ABSTRACT
As large and plump seeds have higher market value, Ayoub et al. (2002) used image analysis to measure kernel size and shape attributes (area, perimeter, length, width, F-circle and F-shape) in grain samples of 140 doubled-haploid lines from a two-rowed barley cv. Harrington by six-rowed barley cv Morex cross. They used interval mapping to map QTLs affecting the means and width-sample standard deviations of these attributes using a 107-marker genome map. Regions affecting one or more kernel size and shape traits were detected on all seven chromosomes. These included one near the vrs1 locus on chromosome 2 and one near the int-c locus on chromosome 4. Some, but not all, of the QTLs exhibited interactions with the environment and some QTLs affected the within-sample variability of kernel size and shape without affecting average kernel size and shape. When QTL analysis was conducted using data from only the two-rowed lines, the region on chromosome 2 was not detected but QTLs were detected elsewhere in the
genome, including some that were detected in the analysis of the whole population. Analysis of only the six-rowed lines did not detect any QTL affecting kernel size and shape attributes.
