ABSTRACT Traditional germination tests which assess seed quality are costly and time-consuming, mainly when performed on a large scale. In this study, we assessed the efficiency of X-ray imaging analyses in predicting the physiological quality of tomato seeds. A convolutional neural network (CNN) called mask region convolutional neural network (MaskRCNN) was also tested for its precision in adequately classifying tomato seeds into four seed quality categories. For this purpose, X-ray images were taken of seeds of 49 tomato genotypes (46 Solanum pennellii introgression lines) from two different growing seasons. Four replicates of 25 seeds for each genotype were analyzed. These seeds were further assessed for germination and seedling vigor-related traits in two independent trials. Correlation analysis revealed significant linear association between germination and image-based variables. Most genotypes differed in terms of germination and seed development performance considering the two independent trials, except LA 4046, LA 4043, and LA4047, which showed similar behavior. Our findings point out that seeds with low opacity and percentage of damaged seed tissue and high values for living tissue opacity have greater physiological quality. In short, our work confirms the reliability of X-ray imaging and deep learning methodologies in predicting the physiological quality of tomato seeds. timeconsuming, timeconsuming time consuming, consuming time-consuming scale study Xray X ray CNN (CNN MaskRCNN (MaskRCNN categories purpose 4 46 (4 lines seasons 2 analyzed vigorrelated vigor related trials imagebased image based variables 4046 4043 LA4047 behavior short ( 404 LA404 40 LA40 LA4

ABSTRACT. To be considered drought-tolerant, a tomato cultivar is required to present some level of tolerance at all developmental stages of plant growth. Since drought tolerance is a stage-specific phenomenon, genotype assessment must be performed separately at all developmental stages. In this study, we used a multi-trait index based on factor analysis and genotype-ideotype distance (FAI-BLUP index) to properly rank 49 tomato genotypes according to their tolerance to drought stress conditions at germination and seedling stages. Seeds of 47 introgression lines, which cultivar M82 is considered drought-sensitive, and the Solanum pennellii wild accession LA 716, which is considered drought-tolerant, were subjected to a control condition, where seeds were soaked in distilled water, and a drought condition, where seeds were soaked in a polyethylene glycol (PEG) solution (-0.3 MPa). Drought stress, induced by PEG, had a significant impact on all nine germination and growth performance-related traits; there was a reduction in shoot length (SL), total length (TL), initial germination percentage (IGP), final germination percentage (FGP), germination velocity index (GVI), and germination rate (GR). In contrast, the root-to-shoot ratio (R/S) and time to reach 50% germination (T50) increased under drought stress. Root length (RL) was less affected by drought, and in some genotypes, it was even increased. As expected, LA 716 ranked closest to the drought tolerance ideotype. IL 1-4-18, IL 2-3, IL 1-2, IL 9-2, and IL 10-1 were the most drought-tolerant at the germination stage. These results will serve as guidance for breeders who are aiming at developing drought-resistant tomato cultivars.

Resumo Os modelos de estimativa de área foliar descritos na literatura são inapropriados para folhas tipo lanceoladas, na fase de crescimento vegetativo dependente do broto de sucessão, até os 90 dias de idade. Desta forma, objetivou-se ajustar equações para estimativa de área foliar em bananeiras ‘Prata-Anã’ e ‘BRS Platina’ em plantas com folhas lanceoladas. Foram coletadas 212 folhas do tipo lanceoladas com até 10 cm de comprimento de limbo, de bananeiras ‘Prata-Anã’, e 164 de ‘BRS Platina’, com idade de até 90 dias. Foram mensuradas largura (L), comprimento (C), relação largura-comprimento (RLC) e área foliar (AFescâner), determinada em escâner. Através do procedimento backward no SAEG, ajustaram-se equações de regressão linear múltipla, em função das variáveis L, C e RLC e AFescâner. Para avaliar a qualidade da predição da área foliar calculada através das equações (AFpredita), ajustou-se uma correlação linear de Pearson entre a área AFpredita e AFescâner. Após os ajustes das equações de regressão linear múltipla, obtiveram-se modelos que demonstraram alto coeficiente de correlação com AFescâner, a 1% de significância. Os modelos ajustados para a determinação da área foliar de bananeiras jovens, com folhas lanceoladas, são: AFFL(Prata-Anã) = -,0133624 + 0,000489859**C - 0,00183182 **L e AFFL(Platina) 0,00237026 + 0,004781**L - 0,096802** RLC.

Abstract Models for estimating leaf area of bananas found in the literature are not suitable for lanceolate type leaves occurring at the vegetative stage of ratoon suckers dependent of mother plant. The objective was to determine equations for estimating the leaf area of ‘Prata-Anã’ and ‘BRS Platina’ banana plants with lanceolate type leaves. 212 and 164 lanceolate type leaves having 10 cm-wide lamina or less were collected from ‘Prata-Anã’ and ‘BRS Platina’ banana plants of 90 days of age or less, respectively. Width (W), length (L), width/length ratio (WLR), and scanner-measured leaf area (LAscanner) were determined. Using the backward elimination procedure, simple and multiple linear regression equations were fitted to the relationship between leaf dimensions (W, L and WLR) and LAscanner. To evaluate how precise the equations are in predicting leaf area (LApredicted), Pearson correlation coefficients were calculated between LA and LApredictedscanner. The models highly correlated with LAscanner at 1% of significance level. The models are and LALL() = Prata-Anã = - 0,0133624 + 0,000489859**L - 0,00183182 **W and LALL(Platina) 0,00237026 + 0,004781**W - 0,096802** WLR.

Uma forma para se obter diminuição significativa de tempo e recursos despendidos nos programas de melhoramento de Coffea arabica L. é a clonagem de híbridos F1 por meio de estacas caulinares. Alguns estudos, em diferentes instituições, foram realizados buscando-se definir um método eficiente para esse tipo de clonagem. Com o objetivo de verificar-se a presença de barreiras anatômicas ao enraizamento de estacas caulinares do cafeeiro e a origem das raízes adventícias, bem como compara-las às raízes provenientes de plantas obtidas por semeadura, foram realizadas análises anatômicas no Departamento de Biologia da Universidade Federal de Lavras (UFLA), Lavras, MG. Utilizaram-se estacas caulinares de cafeeiro dos cultivares Acaiá e Rubi e mudas obtidas por semeadura direta e por estaquia. Os cortes realizados nas estacas caulinares mostraram não existirem barreiras anatômicas ao enraizamento adventício. Nas estacas enraizadas, a origem do primórdio radicular foi próxima aos tecidos vasculares. Cortes histológicos nas raízes formadas nas estacas e nas raízes de mudas obtidas por semeadura confirmaram que elas apresentam as mesmas estruturas primárias.

F1 hybrid cloning using rooting of cuttings is a way to achieve less expensive and quicker Coffea arabica improvement. Cloning is still a problem in Coffea arabica improvement and has been studied by different institutions to develop an efficient methodology. This work aimed to search for the presence of anatomical barriers to rooting of coffee stem cuttings and the origin of adventitious roots, comparing them with the structure of seedling tap roots. Anatomical analyses of young coffee plants were carried out at the Biology Department of the Universidade Federal de Lavras (UFLA), Lavras, MG. Plants of the cultivars Acaiá and Ruby obtained by rooting of cuttings and seedlings were analyzed. No anatomical barriers to adventitious rooting were found in the rooted cuttings. In the cloning plants the rooting arose in close to the vascular tissues. The anatomical study of adventitious roots in the rooted cuttings and seedlings showed the same primary structures.

Os objetivos deste trabalho foram caracterizar agronomicamente cem clones de mandioca (Manihot esculenta Crantz) e calcular os coeficientes de trilha entre a produção de raízes tuberosas e cinco componentes da produção, de modo a auxiliar na seleção de clones superiores. O experimento foi conduzido em área experimental da Universidade Federal de Lavras, em 2005 e 2006. Cem clones de mandioca foram avaliados em delineamento látice quadrado 10x10. A unidade experimental foi constituída por quatro plantas espaçadas de 1,0x0,6 m. As análises estatísticas foram realizadas considerando-se as oito características individualmente, utilizando-se o teste de Scott-Knott para agrupamento das médias. Os clones 87 e 88 mostraram-se promissores tanto para serem utilizados em cruzamentos quanto para fixação como novas cultivares em virtude do excelente desempenho nas características comprimento, número e produção de raízes tuberosas por planta. A análise de trilha mostrou que o número de raízes por planta e o peso total da parte aérea podem ser utilizados como critérios na seleção indireta para produção de raízes tuberosas em mandioca.

The objectives of this work were to perform the agronomic characterization of 100 cassava plants (Manihot esculenta Crantz) and to obtain path coefficients analysis between the storage root production (basic variable) and five production components (explanative variable). The experiment was conducted at a research farm of Universidade Federal de Lavras, Minas Gerais State, Brazil, in square lattice (10x10), with two replications, using 100 clones and four plants per plot. Statistical analyses were accomplished considering the eight characteristics individually. The Scott-Knott's test was used for grouping the means. Clones 87 and 88 showed to be promising both to be used in crossings and to be fixed as new cultivate, due to their excellent performance in the characteristics: length, number and storage roots production per plant. Path coefficient analysis demonstrates that root number per plant and aerial part weight can be used as criteria of indirect selection for storage roots production.