100 Years of Breeding Guayule - Industrial Crops: Breeding for BioEnergy and Bioproducts - page 354

characters related to biomass production, were highly and consistently correlated to
rubber yield [ 29 , 30 ]. The characters shown to be the best predictors of rubber
content were plant fresh and dry weight and percent dry weight and plant volume,
and the best predictive model for rubber yield includes plant height and width and
volume and dry weight [ 29 ].
Ray et al. [ 31 ] tested the relatedness of apomictic parents and their open-
pollinated, half-sib progeny families for eight components of yield. Heritability
estimates were made by measuring the components of yield in both the parents and
progeny. The parent plants were all open-pollinated progeny of a single-plant
selection made by D.D. Rubis (University of Arizona), and measurements were
made when the parent plants were 3 years old and the progeny plants, 2 years old.
For rubber yield, rubber content, resin content, fresh weight, dry weight, percent dry
weight, height, and width, none of the parent-progeny regressions were signifi-
cantly different from zero. For all characters, a large range of phenotypic variation
was observed, and the range and standard deviation of the parents were greater than
among the progeny. This was probably due to the compounding of environmental
effects (the parent plants were a year older than the progeny plants) rather than a
difference in genetic variability [ 32 ]. Linear correlations were performed to study
the relationship between rubber yield and the other seven characters, and fresh and
dry weights were highly and positively correlated with rubber yield in all
populations. Thompson et al. [ 30 ] found significant correlations between rubber
content and resin content that were higher than correlations of any other character
with rubber content. This high correlation means that breeders should be able to
create new lines that are higher in both rubber and resin than older lines. Because
both rubber and resin are important characters in determining the value of guayule
end products, breeding for simultaneous increases in these traits is important to
insure successful commercialization. Evidence that this is possible is found in the
release of six new germplasm lines that are higher in rubber and resin the older
USDA lines [ 22 ].
Biomass appears to be the best predictor of rubber yield (rubber yield

plant
biomass rubber concentration). Thus, plant growth or biomass production can be
used as a primary selection index for rubber yield. However, selection for large
plant size may be disadvantageous because larger plants may result in mechanical
harvesting problems, increased transportation and handling costs, and reduced
efficiency of rubber extraction in the processing plant. These are all significant
economic factors in the production of rubber from guayule. For this reason,
selection of plants with higher rubber concentration in concert with adequate
biomass production must receive primary attention. Such selection is difficult
because there is often a negative correlation between rubber concentration and
biomass [ 30 ].
Yield trials have been used successfully to evaluate guayule germplasm lines
under various environmental conditions [ 19 , 33 ]. However, this valuable tool has
not been consistently available to breeders due to a lack of continuous funding.
More consistent funding is needed to carry these trials to completion and initiate

¼