Indicative traits of aluminium sensibility in rice. II. Canonical correlation with yield

As a step in the development of a method to access differential Al tolerance, data from grain yield of 20 rice genotypes growing in high and low Al saturation plots, were correlated with morphological characters of the same genotypes grown in nutrient solution. By the canonical correlation it was possible to obtain two linear combinations, among several possibilities. These combinations retained the root area and the maximal leaf length, combined with four Al levels (0, 10, 20 and 30 mg L-1 Al) as the first canonical variable. The second canonical variable considered the weighted difference between grain yield at low and high Al saturation level or the unweighted difference, with r 0.93 and 0.91, respectively. The results showed the possibility to correlate morphological traits, measured in nutrient solution, against field data, reducing uncertainty in choosening the most suitable indicator.

Saved in:
Bibliographic Details
Main Authors: Vicente, Fabiano Mascarenhas Paulo, Rossiello, Roberto Oscar Pereyra, Pereira, Maurício Ballesteiro
Format: Digital revista
Language:por
Published: Pesquisa Agropecuaria Brasileira 1998
Online Access:https://seer.sct.embrapa.br/index.php/pab/article/view/4826
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As a step in the development of a method to access differential Al tolerance, data from grain yield of 20 rice genotypes growing in high and low Al saturation plots, were correlated with morphological characters of the same genotypes grown in nutrient solution. By the canonical correlation it was possible to obtain two linear combinations, among several possibilities. These combinations retained the root area and the maximal leaf length, combined with four Al levels (0, 10, 20 and 30 mg L-1 Al) as the first canonical variable. The second canonical variable considered the weighted difference between grain yield at low and high Al saturation level or the unweighted difference, with r 0.93 and 0.91, respectively. The results showed the possibility to correlate morphological traits, measured in nutrient solution, against field data, reducing uncertainty in choosening the most suitable indicator.