Oil content increase and toxicity reduction in jatropha seeds through family selection

Abstract

The objectives of this study were to verify the existence of genetic variability among half-sib families and select the best families in order to increase oil content and reduce toxicity in jatropha seeds. The experiment was carried out in complete randomized block design with two replications and five plants per plot, during five crop years (2010–2014). Plots were evaluated for grain yield (g/plant) and for weight of 100 seeds (g), and in the year of 2014, seed oil content and phorbol ester concentration (mg/g) were also evaluated. Analysis of variance was carried out considering genotypes as random, and from this analysis, genetic parameters were estimated for the characteristics of interest. Diversity analysis was carried out using the Mahalanobis distance matrix. Tocher and UPGMA clustering methods were also used. Gain with selection was estimated using the method of direct selection on the main variable and indirect selection on the other variables, based on the selection of the former. Different selection indexes wegain) in order to estimate the gain and select the top 20 families. The mulre also used (classic, base index, sum of ranks, multiplicative index and index based on the desired tiplicative index was the best index for jatropha because using this index we obtained gain for all traits together. It was concluded that there is genetic variability among half-sib families, and that selected families by multiplicative index made it possible to increase oil content and decrease toxicity in jatropha seeds.