model. The proposed method is feasible to the development of risk maps of disease occurrence, since confer spatial and temporal variability in relation to the model existing on the literature for the region. / O processamento de imagens digitais tem auxiliado na solução de um expressivo número de problemas do setor agrícola, sobretudo com a evolução dos sistemas de sensoriamento remoto. Este trabalho tem por objetivo apresentar um modelo computacional baseado em processamento digital de imagens e sensoriamento remoto para inferência de risco em ambiente agrícola. Para validação do método, foi realizado estudo experimental sobre o risco de ocorrência da doença fúngica em bananais. Foram utilizadas séries temporais de dados meteorológicos e de monitoramento da doença, organizados em classes para a definição de modelos de distribuição de probabilidades, baseados em funções polinomiais, bem como imagens de satélites, organizadas com técnicas de fusão, correções geométricas e re-amostragem, com interpoladores baseados em técnicas de krigagem. Foram testados os métodos de fusão por IHS (Intensity, Hue, Saturation) e PCA (Principal Component Analysis), bem como os modelos gaussiano, exponencial e cilíndrico para a krigagem ordinária. A técnica de fusão por IHS demonstrou-se mais interessante em relação à técnica por PCA, apresentando coeficientes de correlação entre as bandas 2, 3 e 4 originais e híbridas, de 0,2318, 0,0304 e 0,1800, respectivamente. O método de krigagem ordinária para reamostragem das imagens apresentou melhores resultados quando ajustado pelo modelo gaussiano. A metodologia proposta se apresentou viável e adequada para a elaboração de mapas de risco de ocorrência da doença, uma vez que conferece variabilidade espacial e temporal ao modelo já existente na literatura para aquela região.

Orientador: João Antonio da Costa Andrade / Resumo: O milho é uma das culturas mais extensamente cultivadas em todo mundo. A incidência e a severidade de doenças têm aumentado significativamente nos últimos anos acarretando perdas no rendimento e afetando a qualidade dos grãos. Muitos trabalhos têm sido desenvolvidos na tentativa de identificar híbridos resistentes às principais doenças que acometem a cultura do milho, mas poucos são os relatos de estudos com linhagens. Dessa forma, o objetivo deste estudo foi: 1) identificar linhagens resistentes e susceptíveis com base na área abaixo da curva de progresso de doenças (AACPD) para os sintomas de ferrugem tropical (TR), ferrugem polissora (SR), cercosporiose (GLS), helmintosporiose (NLB), mancha marrom (PBS) e mancha branca (PLS); 2) identificar linhagens resistentes e suscetíveis com base nos parâmetros de adaptabilidade e estabilidade fenotípica para os sintomas de cercosporiose, helmintosporiose, mancha marrom e mancha branca; 3) identificar as melhores datas de semeadura, com a maior ocorrência das doenças, para fins de avaliação de linhagens e outros genótipos para resistência. Cinquenta linhagens, derivadas de populações com grãos flint e dent, foram avaliadas em blocos casualizados com três repetições, aos 45, 60, 75 e 90 dias após a semeadura em duas épocas, para medição da AACPD. Para a análise de adaptabilidade e estabilidade, 41 linhagens foram avaliadas em blocos casualizados com três repetições, 30 dias após o florescimento feminino, em onze épocas de semeadura, us... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor

La banane plantain (Musa paradisiaca) est l’aliment de base de millions de personnes dans le monde et sa culture génère des revenus permanents pour un grand nombre d’agriculteurs, dans des plantations de taille petite ou moyenne. Comme dans d’autres régions tropicales, la culture du plantain en Guadeloupe et en Haïti est soumise à de fortes contraintes parasitaires aussi bien telluriques (i.e. du sol : nématodes phytoparasites et charançon du bananier) qu’aériennes (Cercosporiose noire notamment). Les moyens de lutte conventionnels reposent sur l’usage de produits de synthèse dont les effets néfastes sur l’environnement (sols, eaux, animaux) comme sur la santé humaine ne sont plus supportables. Il est donc urgent de réfléchir à des solutions agro-écologiques permettant de rétablir les équilibres biologiques, de maintenir une bonne qualité des sols et une production optimale dans les systèmes de culture plantains. C’est le but de ce travail de thèse qui couple la réalisation d’un diagnostic agro-écologique dans des parcelles paysannes, et le test d’alternatives agro-écologiques en milieu semi-contrôlé. Pour ce faire, une typologie des systèmes de culture plantains a été réalisée à l’issue d’une enquête agro-environnementale dans les deux zones d’étude. Elle a permis de sélectionner 23 parcelles en Guadeloupe et 12 en Haïti dans lesquelles un diagnostic agro-écologique a été conduit. Sur la base de ce diagnostic et de la recherche d’alternatives agro-écologiques à l’usage des produits chimiques, une expérimentation au champ a été mise en place en Guadeloupe en station de recherche, permettant le test de trois pratiques culturales innovantes pour le plantain (seules et combinées), à savoir : i) l’introduction d’une plante de service Paspalum notatum pour la gestion des adventices et la réduction de l’utilisation d’herbicides ; ii) l’apport de vermicompost pour le contrôle des nématodes phytoparasites inféodés au bananier plantain et la fertilisation de celui-ci ; iii) l’utilisation de plants sains PIF (Plants Issus de Fragments de tiges) indemnes de nématodes et de larves de charançon du bananier. La typologie des systèmes de culture plantains révèle que : i) en Guadeloupe les précédents : jachère, ananas et banane plantain sont prédominants avec un niveau d’intensification faible (apports d’intrants chimiques faibles et peu fréquents) ou élevé (apports d’intrants chimiques élevés et plus fréquents) ; ii) en Haïti, les précédents : jachère, banane plantain et manioc prédominent avec un niveau d’intensification faible ou nul (apport d’intrants inexistant). Les résultats du diagnostic agro-écologique montrent que, i) lorsque le niveau d’intensification est faible, les bananeraies plantains pérennes et le précédent-ananas permettent de maintenir une bonne qualité du sol et une bonne régulation des parasites telluriques ; ii) lorsque le niveau d’intensification est fort, les populations d’ingénieurs du sol diminuent drastiquement, alors que le cortège parasitaire tellurique augmente sans que cela n’affecte l’obtention de bons niveaux de rendement instantannés (parcelles précédées d’ananas ou d’une jachère principalement) ; iii) en absence totale de fertilisation, il résulte une diminution de l’activité biologique du sol mais aussi du rendement du plantain, exacerbé par le choix des précédents-culturaux (manioc ou banane plantain) en lien avec les contraintes pédoclimatiques et la maladie des raies noires (Cercosporiose noire) causée par Mycosphaerella fijiensis, notamment au sein des parcelles Haïtiennes ; iv) la succession plantain/plantain est la plus pénalisante vis-à-vis de la culture du plantain, car quelque soit le niveau d’intensification, le rendement reste relativement faible, en lien avec une dégradation de l’état sanitaire, comparativement aux autres précédents. / Plantain (Musa paradisiaca) is the staple food of millions of people worldwide and its cropping generates ongoing revenues for many farmers who are planting small or medium size areas. As in other tropical regions, plantain cultivation in Guadeloupe and Haiti is under heavy parasitic constraints terrestrial (plant-parasitic nematodes and banana weevil) as well as aerial (black Sigatoka in particular). Conventional means of control based on the use of synthetic products which adverse effects on the environment (soil, water, animals ...) as on human health are not bearable any more. It is thus urgent to think about agroecological solutions allowing to restore the biological balances, to maintain good soil quality and optimal plantain cropping systems.This is the ultimate goal of this thesis which couples the realization of an agroecological diagnosis in peasants’ plots, and the test of agroecological alternatives in semi-controlled conditions. To do this, a typology of plantain cropping systems was carried out from an agrienvironmental survey in the two study areas. It allowed to select 23 plots in Guadeloupe and 12 in Haiti in which an agroecological diagnosis was implemented. Based on this diagnosis and the research of agroecological alternatives to the use of chemicals, a field experiment was set up in Guadeloupe, in an experimental station allowing the test of three innovative practices for plantain cultivation (alone and combined), namely : i) the introduction of a cover-crop Paspalum notatum for weed control while reducing the use of herbicides ; ii) the input of worms’ compost to control plant-parasitic nematodes specific to plantain and to fertilize it ; iii) the use of healthy “PIF” plants (plants issued from stem fragments) free from telluric pests (nematodes and weevil’s larvaes).The typology of plantains cropping systems shows: i) in Guadeloupe the previous crops are: fallow pineapple and plantain predominate with a low level of intensification (low and infrequent chemical inputs) or high (high and frequent chemical inputs); ii) in Haiti, the previous crops are: fallow, plantain and cassava predominate with a low level or no intensification at all (no inputs). The results of the agroecological analysis show that : i) when the level of intensification is low, perennial plantain and pineapple as previous crops help maintaining a good soil quality and a good regulation of the telluric pests ; ii) when the level of intensification is strong, the soil engineers drastically reduce, while the density of telluric parasites increases without affecting good levels of instantaneous yields (plots where the previous crop is pineapple or mostly fallow) ; iii) when the fertilization is totally missing, it decreases the biological activity of the soil furthermore the plantain yields, exacerbated by the choice of the previous crop (cassava or plantain), in connection with soils and climate constraints and the black Sigatoka caused by Mycosphaerella fijiensis, especially in the Haitian plots ; iv) the crop succession plantain/plantain is the most critical regarding the plantain’s cropping, because whatever the level of intensification, the yields remain relatively low in connection with a degradation of the health state, compared to other previous crops.The driving of an experiment in a research station shows that on the scale of one year, the three tested innovative practices allow maintaining a good soils quality. Healthy plants "PIF" have a better health state (absence of plant parasitic nematodes in the roots) which helps a significant increase of the yields. Cover-crop P. notatum helps the weeds and the soil pests control and favors the improvement of soil biological activity and plantain yields. Worms’ compost contributes to the maintenance of a better soils quality while allowing the regulation of the populations of plant-parasitic nematodes of the plantains.