Plant breeding in guayule, a rubber and latex producing plant, has not been
as effective as predicted or desired. A surprisingly large amount of varia
bility has been reported in this apomictic crop for traits such as plant he
ight, width, resin, rubber, and latex contents. This study was designed to
calculate the proportion of the total measured variability due to the envir
onment and the proportion due to genetic influences within and between thre
e-released germplasm lines. Plant heights were measured at 1, 2 and 3-years
-of-age; plant width, and latex contents at 2 and 3-years-of-age; and resin
and rubber content at 2-years-of-age. Broad-sense-heritabilities were esti
mated for each trait and year by dividing the genotypic by the total varian
ce. To estimate the genetic component of the measured variance for each tra
it, the environment effects (variance from clonally propagated plants) were
subtracted from the total variance (variance from open-pollinated (OP) see
d propagated plants). In general, the variances of the means for the measur
ed traits were lower in the clonally propagated plants compared with the ap
omictic OP seed propagated plants. The heritability estimates calculated fo
r each trait differed from year to year. For instance, heritability for pla
nt height was estimated in line AZ-2 to be 0.84 at 1 year of age; 0.47 at 2
years of age; and 0.0 at 3 years of age. These values imply that a large p
ortion of the observed variation in this line is attributed to genetic effe
cts in the first 2 years of growth. As the plant grows over several seasons
, the environment effects compound, masking the genetic effects, making eff
ective selection choices more difficult. Heritability for latex content for
the same line was estimated to be 0.97 for the second year and 0.55 for th
e third year. Selections for the measured traits in this study appear to be
most effective during the first and second years of growth, with effective
ness diminishing during the third year. Most selections previous to this st
udy were performed between 3 and 5 years of growth, thus suggesting one rea
son for the lack of significant progress in most breeding programs. (C) 200
1 Elsevier Science B.V. All rights reserved.