Análises biométricas da produtividade, composição de açúcares solúveis e polissacarídeos não-amídicos (PNAs) na seleção de genótipos de soja

• Autor(a) principal: Bueno, Rafael Delmond
• Data de Publicação: 2009
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/2414
• topic: Soja; Açúcares solúveis; Polissacarídeos; Soybean; Soluble sugar; Polysaccharides; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::FITOTECNIA::MELHORAMENTO VEGETAL

Soybeans have high nutritional values and are used as a base for various products; therefore there is a growing demand for soybean genotypes with high productivity and specific quality characteristics. For this reason it is important to characterize soybean genotypes for biochemical components related to quality and also antinutritional factors which limit its use for human and monogastric animal consumption. The objective of this study had two general objectives: evaluate genotype x location interaction in terms of productivity, and concentrations of proteins, sucrose, raffinose and stachyose in soybeans, estimating correlations between analyzed factors in the present study and; select soybean genotypes which can be used in programs which intent to increase the nutritional quality of soybeans for humans and monogastric animals. A total of 18 soybean genotypes were evaluated, gathered from COOPADAP experimental sites in the municipalities of Mineiros-GO, Perdizes-MG, São Gotardo-MG and Presidente Olegário-MG during the 2005/2006 harvest. Experiments were performed in randomized blocks with three repetitions. Average yield (Kg ha-1) of the 18 genotypes at the four different locations were determined as well as concentrations of protein using near infrared spectroscopy (NIR) and sucrose, raffinose and stachyose by HPLC. Statistical analyses were performed with the help of a genetic and statistical computational program (GENES Program). During analysis of joint variance, significant effects were verified for location and genotype x location (p<0.01), for all five evaluated characteristics. Significant effects of the genotype were detected (p<0.01) for yield and protein and (p<0.05) for sucrose and raffinose. Adaptability and stability analyses were performed using the Lin and Binns (1988) method, modified by Carneiro (1998), and the Centroid method. Both models showed coherency and allowed for identification of the most stable genotypes and responses for the best location for each genotype as well as greatest yields and concentrations of protein, sucrose, raffinose and stachyose. The genotypes which stood out due to their wide adaptability according to the two utilized methodologies in the present work were: for productivity, the VENCEDORA and MSOY 8001 varieties; for protein concentration, the ELITE variety, for sucrose concentration, the CS 01 736 line, for raffinose concentration, the CS 02 988 line and for stachyose concentration, the MSOY 8001 variety. Yield and protein concentration presented negative correlations, which showed that selection of one determined characteristic provoked a decline in the other. Correlation between yield and sucrose concentration was insignificant. The same behavior was observed between protein concentration and sucrose concentration and between protein concentration and stachyose concentration. The fact that these characteristics are not correlated facilitates the selection of a genotype with high sucrose concentration and high yields, or genotypes with high concentrations of sucrose and protein. The correlation between yield and raffinose concentration and between yield and stachyose concentration were both negative and significant. This negative correlation and significance facilitated the selection of genotypes with high yields and low concentrations of raffinose and stachyose. Sucrose concentration presented high and positive correlation coefficients with the concentrations of raffinose and stachyose, indicating the possibility of making a direct selection of individuals with high sucrose concentrations and low raffinose and stachyose concentrations. The second part of this study was performed using six soybean genotypes from Mineiros-GO. It was demonstrated that the methodology adapted to quantify thermostable antinutritional factors, such as soybean non-starch polysaccharides, using HPLC, showed excellent precision. No significant correlation was encountered between yield and NSP, vegetable oil and NSP, and protein and NSP between the six analyzed genotypes. The fact that these characteristics are not correlated facilitates genotype selection with the best nutritional quality, since it allows for diminishing the total NSP content, insoluble NSP and principally, soluble NSP, without affecting the yield and concentrations of oil and protein. The yield and oil concentration variables showed to be high and have positive correlation coefficients with the total carbohydrate concentrations and soluble sugars, indicating difficulty to make a direct selection of varieties with high yields or high oil concentrations and at the same time, low total carbohydrate and soluble sugar concentrations. Protein correlated negatively with total carbohydrate and soluble sugar concentrations. This result favored the selection of genotypes with high protein concentrations and lower total carbohydrate and soluble sugar concentrations.