engleski

GxE interaction in Croatian cereal variety trials : I. Estimation of variance components for yield

Variety testing in Croatia is organized by the Ministry of Agriculture and Forestry with the engagement of different institutions, persons and firms. Field experiments planning and statistical data analyses is confided to the Department of Plant Breeding, Genetics and Biometrics. A relatively small number and the closeness of the sites, raised suspicions about the narrowness of necessary environmental variability. Aware of these weaknesses and many other irregularities, this paper is attempt to present cereal (spring oats,winter wheat, spring barley and winter barley) trial yield results of five years and compare them with some European countries. Unbalancedness of data set required REML as a method of estimating variance components. Comparing the results of the analyses for different species, we noticed certain pattern of variability breakdown. The highest portion of variability always stands for site x year interaction and site, but all other variance components have similar relative values as well, except variety and year components. Despite the different distribution pattern, the variety x environment components generally have similar values. Comparison with the results from other countries' trials showed likeness. Inspite of our complex situation range of environments is still wide enough to include all extremes, despite their relatively small number.

cereals; variety testing; GEI; variance components; REML

nije evidentirano