Research Article

Phenotypic characteristics for discrimination between advanced genotypes of Brachiaria ruziziensis

Published: March 31, 2016
Genet. Mol. Res. 15(1): gmr7774 DOI: 10.4238/gmr.15017774

Abstract

The aim of this study was to use multivariate methods and Pearson and partial correlations to disregard phenotypic characteristics that contribute little to differentiation between Brachiaria ruziziensis genotypes. Eighty-one genotypes of B. ruziziensis were assessed in completely randomized blocks with three replications. Ten phenotypic characteristics were assessed: plant height, leaf length, leaf width, sheath length, length of the flower stem, length of the inflorescence axis, number of racemes per inflorescence, length of the basal raceme, number of spikelets per basal raceme, and width of the rachis. The best traits for differentiation between genotypes were determined by assessing relative contribution to diversity, canonical variables, as well as Pearson and partial correlations. Four canonical variables were found to account for 57% of the overall variation, while plant height, sheath length, and number of racemes per inflorescence were considered traits that could potentially be disregarded in future assessments.

The aim of this study was to use multivariate methods and Pearson and partial correlations to disregard phenotypic characteristics that contribute little to differentiation between Brachiaria ruziziensis genotypes. Eighty-one genotypes of B. ruziziensis were assessed in completely randomized blocks with three replications. Ten phenotypic characteristics were assessed: plant height, leaf length, leaf width, sheath length, length of the flower stem, length of the inflorescence axis, number of racemes per inflorescence, length of the basal raceme, number of spikelets per basal raceme, and width of the rachis. The best traits for differentiation between genotypes were determined by assessing relative contribution to diversity, canonical variables, as well as Pearson and partial correlations. Four canonical variables were found to account for 57% of the overall variation, while plant height, sheath length, and number of racemes per inflorescence were considered traits that could potentially be disregarded in future assessments.