The power of a backcross design to detect genetic variation associated
with a single chromosome was investigated. A simple chromosomal test
was suggested in which the phenotypic observations are regressed onto
genotypic information from multiple markers. It was shown that the opt
imum marker spacing depends on the underlying genetic structure and ch
romosome length. A sparse marker map, with markers approximately every
50 CM, is sufficient to detect chromosomal variation if the nature of
the genetic variance is coupled polygenes, whereas the optimum marker
spacing to detect a single QTL somewhere on the chromosome is slightl
y denser, about 20-40 cM. Although the method was demonstrated for lin
e crosses, it can equally be applied to other populations, for example
four-way crosses and half-sib designs.