Index | Search | Home | Table of Contents

McLaughlin, S.P. 1996. Domestication of Hesperaloe: Progress, problems, and prospects. p. 395-402. In: J. Janick (ed.), Progress in new crops. ASHS Press, Arlington, VA.

Domestication of Hesperaloe: Progress, Problems, and Prospects

Steven P. McLaughlin*

    1. Growth and Biomass Production
    2. Physiology
    3. Genetic Variation
    1. Progress
    2. Problems
    3. Prospects
  5. Table 1
  6. Table 2
  7. Table 3
  8. Fig. 1
  9. Fig. 2
  10. Fig. 3
  11. Fig. 4
  12. Fig. 5
  13. Fig. 6

Representatives of a major United States pulp and paper company, the James River Corporation (JRC), approached the University of Arizona in late 1985 about establishing a domestic sisal (Agave sisalana) industry. At that time JRC was a large, diversified pulp and paper company; sisal was one of the raw materials used by their specialty paper business. Sisal is a tropical hard fiber crop originally domesticated as a source of cordage fibers. When pulped, sisal cordage yields long, thin fibers that command a high price in the paper industry (Table 1). JRC initially was interested in developing a domestic sisal industry in order to obtain a more reliable supply of this fiber at a low cost.

JRC representatives were advised that sisal could not be successfully grown in Arizona because of its lack of frost tolerance. Bulbils of sisal obtained from the Huntington Botanical Garden were planted in Tucson as a demonstration and they deliquesced and died immediately after temperatures first dropped to 0°C in Dec. 1985. Sisal was selected for domestication because it made good rope, not because it made good paper, but many Southwestern relatives of sisal have frost tolerance and were used by indigenous peoples for cordage products (McLaughlin and Schuck 1991). We jointly decided therefore to screen various members of the Agavaceae to determine if any (1) possessed high-quality fibers and (2) were suitable for production in the temperate climates of the United States.

During 1986 leaf samples were collected from the wild and from several botanical gardens, including the Desert Botanical Garden in Phoenix, the Huntington Botanical Garden in San Marino, California, and the Boyce-Thompson Southwestern Arboretum in Superior, Arizona. The screen included over 100 collections of 62 species from six genera in the Agavaceae: Agave, Dasylirion, Furcraea, Hesperaloe, Nolina, and Yucca. (Many systematists today place Dasylirion and Nolina in a separate family, the Nolinaceae). All leaf samples were sent to James River's Neenah Technical Center in Wisconsin, where they were pulped and made into paper samples that were tested for strength properties.

The best prospects to emerge from the screen were Hesperaloe funifera and Hesperaloe nocturna, both because they had superior fiber characteristics and because they had favorable agronomic traits (McLaughlin 1993). The results of this screening study were never published. In fact, JRC consistently discouraged any disclosure of their participation in the research and development of Hesperaloe. Populations of several species were subsequently resampled and fiber lengths, fiber widths, and cell-wall dimensions were determined at the University of Arizona (McLaughlin and Schuck 1991). Fibers of both Hesperaloe species were longer and thinner than those of any other species examined. Indeed, the length-width ratio of fibers from Hesperaloe funifera is greater than those of most other paper-making fibers and comparable to that of abaca (Musa textilis), the premium paper-making fiber (Table 2).

Based on the results of the 1986 screen, JRC and the University of Arizona selected Hesperaloe funifera and H. nocturna for further research and development. JRC worked on pulping, paper-making, and product development while the University of Arizona worked on various aspects of the agronomy and biology of the plants. Although both Hesperaloe species were not common in nature, we obtained sufficient seed of H. funifera to initiate long-term biomass production studies in 1988 but similar studies could not be started with H. nocturna until 1990.


Growth and Biomass Production

Experimental details of the long-term growth and biomass studies have been presented previously (McLaughlin 1993, 1995). We estimated fresh weights on a sample of 20 plants of Hesperaloe funifera at 4-week intervals over a period of five years (Fig. 1). Hesperaloe species are acaulescent and their leaves are tightly packed into clusters of several rosettes. In its first year H. funifera grew very slowly. During years 2 and 3 growth was most rapid in the latter half of the year with maximum growth rates about 10 g FW (fresh weight) plant-1 day-1. Approximately half of the plants flowered during their third year. This was a critical stage because lateral rosettes were also initiated at about the time of flowering. More rapid growth followed the development of these lateral rosettes during years 4 and 5, when maximum growth rates were about 40 g FW plant-1 day-1 (Fig. 1). Within-stand variation in plant size appeared to reach a maximum in year 4 and decline in year 5.

Biomass production for Hesperaloe funifera has been examined in six 300-m2 plots for seven years. The plots were established at three densities and have been monitored by randomly sampling plants from within the plots at the end of each growing season (McLaughlin 1995). The plots at the highest density (27,000 ha-1) produced a yield of 192 t FW ha-1 after five years (Fig. 2). Very little aboveground growth was observed during the first year. Plots harvested at the end of year 5 regrew slowly in year 6; growth rates in year 7 were comparable to those observed in year 5 prior to harvest.

The relationship between stand density and individual plant biomass is presented in Fig. 3. There was little evidence of competition among Hesperaloe funifera plants in these plots during their first three years, i.e., there was no discernable relationship between stand density and mean plant size. However, effects of crowding were evident in years 4 and 5 after plants flowered and produced their first group of lateral rosettes. The highest density used in this study was achieved using a 61 cm row spacing. Arizona cotton growers most often use 102-cm rows. Hesperaloe planted on 46 cm centers within 102 cm rows would correspond to a density of approximately 21,000 plants ha-1. From Fig. 3 it can be estimated that a stand of 21,000 plants ha-1 should have an average plant size of 8.5 kg FW plant-1 and a standing crop at first harvest of 180 t FW ha-1.

The crop cycle for Hesperaloe is still unclear. Biomass yields can be optimized by delaying the initial harvest until the end of year 5. It appears now that the first reharvest would be at year 8, not year 7 as previously suggested (McLaughlin 1995). A second reharvest might be achieved at year 10 (Fig. 4). Plants should regrow more rapidly after the second cut at year 7 than after the first cut at year 5 because they will have more meristems (rosettes) from which leaves can be produced.


Damian Ravetta studied gas exchange in Hesperaloe species (Ravetta 1994; Ravetta and McLaughlin 1993), and determined that these species have Crassulacean Acid Metabolism (CAM). CAM plants close their stomata during most of the daylight hours but open them at night, absorbing CO2 and transpiring water when temperatures are lower and relative humidity is higher. Hesperaloe species were found to have low maximum photosynthesis rates, low transpiration rates, and very high water-use efficiency (WUE), all consistent with CAM. The new leaves of H. funifera emerge from the central bud at a 90° angle to the ground surface and remain at high angles for 2-3 years. This vertical-leaf orientation is also consistent with CAM, because vertical leaves absorb little solar radiation at midday when their stomata are closed, thus avoiding heat stress.

Ravetta (1994) found that photosynthetic rates in H. funifera were highest during the fall months. High fall photosynthetic rates are consistent with the high daily growth rates that are observed in the fall (Fig. 1). Solar angles and night temperatures are lower in the fall than in the summer months and this may promote the higher photosynthetic rates. There may also be enhanced sink capacity in the fall when the lateral rosettes which emerged during the summer are growing at a rapid rate.

The agronomic significance of CAM is high WUE, and our initial biomass production trials confirmed that the high physiological WUE of Hesperaloe funifera did indeed translate into low water requirements (McLaughlin 1995; Table 3). Water-use efficiency was very low during the first year of stand establishment from transplants and very high during the fifth year (0.29 to 0.44 t DW cm-1). For comparison with the WUE values shown in Table 3, alfalfa grown in Arizona has a WUE of about 0.09 t DW cm-1. Averaged over 5 years, including year 1 when the crop uses all resources inefficiently, WUE in H. funifera is about twice that of C3 crops. Regrowing stands are expected to have a higher WUE than newly established stands.

Genetic Variation

A collection of 10 accessions of Hesperaloe funifera was assembled from populations from the wild and from cultivation (McLaughlin and Schuck 1992; McLaughlin 1994). These accessions were grown in small, replicated plots in a common-garden study in Tucson. There were significant differences among these populations in plant basal circumference, leaf lengths, leaf widths, plant fresh weights, number of rosettes at year 3, fiber lengths, fiber widths, and cell-wall thickness. No accession combined the best agronomic traits with the best fiber traits. The phenotypic variation observed in the common garden indicated a potential for improving the crop through classical breeding and selection. Because of the long life cycle of Hesperaloe, genetic improvement would be a long-term process that would likely follow, rather than precede, successful commercialization. Repeated measurements from individual plants showed good correlations between early plant fresh weights before first flowering (year 2), at an initial harvest (year 4), and at reharvest (year 6). Thus it may be possible to make early selections in a breeding program. For example, a 3-year cycle for crop improvement could involve making selections at year 2 followed by culling and open-pollination of selected plants in year 3.


JRC sold its specialty paper business in 1991; however, they continued to investigate Hesperaloe fibers for other applications. By the end of 1992 it was established that (1) fibers of Hesperaloe funifera have exceptional strength and texture; (2) high biomass yields can be obtained at year 5, and (3) the crop does have low water requirements consistent with CAM, which should make it appealing to potential growers. At this same time the newly established USDA Alternative Agriculture Research and Commercialization (AARC) Center issued its first call for proposals. The University of Arizona, JRC, and Sundance Farms of Coolidge Arizona submitted a proposal to the AARC Center in 1993 to establish a demonstration farm and conduct pulping and paper-making on a pilot scale. AARC Center funds would have gone into the demonstration farm and into development of appropriate agricultural machinery. JRC was proposing to manufacture a line of premium tissue and towel products which they had developed using Hesperaloe fibers (Reeves et al. 1994). Softness, absorbency, and thickness are usually inversely related to fiber strength. However, sanitary papers made with a blend of Hesperaloe fibers have both greater strength and softness (low coarseness), and also greater bulk, flexibility, and improved absorbency (Reeves et al. 1994).

This proposal was reviewed favorably by the AARC Center. However, during subsequent negotiations, JRC decided not to commit to a new major business venture at that time. The primary reason was the price of pulp, which hit an industry-wide low in the 4th quarter of 1993 (Fig. 5); major pulp manufacturers were not making a profit with softwood kraft pulp selling for less than $500/t.

Based on preliminary statements of interest from several paper makers specializing in higher-value products, the AARC Center agreed to fund the University of Arizona for a one-year period (Jan.-Dec. 1994) to develop a consortium of private-sector parties to take over the commercialization of Hesperaloe. The AARC Center stipulated that this consortium should consist of no fewer than five companies in the pulp and paper industry. A consortium was successfully organized with Arbokem, Inc., a specialty pulp manufacturer, as the lead organization. Six paper makers initially agreed to participate in the consortium.

The consortium submitted a second proposal to the AARC Center in August 1994, but this proposal was not accepted. The new AARC Center Board of Directors, under pressure for early commercial successes, felt that commercialization of Hesperaloe was too much of a long-term project (B. Crain pers. commun.). Following the AARC Center's rejection of this proposal, several of the paper makers involved decided to terminate their participation in the consortium.

We are currently working with a single pulp manufacturer (Arbokem, Inc.), and a single paper maker; the latter company prefers not to disclose its participation at this time. Under the direction of Wayne Coates of the University of Arizona, enough Hesperaloe funifera biomass will be harvested to ship approximately 2-3 t of decorticated fibers to Arbokem for pilot-scale pulping trials. The pulp will then be used for pilot-scale paper-making trials. If the results of these pilot-scale trials are satisfactory, a decision to move forward with a commercialization program could be made.

The major limitation on commercialization at this time is seed supply. We have been harvesting and cleaning seed from our experimental plots and have accumulated enough seed to plant about 150 ha. The first plantings of Hesperaloe funifera on demonstration farms will have to accomplish several objectives, including further development and testing of best management practices, seed production, and biomass production. Commercial-scale fiber production, approximately 20,000 to 25,000 t bone-dry pulp per year, would require about 5000 ha of cropland. However, assuming that the production cycle for Hesperaloe entails harvests at ages 5, 8, and 10, it is difficult to devise a planting scenario that would produce the exact quantity of biomass required each year by the pulper. One potential scenario is shown in Fig. 6, which assumes a yields of 180 t FW ha-1; an initial harvest at 5 years, and a productive lifetime for the stand of just 10 years. Under this plan, full production for a single processing plant would be reached at year 8. There would be excess production in years 13, 15, and 20, and sufficient biomass production to double the pulping capacity in year 26. Many different commercialization plans could be devised, but they would all have to share 2 basic features: an initial period of demonstration-scale production (ca. 50-200 ha) for increased seed production, followed by staggered plantings at full commercial scale. Sizes of plantings, planting dates, harvest dates, and the number of years that plantings remain in production could all be varied to adjust the annual supply of raw materials.



Much progress has been made in understanding how Hesperaloe species grow and develop, in determining likely biomass yields, and in defining the likely production cycle for this crop. We have also learned much about the physiology and reproductive biology of these species. Sufficient seed has been produced to move from the experimental scale to a demonstration-farm scale. Considerable research and development work also has been carried out by the private sector, although none of this research has been published in the scientific or technical literature. Efforts have concentrated on determining the critical strength, texture, and brightness properties of paper samples made from fibers of Hesperaloe; work has also been done on pulping and development of products. There is one patent on the use of Hesperaloe fibers in tissue and towel products (Reeves et al. 1994).


The comparatively large size and perennial habit of Hesperaloe funifera have posed major problems for conducting agronomic research with this plant. The large plant size necessitates large plot sizes and thus substantial commitments of land and effort for agronomic experimentation. The slow development and long production cycle of the crop necessitate that experimental plantings be evaluated for several years. One cannot redesign and replant a study the next year based on the current year's results, as with annual crops. In fact, we are continuing to learn more about the growth, biomass production, and reproduction of H. funifera from the first studies we established. Another 3 years of observation and monitoring will be needed just to determine the optimum growth cycle (Fig. 4).

We established expanded plantings of Hesperaloe funifera at the University of Arizona Maricopa Agricultural Center this past year. We have encountered many problems, including production of vigorous transplants, adequate preparation of beds, effective transplanting with available equipment, and weed control. The slow growth rate of Hesperaloe during its first year following transplanting (Fig. 1, 2) make this crop a very poor competitor with weeds. All weed control during stand establishment will have to be accomplished using cultivation and chemical control. William McCloskey and Ramon Cinco-Castro have conducted preliminary work on herbicide evaluations. All pre-emergent herbicides tested proved to be phytotoxic to Hesperaloe funifera, but the plant did appear to tolerate several post-emergent herbicides.

Over the near term, lack of adequate supplies of seed and the long period of stand establishment (5 years to first harvest) will limit how rapidly commercialization of Hesperaloe can proceed. The private sector, like the AARC Center, may balk at investing in a project with a comparatively long payback period. Growers, processors, and paper makers will have to find a mutually acceptable way to share the risks and establishment costs.


In general the prospects for increased utilization of nonwood fibers, both worldwide and in the United States, are favorable (Atchison 1995; Hurter 1991; Judt 1993). Annual crops such as kenaf, hemp, roselle, and others are likely to play an important role in meeting the growing worldwide demand for paper products. Novel perennial fiber crops will have to have special properties in order to succeed commercially. The unusual combination of improved strength and texture provided by Hesperaloe fibers (Reeves et al. 1994) should guarantee a market if an adequate supply can be developed at a reasonable cost.

Basic research on Hesperaloe, including continued studies on biomass production, agricultural engineering, water use, and weed control, is currently being supported by the USDA Cooperative State Research, Education, and Extension Services (CSREES). Lack of longer-term AARC Center support for commercialization is probably not a major obstacle to successful commercialization. Potential support from the AARC Center provided a good incentive for many private-sector parties at a time when market pulp prices were very low (Fig. 5). With an economically healthier pulp and paper industry, federal support for commercialization is less critical.


*Many coworkers at the University of Arizona have contributed to the research and development efforts on Hesperaloe. I thank Damian Ravetta, Abed Anouti, Susan Schuck, Wayne Coates, John Nelson, Bill McCloskey, and Ramon Cinco-Castro. Financial support was received from the James River Corporation, the USDA AARC Center (93AARC20030), and the USDA CSREES (94-COOP-1-0036). Mr. Alfred Wong of Arbokem, Inc., has played a major role in defining the pulping and paper-making characteristics of Hesperaloe fibers, as well as in maintaining private-sector involvement in this project.
Table 1. Prices for pulps, 1991 (from Judt 1993).

Pulp type US $/Mg
Well-cleaned, bleached fiber pulps (abaca, hemp, flax, sisal) $1800-$2400
Not so-well cleaned, unbleached fiber pulps (hemp, flax, kenaf, cotton lint) $1200-$1800
Special softwood pulps $750-$850
Normal softwood pulps $550-$750
Normal hardwood pulps $450-$550

Table 2. Dimensions of paper-making fibers. Data for Hesperaloe funifera are from McLaughlin and Schuck (1992) for accession A-184; data for other species are from Hurter (1991).

Fiber Mean length (mm) Mean width (µm) L/W Ratio
Abaca 6.00 20 300:1
Hesperaloe funifera 3.20 15 215:1
Sisal 3.03 17 180:1
Cotton linters 3.50 21 165:1
Kenaf bast fiber 2.74 20 135:1
Wheat straw 1.48 13 110:1
Softwoods 3.00 30 100:1
Jute 2.00 20 100:1
Hardwoods 1.25 25 50:1

Table 3. Water applications, biomass production, and water-use efficiency (WUE) in Hesperaloe funifera, from planting until first harvest at year 5. MD = medium density (13000 plants ha-1); HD = high density (27000 plants ha-1). Dry weights are estimated as 0.325 x fresh weight.

Water applied (cm) Biomass production (Mg DW ha-1 y-1) WUE (Mg DW cm-1)
1 30 42 0.3 0.5 <.01 <.01
2 37 46 4.9 9.8 0.13 0.21
3 46 74 9.2 12.5 0.20 0.17
4 69 114 11.7 15.5 0.17 0.13
5 41 83 18.3 24.2 0.44 0.29
Mean 44.6 71.8 8.8 12.5 0.20 0.17

Fig. 1. Cumulative biomass production and daily growth rates in fresh weight measured over a 5-year period in a sample of 20 plants of Hesperaloe funifera.

Fig. 2. Biomass production of Hesperaloe funifera at three stand densities measured over a 7-year period. Plots were completely harvested at the end of 1992 after 5 years of growth.

Fig. 3. Relationship between plant size and stand density in Hesperaloe funifera.

Fig. 4. Potential crop cycle for Hesperaloe funifera.

Fig. 5. Prices of market pulps, 1989-1995; adapted from Presley (1995).

Fig. 6. An agricultural production plan for commercialization of Hesperaloe. Initial demonstration farm plantings will provide the seed for full commercial scale-up. A sequence of plantings will be required to provide a uniform supply of materials to a pulping facility. Each box represented a planting lasting 10 years and producing 18 t pulp ha-1 (from 180 t FW ha-1) at years 5, 8, and 10.

Last update August 22, 1997 aw