Resumo O objetivo deste trabalho foi definir blocos experimentais para experimentos com cana-de-açúcar, com uso de técnicas de geoestatística, análise de componentes principais e técnicas de agrupamento aplicadas às propriedades do solo. Para isso, foram utilizados dados de propriedades químicas do solo de um experimento com cana-de-açúcar. As técnicas de geoestatística foram aplicadas para identificar a variabilidade espacial dessas propriedades e estimar os valores para locais não amostrados por meio de krigagem. A análise de componentes principais foi aplicada para redução dimensional, e, com as novas variáveis obtidas, realizou-se a análise de agrupamento pelo método k-means, para determinar os blocos experimentais com duas a cinco repetições. Das 12 variáveis analisadas, 10 apresentaram dependência espacial. A análise de componentes principais permitiu reduzir a dimensionalidade dos dados para duas variáveis, que explicaram 82,27% da variância total. Os blocos obtidos apresentaram formas poligonais irregulares, com diferentes formatos e tamanhos, e alguns mostraram descontinuidade. A metodologia proposta tem potencial para identificar áreas mais homogêneas em termos de propriedades químicas do solo, para alocar blocos experimentais de cana-de-açúcar. canadeaçúcar, canadeaçúcar cana açúcar, açúcar cana-de-açúcar isso canadeaçúcar. açúcar. krigagem dimensional obtidas realizouse realizou se kmeans, kmeans k means, means k-means repetições 1 analisadas 8227 82 27 82,27 total irregulares tamanhos descontinuidade 822 8 2 82,2 82,

Abstract The objective of this work was to define experimental blocks for sugarcane experiments using geostatistical techniques, principal component analysis, and clustering techniques applied to soil properties. For this, data of soil chemical properties from a sugarcane experiment were used. Geostatistical techniques were applied to identify the spatial variability of these properties and to estimate the values for non-sampled locations through kriging. The principal components analysis was used for dimensional reduction, and, with the new variables obtained, the cluster analysis was performed using the k-means method to determine the experimental blocks with two to five replicates. Of the 12 analyzed variables, 10 showed spatial dependence. The principal component analysis allowed reducing the dimensionality of the data to two variables, which explained 82.27% of total variance. The obtained blocks presented irregular polygonal shapes, with different formats and sizes, and some of them showed discontinuities. The proposed methodology has the potential to identify more uniform areas in terms of soil chemical properties to allocate experimental blocks for sugarcane. nonsampled non sampled kriging reduction kmeans k means replicates 1 dependence 8227 82 27 82.27 variance shapes sizes discontinuities 822 8 2 82.2 82.

Abstract This study aimed to assess the effect of red, pearl and aluminet shading nets on the photosynthetic efficiency of Capsicum chinense and C. frutescens. The experiment was carried in entirely randomized block design. C. chinense and C. frutescens were cultivated under red, pearl and aluminet with 35% shading and direct sunlight as control for 150 days. Leaf area, area duration and area ratio; specific leaf area and harvest index were higher in C. frutescens. Net CO2 assimilation and stomatal conductance were higher in C. frutescens cultivated under red net and C. chinense cultivated under aluminet net. Transpiration was higher in C. chinense cultivated under the aluminet net. Water use efficiency and relative chlorophyll content and intrinsic water use efficiency were higher in C. frutescens cultivated under the red net. Initial and maximum chlorophyll were also higher in C. frutescens cultivated under red net. Maximum quantum efficiency and chlorophyll fluorescence rate decrease were lower in C. chinense cultivated under red net. Linear flux electron was lower in C. chinense cultivated in the control. PSII open centers fraction was higher in control, pearl net for C. frutescens. Non-photochemical extinction coefficient and dissipation of absorbed light were higher in C. chinense cultivated under red net. C. frutescens cultivated under red net had higher photosynthetic efficiency, considering water use efficiency in dry matter conversion and several mechanisms to maximize the photosynthetic process under shaded conditions. C design 35 15 days ratio CO Nonphotochemical Non photochemical conditions 3 1

RESUMO: A divisão de subconjuntos de treinamento e teste é parte fundamental da otimização de modelos. O objetivo deste trabalho foi avaliar a influência da escolha do subconjunto de treinamento na construção dos modelos, bem como sua validação. Os métodos Kennard-Stone (KS) e a amostragem aleatória (AA) foram avaliados em dados de espectroscopia no infravermelho próximo (NIR) e em dados de marcadores SNPs (Single Nucleotide Polymorphisms). Vale destacar, que em dados SNPs, não há conhecimento de relatos na literatura a respeito da utilização do método KS. Para a construção e validação dos modelos, o método de estimação dos mínimos quadrados parciais (PLS) e Lasso bayesiano (BLASSO) mostraram-se mais eficientes para os dados NIR e para os dados SNPs, respetivamente. A avaliação da capacidade preditiva dos modelos obtidos após a partição dos dados ocorreu por meio da correlação entre os valores preditos e os valores reais, e da raiz quadrada do erro quadrático médio de predição. Para ambos os conjuntos de dados, os resultados indicam que os métodos KS e AA diferem estatisticamente entre si pelo teste F (valor P < 0.01), com o KS mais eficiente do que o AA em praticamente todas as repetições. Além disso, o método KS possui a vantagem de ser fácil e rápido de ser aplicado e também de selecionar sempre as mesmas amostras, o que proporciona grandes benefícios em futuras análises.

ABSTRACT: Splitting the whole dataset into training and testing subsets is a crucial part of optimizing models. This study evaluated the influence of the choice of the training subset in the construction of predictive models, as well as on their validation. For this purpose we assessed the Kennard-Stone (KS) and the Random Sampling (RS) methods in near-infrared spectroscopy data (NIR) and marker data SNPs (Single Nucleotide Polymorphisms). It is worth noting that in SNPs data, there is no knowledge of reports in the literature regarding the use of the KS method. For the construction and validation of the models, the partial least squares (PLS) estimation method and the Bayesian Lasso (BLASSO) proved to be more efficient for NIR data and for marker data SNPs, respectively. The evaluation of the predictive capacity of the models obtained after the data partition occurred through the correlation between the predicted and the observed values, and the corresponding square root of the mean squared error of prediction. For both datasets, results indicated that the results from KS and RS methods differ statistically from each other by the F test (P-value < 0.01). The KS method showed to be more efficient than RS in practically all repetitions. Also, KS method has the advantage of being easy and fast to be applied and also to select the same samples, which provides excellent benefits in the following analyses.

Abstract This study aimed to investigate genetic parameters and selection strategies for sugarcane resistance to borer and determine the effects of selection for borer resistance on yield traits. The infestation index by borer (IIB) and yield traits were assessed over 3 years in 35 sugarcane clones planted in four locations in southeastern Brazil. There was no genotype × environment (year or location) interaction for borer resistance. Consequently, selection may be performed at only one location and year or by the mean of clones in several locations and/or years. The borer population affected the genotypic variance for borer resistance but when genotypic variance was significant, heritability was elevated and suitable genetic gain was achieved. Selection for borer resistance reduced the genetic gains for tons of cane per hectare, theoretical recoverable sugar, and tons of sugar per hectare; however, satisfactory selection gains for both borer resistance and yield were achieved using selection index.

ABSTRACT The sugarcane spittlebug, Mahanarva fimbriolata (Walker) (Hemiptera: Cercopidae) is an economically important pest of sugarcane in Brazil. The purpose of this study was to develop and validate a greenhouse methodology to screen large sugarcane populations for resistance to spittlebug Mahanarva fimbriolata. A resistant and a susceptible genotype (H. Kawandang and SP81-3250) were first used to determine adequate days after infestation and levels of infestation (number of nymphs per plant) for comparing the resistance of genotypes. Then, 74 sugarcane genotypes including three susceptible and three resistant controls were screened for resistance. The screening method consists in infesting single-tiller plants supported in a small plant growth unit and assessing the damage by using a 1-5 visual damage score. Our data suggest screening with four to six nymphs per plant and the damage score assessment at least 21 days after infestation. The screening technique was proved reliable as susceptible and resistant controls were placed in their respective resistance category. Three genotypes were classified as resistant while the majority of genotypes were classified as susceptible to spittlebug, indicating the need of breeding for resistance.

Abstract The objective of this study was to select a population of sugarcane clones to develop energy cane cultivars. The genotypic means of 124 clones were estimated for the yield-related traits: mean number of stalks per plant (NS), mean stalk weight (MSW), fiber content (FIB), lignin content (LIG), and sucrose content in cane juice (SPC). From these data, the traits tons of cane per hectare (TCH), tons of fiber per hectare (TFH) and tons of lignin per hectare (TLH) were estimated. The clone effects were statistically significant (p <0.01) for all traits except TLH and LIG. Clone UFV09251 had the highest genotypic values for TFH and TCH. Clone UFV09270 had a high fiber content (15.04%) and clone UFV09225 a high sucrose content (14.05%). The correlation between TFH and TCH was high (0.89). Clone UFV09251 is highly promising as energy cane cultivar, for producing most TCH and FIB.

RESUMO A broca da cana-de-açúcar Diatraea saccharalis Fabr. (Lepidoptera: Crambidae) é uma das mais importantes pragas da cana-de-açúcar nas Américas. Este estudo foi realizado para investigar múltiplos traços de resistência em oito genótipos de cana-de-açúcar contra D. saccharalis, incluindo cinco cultivares comerciais e três germoplasmas exóticos com potencial para servir de fontes de resistência contra esta praga. A preferência de oviposição foi avaliada por meio de testes de escolha livre e de não-escolha. O desempenho de larvas de estágio inicial que se alimentam das folhas e de larvas de estádio tardio alimentando-se com os colmos também foi avaliado. Houve diferenças entre os genótipos para o número de ovos e posturas no teste de livre escolha, enquanto não foram observadas diferenças no teste de não-escolha. Houve também diferenças na sobrevivência de larvas de estádios iniciais que se alimentam das folhas, índice de lesões e dano no colmo. O genótipo IM76-228 foi o menos preferido para a oviposição e parece que a largura da folha teve alguma influência sobre a preferência dos adultos em vez da intensidade de verde das folhas. IM76-228 e RB867515 provocaram uma mortalidade mais elevada de larvas de estádio inicial alimentando-se dos colmos, enquanto que IM76-228 e RB985523 apresentaram menor dano nas folhas e nos colmos. O genótipo IM76-228 foi o mais resistente a D. saccharalis e pode servir como fonte de genes para resistência em programas de melhoramento de cana-de-açúcar.

ABSTRACT The sugarcane borer Diatraea saccharalis Fabr. (Lepidoptera: Crambidae) is one of the most important pests of sugarcane in the Americas. The purpose of this study was to assess multiple resistance traits in eight sugarcane genotypes against D. saccharalis, including five commercial cultivars and three exotic germplasm with potential to serve as resistance sources against this pest. The oviposition preference was assessed by using both free-choice and non-choice tests. The performance of both early stage larvae feeding on the leaves and late stages larvae feeding within the stalks were also assessed. There were differences among genotypes for number of both eggs and egg cluster in the free-choice test while no differences in non-choice test were observed. There were also differences in survival of early stage larvae feeding on the leaves, foliar injury rating and stalk damage. The genotype IM76-228 was the least preferred for oviposition and it seems that leaf width had some influence on adults’ preference rather than greening of the leaves. IM76-228 and RB867515 causes higher mortality of early stage larvae feeding in the stalks while IM76-228 and RB985523 had lower damage in both leaves and stalks. The genotype IM76-228 was the most resistant to D. saccharalis and could serve as genes sources for resistance in sugarcane breeding programs.

Abstract The objective of this study was to identify superior parents for crosses and the best families to breed new clones of energy cane. The best-performing parents were RB867515, RB93509, KRAKATAU, IM76-228, IM76-229, and US85-1008. The heritability and accuracy values for the traits tons of cane per hectare, fiber, and sucrose content indicate high correlation between the predicted genotypic means and observed values, enabling efficient selection of the best cane energy families. The extensive genetic variation detected and the presence of promising seedlings in the selected segregating population indicate the possibility of using new crosses in some clones and even in future commercial plantings. It was possible to identify the best parents involved in the evaluated crosses, and to select, in the best families, seedlings with a high fiber and good sucrose content, seedlings with high fiber and low sugar content, as well as seedlings with the same fiber and sucrose contents as the current cultivars.

RESUMO. O objetivo desse trabalho foi avaliar a importância dos caracteres agronômicos durante o processo de seleção individual em cana-de-açúcar, além de avaliar o potencial do uso da regressão logística e da árvore de decisão na identificação dos melhores genótipos. Foram avaliados 7.719 seedlings na primeira fase de teste (T1) e 659 clones na segunda fase de teste (T2). Os caracteres número de colmos, saliência da gema e comprimento do entre-nó foram os de maior importância no processo de seleção massal realizado no T1. Vigor da planta, diâmetro de colmo e altura de colmo foram os mais importantes na seleção do T2. Pelo método massal foram selecionados 174 indivíduos no T1 e 113 no T2, enquanto a regressão logística selecionou 153 indivíduos no T1 e 79 no T2. A taxa de erro aparente dos modelos logísticos ajustados para as seleções no T1 e T2 foram de 0,8 e 5,10%, respectivamente. Com o uso da árvore de decisão foram selecionados 67 clones dentre os mais produtivos na fase T2. A elaboração de árvores de decisão apresenta grande aplicabilidade na identificação de clones potenciais nas fases iniciais dos programas de melhoramento genético da cana-de-açúcar.

ABSTRACT. The aim of this study was to evaluate the importance of agronomic traits during the selection of sugarcane (Saccharum spp.), as well as to evaluate the potential for using logistic regression and decision trees to identify the best genotypes. A total of 7,719 seedlings of 128 half-sib families were evaluated during the first test phase (T1), and 659 clones were selected for the second (T2). Logistic regression was applied to both populations. The number of stalks, bud prominence and length of the internode were the most important selection traits in the T1 population. The plant vigor, stalk diameter and stalk height were the most important selection traits in the T2 population. There were 174 individuals selected when using the mass selection method in T1 and 113 individuals in T2, whereas a logistic regression selected 153 individuals in T1 and 79 in T2. The apparent error rates of the logistic models fitted to the selections in T1 and T2 were 0.8 and 5.10%, respectively. By using a decision tree, 67 clones were selected among the most productive ones in phase T2. Therefore, the formulation of decision trees is highly applicable to identifying potential clones during the initial phases of breeding programs.

Resumo: O objetivo deste trabalho foi avaliar a eficiência da análise estatística espacial, em comparação a análises usuais em blocos ao acaso e em látice, na classificação de famílias de feijoeiro ( Phaseolus vulgaris), sob diferentes cenários de dependência espacial e de precisão experimental. Foram considerados 12 cenários, formados por quatro classes de dependência espacial entre erros (nula, baixa, média e alta) e três classes de precisão experimental (moderada, alta e muito alta). Para as três classes de precisão experimental, foram definidos os valores de acurácia seletiva de: 0,60, 0,80 e 0,95, respectivamente. Já as quatro classes de dependência espacial entre erros assumiram, respectivamente, os seguintes valores de alcance do modelo geoestatístico exponencial: 0, 10, 20 e 40 m. Para experimentos com precisão experimental muito alta ou ausência de dependência espacial, a eficiência da análise espacial para classificação de famílias do feijoeiro é semelhante à eficiência das análises usuais. Para experimentos com alta e, principalmente, com moderada precisão experimental, a análise espacial é mais eficiente que as análises usuais para classificação de famílias de feijoeiro, quando os erros apresentam dependência espacial.

Abstract: The objective of this work was to evaluate the efficiency of spatial statistical analysis, compared to the usual analyzes in randomized blocks and in lattice, for the classification of common bean ( Phaseolus vulgaris) families under different scenarios of spatial dependence and experimental precision. Twelve scenarios were considered, formed by four classes of spatial dependence between errors (null, low, medium, and high) and three classes of experimental precision (medium, high, and very high). For the three classes of experimental precision, the following values of selective accuracy were defined: 0.60, 0.80, and 0.95, respectively. The four classes of spatial dependence between errors assumed, respectively, the following range values for the exponential geostatistical model: 0, 10, 20, and 40 m. In experiments with very high experimental precision or with the absence of spatial dependence, the efficiency of the spatial analysis for the classification of common bean families is similar to that of the usual analyses. In experiments with high and, especially, with medium experimental precision, the spatial analysis is more efficient than the usual analyses for the classification of common bean families, when the errors show spatial dependence.

O objetivo desse trabalho foi avaliar o uso de Redes Neurais Artificiais (RNA) na seleção dentro de famílias de cana-de-açúcar. O melhor modelo de RNA não apresentou erro, sendo capaz de classificar corretamente todos os genótipos, ou seja, a rede tomou a mesma decisão de seleção realizada pelo melhorista durante a aplicação do BLUP individual simulado (BLUPIS), demonstrando a capacidade de aprendizado da RNA a partir das entradas e saídas informadas nas fases de treinamento e validação. Tendo em vista que a seleção via RNA facilita a identificação dos melhores indivíduos e visando desenvolver uma nova estratégia de seleção dentro das melhores famílias, de forma a garantir que os melhores genótipos da população sejam avaliados nas próximas fases do programa de melhoramento, recomendamos, ranquear as famílias via BLUP, selecionar as melhores e realizar a seleção individual via RNA, a partir das informações coletadas em nível individual nas melhores famílias.

The objective of this study was to evaluate Artificial Neural Networks (ANN) applied in an selection process within sugarcane families. The best ANN model produced no mistake, but was able to classify all genotypes correctly, i.e., the network made the same selective choice as the breeder during the simulation individual best linear unbiased predictor (BLUPIS), demonstrating the ability of the ANN to learn from the inputs and outputs provided in the training and validation phases. Since the ANN-based selection facilitates the identification of the best plants and the development of a new selection strategy in the best families, to ensure that the best genotypes of the population are evaluated in the following stages of the breeding program, we recommend to rank families by BLUP, followed by selection of the best families and finally, select the seedlings by ANN, from information at the individual level in the best families.

O objetivo dessa pesquisa foi avaliar a diversidade genética e a relação de parentesco entre 50 genótipos de canade- açúcar, além de identificar genitores potenciais para a obtenção de cultivares cana energia. A análise de diversidade foi realizada a partir da avaliação de caracteres morfológicos e agronômicos. O coeficiente de parentesco foi obtido a partir das informações de pedigree. De acordo com as análises realizadas os genótipos foram separados em dois grupos. O grupo G1 foi formado por 13 genótipos descendentes das espécies Saccharum spontaneum e Saccharum robustum. Os outros 37 genótipos descendem de retrocruzamentos com Saccharum officinarum e foram alocados no grupo G2. Os genótipos apresentaram baixos valores de similaridade genética e de coeficiente de parentesco, indicando ampla variabilidade genética na população. A condução de cruzamentos envolvendo genótipos do grupo G1, principalmente aqueles com teor de fibra acima de 17% (IJ76-293, 57NG12, IN84-82, IN84-88, IM76-228 e UM69/001), com os genótipos do grupo G2, que apresentam elevada produtividade de colmos (RB92579, RB83102, RB047232, RB867515, RB971723, RB937570, RB011941, RB75126, MEX68- 200, Co62175 e CP691052), devem ser explorados com a finalidade de desenvolver cultivares cana energia. As análises de diversidade e do coeficiente de parentesco permitiram a identificação de dois grupos heteróticos. Além disso, foi possível identificar nos dois grupos genitores potenciais para a obtenção de cultivares cana energia. A utilização das distâncias genéticas com base em dados morfo-agronômicos e do pedigree devem ser usadas de forma complementar, visando agregar maior conhecimento na escolha dos melhores genitores.

The aim of this study was to evaluate the genetic diversity and coefficient of kinship in 50 sugarcane genotypes, in addition to identifying potential parents for obtaining cultivars of energy cane. Diversity analysis was carried out based on the evaluation of morphological and agronomical characteristics. The coefficient of kinship was obtained from information on pedigree. According to analyses carried out, genotypes were separated into two groups. Group G1 consisted of 13 genotypes from the species Saccharum spontaneum and Saccharum robustum. The other 37 genotypes were from back-crosses with Saccharum officinarum, and were allocated to group G2. The genotypes displayed low values for genetic similarity and coefficient of kinship, indicating broad genetic variability in the population. Carrying out crosses involving genotypes from group G1, especially those with a fibre content of over 17% (IJ76-293, 57NG12, IN84-82, IN84-88, IM76-228 and UM69/001), with genotypes from group G2 which have high stalk yield (RB92579, RB83102, RB047232, RB867515, RB971723, RB937570, RB011941, RB75126, MEX68-200, Co62175 and CP691052), should be explored, with the aim of developing energy cane cultivars. Analyses of diversity and of the coefficient of kinship made it possible to identify two heterotic groups. Moreover, it was possible to identify two potential parent groups for obtaining energy cane cultivars. Genetic distances which are based on both morpho-agronomical data and on pedigree, should be used in a complementary way, with a view to having more information when choosing the best parents.

A eficiência do método BLUP individual simulado (BLUPIS) na seleção de genótipos dentro de famílias de irmãos germanos de cana-de-açúcar no estágio de cana soca, comparada com a seleção utilizando o método BLUP individual, foi avaliada. O número ótimo de genótipos a serem selecionados na melhor família foi estabelecido para as características massa media de colmos, teor de sólidos solúveis totais (BRIX), tonelada de colmos por hectare e tonelada de BRIX por hectare. Foram avaliadas dezessete famílias de irmãos germanos no Centro de Experimentação em Cana-de-açúcar, localizado em Oratórios, MG, Brasil. A metodologia de modelo misto foi utilizada para predizer os efeitos genotípicos de cada família e os valores genotípicos de cada individuo dentro de família. O método BLUPIS mostrou-se eficiente para a seleção individual. O número ótimo de genótipos a serem selecionados na melhor família para obter a maior eficiência do método BLUPIS foi 100, na maioria dos casos.

The aim of this study was to assess the efficiency of the simulated individual BLUP (BLUPIS) method in selecting genotypes within full-sib families of sugarcane in ratoon stage, through comparison with selection using the individual BLUP method. The optimal number of genotypes to be selected in the best families were established for mean stems mass, total soluble solids assay (BRIX), ton of stalks per hectare, and BRIX tons per hectare traits. Seventeen full-sib families were assessed in the Centre for Experimentation in Sugarcane, located in Oratórios, MG, Brazil. Mixed model methodology was used to predict the genotypic effects of each family and the genotypic values of each individual within family. BLUPIS method is efficient for individual selection. The optimal number of genotypes to be selected in the best family for obtaining higher efficiency of the BLUPIS method is 100 in the majority of cases.

Para a avaliação de experimentos com cana-de-açúcar, tem-se recomendado a colheita total de parcelas, algo que, devido aos seus custos, tem se mostrado inexequível. Alternativamente, pode-se estimar a produção com base na tomada de amostras que representem as parcelas. O objetivo deste trabalho foi estudar a suficiência amostral em cana-de-açúcar para a estimação da massa dos colmos de parcelas experimentais. Reamostragens com reposição subsidiaram o uso dos modelos da curva potencial, linear e quadrático segmentados com resposta platô e do modelo do coeficiente de variação relativo. Também foram realizados estudos simulados de análise de variância. Os resultados indicaram que amostras de 20 colmos coletados aleatoriamente na parcela experimental permitem estimar adequadamente a variável massa do colmo. Com este valor, serão obtidas avaliações confiáveis da produtividade em colmos

During the evaluation of the experimental phase, which corresponds to competition experiments and indication of cultivars, is recommended total harvest of plots, which, owing to their costs, can become impracticable. Optionally, one can estimate the production based on taking samples representing plots. The aim of this study was to research the sample sufficiency in sugarcane for estimating the mass of stalks of experimental plots. One used the maximum curvature of the power model, the linear and quadratic segmented models with response plateau and the relative CV model. Studies based on simulated analysis of variance were done also. The application of these models and the ANOVA was supported by bootstrap resamplings. The results indicated that a sample made by 20 stalks harvested in plot can be used for estimating properly the variable mass of the stalk. This value may be used for estimating the sugarcane yield per hectare