COMPARING INTRA- AND INTERENVIRONMENTAL PARAMETERS OF OPTIMAL SETTING IN BREEDING EXPERIMENTS

A series of biometrical and quantitative-genetic parameters, not well known in Croatia, are being used for the most important agronomic traits to determine optimal genotype setting within a location as well as among 7locations. Objectives of the study are to estimate and to compare 1) parameters of intra-environment setting (effective mean square error EMSE, in lattice design, relative efficiency RE, of lattice design LD, compared to randomized complete block design RCBD, and repeatability Rep, of a plot value), and 2) operative heritability h2, as a parameter of inter-environment setting in an experiment with 72 maize hybrids. Trials were set up at four environments (two locations in two years) evaluating grain yield and stalk rot. EMSE values corresponded across environments for both traits, while the estimations for RE of LD varied inconsistently over environments and traits. Rep estimates were more different over environments than traits. Rep values did not correspond with h2estimates: Rep estimates for stalk rot were higher than those for grain yield, while h2for grain yield was higher than for stalk rot in all instances. Our results suggest that due to importance of genotype × environment interaction, there is a need for multienvironment trials for both traits. If the experiment framework should be reduced due to economic or other reasons, decreasing number of locations in a year rather than decreasing number of years of investigation is recommended.