The sex segregation ratio was checked in biparental families of Actini
dia deliciosa (2n=6x=l74) obtained by crossing four females (A12, Mo3,
Br4, Hw1) with two males (T2, M1) and one fruiting male (M3h, subandr
oecious) according to a factorial mating design. The M3h fruiting male
was also selfed. The sex ratio was checked in maternal families of A.
kobomikta (2n=2x) and A. chinensis (2n=2x) as well as in A. deliciosa
. Seedlings of both diploid species took 3-4 years to progress beyond
juvenility, whereas a noticeable number of seedlings from biparental c
rosses of A. deliciosa involving A12 and Hw1 as seed parents were stil
l non-flowering after seven growing seasons. Open-pollinated families
of both diploid and hexaploid species as well as most families from bi
parental crosses showed a sex segregation ratio approaching 1.1. Suban
droecious lines with different degrees of ovary and pistil development
appeared in proportions of 0-4.2%, depending on the cross, but only 6
of the 2567 male vines checked were capable of setting fruit. No case
of self-fertility or apomixis was detected among 1866 bagged female v
ines. Selfed M3h progenies gave only female and male phenotypes in a r
atio of I female to 3 males. No off-type vines were found among these
progenies. The same disomic sex segregation ratio seems to be operatin
g at different ploidy levels in the genus Actinidia. Since selfed frui
ting males produced both female and male individuals, the male sex app
ears to be the heterogametic one. Such evidence indicates that a monof
actorial system based on one or more linked genes or on an X/Y chromos
ome set must be controlling sex expression. How a monofactorial sex-de
termining mechanism could operate in polyploids to give a 1:1 female:
male ratio is discussed. Minor modifying gene(s) seem to be responsibl
e for the feminization of males, and their expression appears enhanced
by environmental conditions. Masculinizing gene(s) seem to be lacking
in female genotypes.