Desarrollo y predicción de sintéticos de cruzas dobles de maíz.
Development and prediction of double cross synthetics of mayze. Hybrid varieties (VH) have been identified among the best, under optimal conditions, to exploit the genetic potential of several crop species. Sowing F2 plantings or more advanced seed, does not regenerate the VH. A population similar t...
|Main Authors:||Sahagún Castellanos Ruiz, Jaime, Rodríguez Pérez, Juan Enrique, Peña Lomeli, Aureliano|
Universidad de Costa Rica
Development and prediction of double cross synthetics of mayze. Hybrid varieties (VH) have been identified among the best, under optimal conditions, to exploit the genetic potential of several crop species. Sowing F2 plantings or more advanced seed, does not regenerate the VH. A population similar to a synthetic variety (VS) is obtained instead, particularly if the hybrid is a double cross. A VS is characterized by possessing a population buffering which makes, it to be adequate for low input agricultural conditions. Considering that building up VS with double crosses is a procedure assessing the capacity of a higher number of lines to form superior VS, and since it is important to increase the theoretical knowledge about performance prediction, six prediction formulae (FP) for a VS built-up with n double were derived. Three formulas (FPAA) were based on the field evaluation of the progenies produced by the direct crosses among parents [the n double-cross hybrids (CD), the involved 2n single crosses (CS), or the 4n lines] and the populations produced by randomly mating (PAA) in isolation. The individuals representing each parent. In the additional three FPR, each PAA population was evaluated in two parts: one produced by selfing and the remaining. All six were unbiased, but between types and among parents the FPR methods and lines, respectively, showed the lowest variances. The superiority of the FPR in relation with the FPAAmethods ranked from 8 to 40 %. In addition, for n ≥ 2, the estimation that required less plots was based on CD, but the precision was the lowest. On the other hand, for n=1 and a fixed number of plots the most precise estimator was based on CD as well.