Incomplete Steiner Triple Systems over balanced square lattices to diallel analysis / Sistemas triplos de Steiner incompletos versus Látices Quadrados Balanceados para a análise dialélica

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

The aim of this work was to compare a fraction of a 15 treatments Steiner Triple Systems (STS) to a 4x4 balanced square lattice design (LQB). We simulate experiments with 15 or 16 treatments. To the former, 5 superblocks of the STS was augmented of 1 treatment each. To the LQB one treatment was repeated if only 15 was in the experiment. To simulate treatment effects, a diallel cross was assembled with 6 parents (15 progeny). An adictional progeny of independent parents was added when an extra treatment was needed. LQB was balanced with 5 superblocks (80 plots). THe fraction of the STS used just 75 plots. For each experimental situation 5000 runs where simulated. REML estimates of variance components, as well as additive and dominance effects where worked out. Confidence limits for all estimates where very similar in all cases. To an equivalent size experiment, STS was more efficient than LQB in minimizing the error on parent selection in several combinations of additive and dominance variance components. STS are as efficient as LQB with fewer plots and should be used as a good alternative for diallels when treatment number is suitable.

ASSUNTO(S)

design efficiency diallels estatistica dialelos incomplete block designs blocos incompletos mixed models modelos mistos eficiência de delineamentos

ACESSO AO ARTIGO

Documentos Relacionados