Index | Search | Home | Table of Contents

Ray, D.T., D.A. Dierig, and A.E. Thompson. 1990. Facultative apomixis in guayule as a source of genetic diversity. p. 245-247. In: J. Janick and J.E. Simon (eds.), Advances in new crops. Timber Press, Portland, OR.

Facultative Apomixis in Guayule as a Source of Genetic Diversity

Dennis T. Ray, David A. Dierig, and Anson E. Thompson

  6. Fig. 1
  7. Fig. 2


The United States is totally dependent on imports of natural rubber for industrial purposes. The annual cost of importing rubber amounts to nearly $1 billion (Thompson and Ray 1988). Guayule (Parthenium argentatum Gray, Compositae), a small rubber-producing perennial shrub native to the Chihuahuan desert of north-central Mexico and southwestern Texas, has been long-recognized as a promising alternative source of natural rubber with the potential for cultivation in the arid and semi-arid environments of the American southwest.

Successful commercialization of guayule depends largely upon the development of higher yielding cultivars from the available germplasm. Initially, it appeared that the generic base from which improvement depended was narrow. Fifteen of the 26 available families of germplasm are selections from bulk-seed from a five-plant population collected at Durango, Mexico (Thompson and Ray 1988). Materials derived from this collection have exhibited extreme variability both within and between lines for rubber quality and quantity, dry weight, resin content, yield, and ploidy level (Ray and Thompson 1986, Thompson and Ray 1988, Thompson et al. 1988).


Guayule reproduces predominantly by apomixis (asexual reproduction by seed). Although it has been assumed in guayule that apomixis assures genetic uniformity from generation to generation, its facultative nature (apomixis and sexuality coexisting) and the high amount of heterozygosity in individual plants and the heterogeneous make-up of populations, results in the release of considerable variation whenever sexual reproduction (amphimixis) occurs (Powers and Rollins 1945).

Polyploidy and aneuploidy are common in guayule (Powers 1945, Bergner 1946, Ray and Thompson 1986). Generally, diploids reproduce sexually and polyploids reproduce by apomixis. In apomictic guayule the embryo sac develops directly without meiosis from the megaspore mother cell (MMC). Pollination is not necessary for embryo development, but is needed for normal endosperm development (Powers 1945). Meiosis in the pollen mother cells is normal, resulting in reduction of the chromosome number in the male gametophytes.

The facultative nature of apomixis in guayule results in four classes of progeny (Esau 1946). The origin and relative chromosome numbers of these four classes from tetraploid (2n = 4x = 72) parents illustrates the complexity of reproduction and the potential for release of genetic variability in this species (Fig. 1). The predominant class of progeny arise from non-reduction of the MMC, without fertilization. These are apomictic tetraploid progeny and are identical generically to the maternal parent. Progeny from fertilized, unreduced MMCs will include plants with increased ploidy levels. In our example, these progeny would be hexaploid (2n = 6x = 108). Polyhaploid (2n = 2x = 36) plants are the result of meiotic reduction of the tetraploid MMC, and embryo development without fertilization. The final class would be amphimictic tetraploid (2n = 4x = 72) progeny that arise from normal reduction and fertilization. Thus, two reproductive modes produce tetraploid progeny, one by apomixis and the other by sexual reproduction. The remaining two progeny classes vary in chromosome number from the parental population.


Our objectives were to quantify variation existing within and between guayule breeding lines derived from single plants, and to understand how apomixis may act to affect this variation. In 1986, 42 breeding lines were selected for further study on the basis of superior yield, rubber concentration, and top regrowth. Plants consisted of four-year old progeny rows from single plant selections of facultatively apomictic plants. In February 1987, ten randomly selected plants were individually harvested from each of these 42 lines and 17 characters were measured and described by Dierig et al. (1989a, b). Rubber content (%), rubber yield (g/plant), dry weight (kg/plant), and height (cm) were analyzed (Fig. 2).

Variation for the reported characters suggest that there are genotypic differences between lines that could be exploited through further selection (Dierig et al. 1989a, b). For example, the means of the 42 lines for rubber content ranged between 4.9 and 9.8% (Fig. 2). Significant variation was also observed within lines. Variation within some lines was less than 1% rubber (e.g. 8.5 to 9.2%), while others were over 5% rubber (e.g. 6.5 to 11.7%). Variation between and within lines for the other characters also varied significantly (Fig. 2). The completion of tests designed to elucidate the contribution of the genetic and environmental components and the analysis of chromosome numbers of progeny will enable us to estimate the amount of variability due to genetic hybridity, recombination, euploidy, and/or aneuploidy


There are two means by which variability can be produced in guayule populations. The first is by genetic segregation from meiotic reduction of a MMC, either with or without fertilization. Since guayule plants are highly heterozygous, this would account for many new gene combinations in the progeny The second is by the genetic imbalance resulting from polyploidy and/or aneuploidy. Thus, the high amount of variability we have found among and within guayule lines indicates that continued progress by selection is feasible.


Fig. 1. Reproduction in guayule resulting in four classes of progeny.

Fig. 2. Variability for (A) rubber content(%),

(B) rubber yield (g/plant),

(C) dry weight (kg/plant), and

(D) height (cm) among and within 42 progeny rows from single plants. The lines are in the same order for each character and within each line the square represents the mean, the vertical line the range, and the dots the standard error for the population.

Last update August 27, 1997 by aw