Mapping of quantitative trait loci (QTLs) in plants was performed using the
data of marker genotypes and phenotypic values of a quantitative trait in
a segregating generation such as F-2 and BC1 derived from the crossing of t
wo pure lines. A large sample must be examined to map QTLs with small effec
ts. However genotyping of many markers for a large number of individuals is
costly. For mapping QTLs using a F-2 population, useful information is inc
luded in the derived F-3 lines. Phenotypic measurements of F-3 lines can be
obtained at a relatively low cost. Therefore, by utilizing the phenotypic
measurements of F-3 in addition to F-2 data, the detection of QTLs could be
improved even when only a small number of F-2 individuals are genotyped. I
n this paper a model for mapping QTLs is proposed, where the data of both F
-2 individuals and F-3 lines are jointly analyzed. The effectiveness of thi
s method is evaluated theoretically and numerically and it is shown that th
e ability of detecting QTL can be remarkably enhanced, especially when the
heritability of the trait is low.