USE OF MOVING MEANS IN GRASS YIELD TRIALS

Open Access
Publisher Website
Google Scholar
Abstract
Analyses of data from three grasses of contrasting growth-type (intermediate wheatgrass, Agropyron intermedium (Host.) Beauv.; Russian wild ryegrass, Elymus junceus Fisch.; and Altai wild ryegrass, Elymus angustus Trin.), seeded in different row spacings, showed that the moving mean adjustment of plot yields prior to analysis gave improved control of the experimental error as compared to variance analysis of unadjusted values. The optimum number of adjacent plot yields to include in the moving mean varied widely from test to test as well as from year to year in the same test. Plant breeders may have to run several analyses to obtain the most accurate adjustment. However, this can be easily accomplished in one computer run by inserting appropriate values into the program. A lattice analysis made on several trials suggested that the lattice analysis and the use of the moving mean adjustment in a random block design were comparable in relative efficiency, but the moving mean adjustment in a random block design is more flexible in terms of number of lines that can be assessed.

This publication has 0 references indexed in Scilit: