Assessment of energy production in full-sibling families of elephant grass by mixed models

Abstract The search for alternative energy sources has grown recently. Surely, elephant grass ( Pennisetum purpureum Schum.) is a tropical grass of increased efficiency in the use of light and presents a high production of biomass, being cost-effective for direct combustion and charcoal obtainment. This study aimed to select the best families and individuals in full-sibling families of elephant grass by estimation of genetic parameters and prediction of genetic values by methods of restricted maximum likelihood (REML) and best linear unbiased prediction (BLUP), respectively, as an alternative of energy easy for Brazil. One hundred twenty progenies of full-siblings from bi-parental crossings were assessed in two harvests and one environment. The experimental design was a randomized block design with five-plant per plots and three replications. The assessed traits were dry matter production, tiller number, plant height, stem diameter, and neutral detergent fiber. For all traits, accuracy showed values above 90% and heritability of family means between 0.81 and 0.92%. Families 3, 1, and 2 were the best ranked, being the most indicated for selection. The REML/BLUP method proved to be suitable for selecting individuals of elephant grass for energetic purposes.