Although Hevea is the dominant rubber crop today, Hevea and guayule have had parallel histories of development. In both, commercialization began with the harvest of wild stands before the establishment of plantations and the initiation of cultural studies. Variability within stands and lowered yields per unit area were problems in both species. These problems continued through the early attempts at cultivation since the populations were very heterogeneous genetically due to their establishment from open-pollinated seed. Annual yields have been increased dramatically in both, from 400 to over 2,000 kg/ha for Hevea (Bonner 1991), and from 300 to 1,000 kg/ha for guayule (Estilai and Ray 1991). The differences in development between the two crops can be associated with the initiation of the Rubber Research Institute of Malaya, in 1925. The Rubber Research Institute has been responsible for over 60 years of continuous increases in Hevea yields and the production of a uniform and reliable industrial product (Bonner 1991). Guayule, on the other hand, has suffered from intermittent research efforts, which have in many cases been undermined by periods of neglect. Guayule researchers have found themselves more than once in the position of "reinventing the wheel."
Guayule is the dominant perennial xerophytic shrub found on the limestone bajadas and hillsides of the Chihuahuan desert of north central Mexico and the Big Bend region of Texas (West et al. 1991). Wild stands contain a natural polyploid series of diploids (2n = 2x = 36), triploids (2n = 3x = 54) and tetraploids (2n = 4x = 72); and under cultivation, individual plants have been identified with chromosome numbers up to octaploid (2n = 8x = 144). Diploids reproduce predominantly sexually, and polyploids reproduce by facultative apomixis. Guayule also has a sporophytic system of self-incompatibility and many plants contain B- or supernumerary chromosomes (Thompson and Ray 1989).
After the closure of the processing facilities in Mexico, commercialization efforts were moved across the border to the United States. Agronomic studies and breeding work was initiated by the newly reorganized Intercontinental Rubber Company, with efforts centered in Arizona and California. By the late 1920s, annual production of guayule rubber, from commercial plantings of over 3,200 ha, was about 1,400 t. Seventeen years (1929) after the project was moved from Mexico to the United States, production ceased because of the Great Depression (Bonner 1991).
The second major effort to domesticate and utilize guayule as a source of natural rubber came with the Emergency Rubber Project of World War II. This was an extensive effort involving over 1,000 scientists and technicians, and 9,000 laborers. Over 13,000 ha of shrub was planted at 13 sites in three states. The effort ended with the end of the war and the development of synthetic rubber (Huang 1991). During the four years of its existence, 1,400 t of guayule rubber was produced, but at the end of the war an additional 10,000 t was destroyed while still in the shrub (Bonner 1991). However, the project was very successful, and from it came the bulk of our knowledge about the basic biology of guayule and the origin of the germplasm upon which current breeding programs are based (Thompson and Ray 1991). If this work had continued, undoubtedly today guayule would have already become a commercial rubber crop.
The third effort, of which we are still a part, arose from the quadrupling of crude oil prices in the early 1970s. This led to the enactment of the Native Latex Commercialization and Economic Development Act of 1978, which has supported the current guayule research for about 12 years (Huang 1991). Although this effort is neither as concentrated, or as urgent, as the Emergency Rubber Project, a tremendous amount of work has been done, and has resulted in significant increases in yield and the refinement of cultural practices to fit today's mechanized agriculture.
Guayule is adapted to hot desert environments, and sites with well-drained calcareous soils and relatively low concentrations of nutrients. Sandy-loam soil are most suitable since root diseases, which are exacerbated by standing water, are one of the few problems encountered in guayule cultivation (Mihail et al. 1991). Fertility treatments have been shown to have little effect on growth, and guayule is only slightly tolerant to soil salinity (Nakayama et al. 1991). The semiarid plateau region of the Chihuahuan desert (1,200 to 2,100 m in elevation) in which guayule occurs naturally has a temperature range between -18 and 49.5°C. High temperature does not appear to affect growth, but temperatures below 4°C induce semi-dormancy and extended freezing temperatures can cause plant death (Thompson and Ray 1989).
Areas with annual precipitation between 280 and 640 mm are preferable for guayule cultivation, but in order to achieve maximum yields, moderate to heavy applications of irrigation are necessary. Both dry matter production, and resin and rubber yields, have been shown to increase proportionally with increased water availability (Nakayama et al. 1991). In addition, irrigation can shorten the time until harvest. However, excess water is harmful to guayule plants of all ages, causing disease, reduced soil aeration, and increased weed competition. These problems are especially damaging to young plants (Nakayama et al. 1991).
Presently, stand establishment is accomplished by transplanting. Seeding transplants are produced in greenhouses and fields are established using typical commercial transplanting systems. Transplanting has been extremely successful, but is estimated to be more expensive than establishment by direct-seeding (Thompson and Ray 1989). Direct-seeding has been successful on an experimental scale, but no commercial scale plantings have been attempted.
Weed and pest control on a commercial scale are problematic since no compounds are presently labeled for guayule in the United States. However, pests have not proven to be a problem if sites were carefully selected, stands established by transplanting, and weeds successfully controlled (Mihail et al. 1991). If establishment via direct-seeding becomes the norm, experimental plot work has shown that damping-off of seedlings and weed control will become major areas of concern.
Mechanized techniques have been developed or adapted for all aspects of guayule cultivation. For example, the cost of transplanting may be reduced by clipping instead of digging whole plants. By clipping, the branches are cut approximately 10 cm above the soil level and regrowth occurs from the root crown. Novel equipment has been developed for this purpose and breeding programs are now selecting lines with high levels and rates of regeneration (Coates 1991).
The first and oldest method is flotation. This is essentially the same methodology used at the turn of the century and during the Emergency Rubber Project. In this procedure, ground shrubs are placed in a large vat of dilute sodium hydroxide until the woody tissue takes-up water and sinks to the bottom and the resinous rubber floats to the top in what are called "worms." These worms are skimmed from the top and the rubber is deresinated with acetone. Flotation was recently employed by the processing facility at Saltillo, Mexico, from which all of the guayule rubber used in test tires until 1990 was produced (W.W. Schloman, Jr. pers. commun.). The second method is sequential extraction, in which the resin is first extracted with acetone or another polar organic solvent, and then the rubber is extracted with hexane. Sequential extraction has only been used experimentally and appears not to be an economically viable method.
The third processing method is simultaneous extraction, in which a mixture of solvents, usually acetone and hexane or pentane, are used. After the initial extraction, more acetone is added to coagulate the high molecular weight rubber. This method has been used at both of the experimental processing facilities built by Texas A&M University at College Station, Texas and at Sacaton, Arizona by the Bridgestone/Firestone Corporation. Although this method has been successful in extracting rubber, engineering difficulties in handling the shrub have plagued both facilities (Wagner and Schloman 1991; N.G. Wright pers. commun.).
Economic forecasts suggest that for guayule to become a crop which can compete without subsidies, rubber yields must be increased and/or commercial utilizations of processing coproducts must be identified and developed (Wright et al. 1991). One potentially valuable coproduct is the low-molecular-weight rubber fraction, which accounts for approximately 25% of the total rubber yield. These low-molecular-weight rubber compounds have high value specialty applications as non-tire rubber (Schloman and Wagner 1991). Another processing coproduct, the resins, are only partially characterized, but are predominantly fatty-acid triglycerides and terpenoids. Resins have been used successfully as wood preservatives, a feed-stock for specialty chemicals (coatings and rubber additives), and as a high value fuel with no ash. Unfortunately, resin composition varies with shrub line, cultivation site, harvest date, and processing history (Schloman and Wagner 1991).
Guayule bagasse was used to fuel the early processing plants in Mexico and unprocessed shrub was used to fuel various processes in the Mexican mining industry. Today, bagasse is still being considered as a cogeneration fuel, as well as, a feedstock for gasification, conversion to liquid hydrocarbons, as a source of fermentable sugars, and as a fiber. These applications are not unique and are typical of other types of waste lignocellulose (Schloman and Wagner 1991).
Rubber yield can be expressed as a product of rubber content (% rubber) and biomass (dry weight/unit area). Thus, rubber yield may be improved by increasing either biomass and/or rubber content. An increase in rubber content is more desirable since it increases the processing efficiency of the shrub. Increased biomass involves additional costs associated with harvest, transportation, and processing (Estilai and Ray 1991). Dry weight has been found in many studies (Thompson and Ray 1989) to be the best predictor of rubber yield, with correlation coefficients between the variables typically over 0.90. Rubber content, on the other hand, has neutral to slightly positive correlation coefficients with yield, up to 0.30. There is a negative correlation observed in most studies, between biomass production and rubber content. However, gains have been made in both simultaneously by using minimum selection thresholds for both traits. In other words, a plant must have a minimum rubber content and biomass before being selected. A plant with either adequate (above the selection threshold) rubber content or biomass, but not both, would not be retained in the program.
Secondary breeding objectives are being addressed by particular breeding programs. Among these are increased resin and biomass production which have been realized in several new breeding lines. In addition, many seed characteristics have been addressed, such as size, germination rate, and yield per plant. Breeding for tolerance to environmental stresses and their interaction with rubber production have been studied, but on a very limited scale.
All breeding approaches depend upon the existing genetic variability found within available guayule germplasm. In guayule, limited genetic variability is not a problem, and has been found for every trait that has been evaluated. In fact, the facultative nature of apomixis in polyploid guayule continually releases new variability which may be exploited by plant breeders (Thompson and Ray 1989).
To date, the most extensively employed breeding approach has been single-plant selections from within apomictic polyploid populations. This has been used very successfully in increasing annual rubber yields, from approximately 300 to 1,000 kg/ha. Hybridization of polyploids is another method that has been suggested, but has been tried only sparingly. Interspecific hybridization has been discussed, but applied only on a limited scale. There is little to be gained by going to the related species to find desirable traits, since the available variability within the guayule germplasm has not yet been totally characterized and/or exploited (Estilai and Ray 1991).
Diploids have been used in guayule breeding in several ways. Because of their sexual (non-apomictic) reproduction, standard breeding methodologies may be employed, however, there are still problems in their use. Most notably, diploids yield significantly lower and are much more susceptible to root diseases than polyploids. Yields have been raised using modified recurrent selection schemes, and Verticillium tolerant lines have been developed through mass selection. These improved diploid lines can either be crossed to polyploids or have their chromosome numbers doubled with colchicine, producing polyploids. Diploids have also been used to release new genetic combinations from apomictic polyploids. Meiosis is normal in the pollen mother cells of apomictic polyploids, yielding potentially new and useful combinations of genes from which selections may be made. High yielding polyploids have been crossed onto diploids resulting in populations with excessive phenotypic variation (Thompson and Ray 1989; Estilai and Ray 1991).
Significant progress has been made, even though not more than 2.8 Scientific Years are presently devoted to guayule breeding each year in the United States. At the end of the Emergency Rubber Project, and through the 1950s, guayule rubber yields on an annual basis were between 220 and 560 kg/ha. This was based on the germplasm with which the current breeding programs started. By the Second Guayule Regional Variety Trials, 1985-1988, annual yields had increased to between 600 and 900 kg/ha, and in breeding plots annual yields of over 1,100 kg/ha have been estimated. These programs have been effective, but due to the limited number of researchers and resources, many aspects which might aid commercialization have been ignored. For example, coproduct production, dryland production, and screening for different environmental stresses have been addressed only indirectly.
On its own, guayule is presently not economical without either greater rubber yields or identification and development of high value coproducts. Under irrigated conditions, assuming rubber at $1.21/kg, resin + low-molecular-weight rubber at $0.44/kg, and bagasse at $0.04/kg, annual rubber yields of 1,450 kg/ha (1,300 lb./acre) would have to be obtained. These are not far from our present yields. Under dryland conditions, and using the same assumptions, annual rubber yields of 640 kg/ha (600 lb./acre) must be obtained, however, this would mean an increase of 2.5 times the present dryland yields. Unfortunately, these are rough estimates at best since both the costs of production and of processing are unknown (Wright et al. 1991).