A genetic model based on a two-level intra- and interspecific mating design
is proposed to estimate the genetic architecture of species differences an
d heterosis for outcrossing species. The underlying genetic analyses make u
se of classical quantitative genetic theories and recent results from molec
ular genetic studies. Gene effects across different quantitative trait loci
(QTL) can be approximated by a geometric series. Under natural selection,
gene effects are often associated with allele frequencies in a particular w
ay, which can be approximated by the gamma distribution. By incorporating t
hese approximations into family structural analyses in the mating design, w
e are able to estimate a number of genetic parameters that contribute to qu
antitative genetic variation based on a nonlinear optimization approach. Th
ese parameters include the number of QTL, their gene effects, and their all
ele frequencies in the parental populations. We perform simulation studies
and illustrate an example to demonstrate the statistical property and proce
dure of the method.