An investigation of two-dimensional yield variability in breeders' small plot barley trials

Summary This paper describes an investigation into yield variability and spatial correlations in 129 spring barley trials using small plots between 2·4 and 6·0 m2 in area. We report on the efficiencies of the generalized lattice designs used, and discuss the effects of plot size and year-to-year variations. Results indicate that the recommendations of Patterson & Hunter (1983) on choice of block size apply equally well to small-plot trials. The observed variability and correlations are then discussed and so is the potential for improvement by using alternative designs. Patterns of variability were not at all consistent from year to year, or even between trials in the same year, but it appears important to take account of column effects as well as row effects.