Pk. Lawrence et Ih. Delacy, CLASSIFICATION OF LOCATIONS IN REGIONAL COTTON VARIETY TRIALS WHERE TRIAL ENTRIES CHANGE OVER YEARS, Field crops research, 34(2), 1993, pp. 195-207
Cultivars and advanced breeding lines evaluated for performance in reg
ional trials usually change over successive years. Cluster analysis pr
ocedures are presented for analysing these unbalanced genotype X locat
ion X year data-sets to help study the effectiveness of locations to d
iscriminate amongst genotypes (lines). To illustrate the use of these
methods, yield data from 12 years of cotton evaluation trials conducte
d at eleven locations in eastern Australia were analysed. The eleven l
ocations were classified according to the way they discriminated among
st lines. The combined classification analysis provides the plant bree
der with a better understanding of the relationships amongst the locat
ions used for regional cultivar trials and provides a rational method
for choosing a few key locations to be used for early generation testi
ng of lines from a breeding program.