Correlação de caracteres de planta e tubérculo nas primeiras gerações de seleção de batata

Abstract

Correlation of plant and tuber traits of potato in early generations of selectionThe aim of this work was to evaluate canonic correlation as a tool to identify plant and tuber traits for early selection of potato clones. The canonical correlation analysis was carried out in two generations (seedling and tuber) with 358 clones, representing nine segregating families. The seedling generation (G1) was produced in a screen house during the winter of 2008, and the tuber generation (G2) in the field during the fall season of 2009. The evaluations were: height of the main stem (EHP), fresh weight of the aboveground part (MFA), stolons (MFE) and tubers (MFT), number of tubers (NTC), tubers with elongated (TAL) and flattened shape (TAC), skin texture (ASP), pointed ends (PON), secondary growth (EMB), eye depth (PRG), days to breaking dormancy (DRD) and number of sprouts per tuber (NBT). The variability among families was significant for all traits in both generations, except for PON in the first and MFA, EMB and PRG in the second generation. Linear correlations between G1 and G2 were significant, positive and high for 12 out of 13 evaluated traits. Based upon canonical correlation analysis, there is an association between plant and tuber traits between generations. However, there is no general rule for early selection of clones applied for all families, the evaluation of clones being the best strategy to identify correlated traits between generations.

Canonical correlations in agricultural research: Method of interpretation used leads to greater reliability of results

“…Many works have been published where the interpretation of the analyzes is carried out via canonical weights (coefficients) as in (Alves et al, 2017b(Alves et al, , 2016Bandinelli et al, 2017;Rigão et al, 2009), e even using a standardized structure, they can not fully report the canonical effects (Nimon et al, 2010). Thus, the use of canonical loadings and canonical cross-loadings for the interpretation of canonical correlations is preferable (Takada et al, 1984).…”