Index | Search | Home | Table of Contents

McLaughlin, S.P. 1993. Development of Hesperaloe species (Agavaceae) as new fiber crops. p. 435-442. In: J. Janick and J.E. Simon (eds.), New crops. Wiley, New York.

Development of Hesperaloe Species (Agavaceae) as New Fiber Crops

Steven P. McLaughlin

    1. Taxonomy
    2. Morphology
    1. Size-Biomass Relationships
    2. Biomass Production
    3. Regrowth
    4. Water Requirements
    5. Fertilizer Requirements
    6. Flowering
    7. Monthly Growth Rates
    8. Other Hesperaloe Species
  6. Table 1
  7. Table 2
  8. Fig. 1
  9. Fig. 2
  10. Fig. 3
  11. Fig. 4
  12. Fig. 5
  13. Fig. 6

"Hard fibers" are the bundles of fiber cells obtained by decorticating the leaves of abaca (Musa textilis Née), sisal (Agave sisalana Perrine), henequen (A. fourcroydes Lem.), and other monocots. These fiber bundles are used mostly in cordage products (rope, twine, canvas, burlap), but can be pulped for use in specialty papers (Clark 1965; Corradini 1979; da Silva and Pereira 1985), which include such products as tissue papers, filter papers, tea bags, currency papers, and security papers. The very long and thin fiber cells of these plants produce papers that are strong yet fine-textured. The crop plants that produce the hard fibers of commerce are all frost-sensitive tropical species. The objective of my research program is to develop domestic production of a cold-tolerant source of hard fibers for use in specialty papers.


Abaca and sisal pulps command premium prices in the paper industry (Clark 1965; Baker 1985). Tensile strength, tearing resistance, and bursting strength, are largely determined by fiber morphology (Horn and Setterholm 1990). The fiber cells of abaca and sisal are as long or longer yet much thinner than those of softwoods. Abaca fibers may average 6.0 mm in length and 24 µm in width, for a length-to-width ratio (L/W) of 250. Sisal fibers are closer to 3.0 mm in length with a L/W of 150.

We examined several species of Agavaceae native to the southwestern United States and northern Mexico to determine if any of these plants possessed fibers similar to those of abaca and sisal (McLaughlin and Schuck in press). An updated summary of our results is presented in Table 1. Native species of Agave, Nolina, and Dasylirion have relatively short fiber cells. Species of Yucca and Hesperaloe have much longer and narrower cells. Those of Hesperaloe species are comparable to abaca in their L/W. From this screen, we selected Hesperaloe for further study.



The genus Hesperaloe consists of three described and two or more as yet undescribed species; all are native to northern Mexico. Hesperaloe is probably most closely related to the larger genus Yucca (Smith and Smith 1970). Hesperaloe funifera (Koch) Trel. (Fig. 1A) is found at lower elevations in the east-central part of the Chihuahuan Desert in Coahuilla and Nuevo Leon. Its Spanish common name is samandoque or zamandoque. H. parviflora Torr. is the most widespread species, occurring at higher elevations in the northern Chihuahuan Desert of Texas, Chihuahua, and Coahuilla. It is widely cultivated as an ornamental plant in the southwestern United States where it is called "red yucca." H. nocturna Gentry (Fig. 1B) is a recently described species from the Sierra Madre Occidental area of northeastern Sonora; it has no Spanish or English common name. There are two recently discovered Hesperaloe forms that probably represent new species, one from the southeastern Chihuahuan Desert of San Luis Potosi and the other from southern Sonora.


Hesperaloe species are long-lived, evergreen, acaulescent plants. As in all Agavaceae, the basic module of growth is the rosette, a cluster of leaves produced from a single meristem. This meristem produces several leaves before switching from vegetative to reproductive mode. Once the flower stalk is produced, the rosette ceases to grow. In Hesperaloe (as in most Yucca) lateral or secondary rosettes are produced from the crown after the primary rosette becomes reproductive. Although an older Hesperaloe plant has the appearance of a closely packed, often grass-like clump of leaves with several flower stalks, the plant actually consists of a cluster of separate but closely-spaced rosettes. The older flowering rosettes are found at the center of the clump; younger vegetative rosettes are on the periphery of the clump.

The species of Hesperaloe differ in their leaf morphology. The leaves of H. funifera (Fig. 1A) and the undescribed species from San Luis Potosi are 1 to 2 m long, 3 to 6 cm wide toward the base, and stiffly erect. Those of H. funifera are cresent-shaped in cross section while those of the undescribed species are more strongly folded into a V-shape. Leaves of H. parviflora are less rigid, arching away from the crown, shorter (mostly <1 m), narrower (1 to 2 cm), and cresent-shaped in cross section. Leaves of H. nocturna (Fig. 1B) and the undescribed species from southern Sonora are long (1 to 2 m), very narrow (mostly <1.5 cm wide), and hemispherical in cross section. Leaves of all species bear marginal fibers. Older plants of H. funifera typically consist of 10 or fewer rosettes while those of H. nocturna and H. parviflora often have many more than 10 rosettes.

All Hesperaloe species produce relatively large flower stalks--those of H. funifera may be 3 to 4 m tall. The flowers of H. parviflora and the undescribed species from Sonora are pink to red; those of the other species are white to green.


There are no published studies on the agronomy of Hesperaloe species. Our initial trials therefore, concentrated on determining the potential biomass production of Hesperaloe funifera. This species was selected because its fibers are very long and thin and small amounts of seed were available from landscape plants growing in Tucson. While seed of H. parviflora are more readily available, its fibers are consistently shorter than those in H. funifera (McLaughlin and Schuck in press).

In the following text, standing crops and yields will be reported as fresh weights. Standing crop refers to the amount of biomass present at any particular time; yield refers to the amount of biomass obtained when the stand is harvested. Because whole, cut leaves do not readily lose moisture, it is most convenient to measure biomass as fresh weights. In addition, it is likely that leaves would be transported and pulped as fresh material. Dry matter and dry fiber contents of fresh leaves are approximately 32.5 and 10%, respectively.

Size-Biomass Relationships

H. funifera would be grown as a perennial crop (Fig. 2). Several years would be required from the time of stand establishment to first harvest. Plants harvested near ground level can regrow by (1) elongation of cut leaves (monocot leaves grow from a basal meristem), (2) production of new leaves from cut rosettes, and (3) production of new rosettes. It seems likely, therefore that cut plants will regrow to produce several subsequent harvests.

Replicated production plots for H. funifera were established at three densities: 6,800, 13,500, and 27,000/ha. Plots measured 9.75 by 30.5 m with two plots at each density level. The low-density plots consist of 8 rows of 25 plants; medium-density plots have 8 rows of 50 plants, and high-density plots have 16 rows of 50 plants. Plant spacings within the plots are: low density, 1.22 m between and within rows; medium density, 1.22 m between rows by 0.61 m within rows; and high density, 0.61 m between and within rows. Plots were established from transplants (3- to 5-month old seedlings) in March 1988; they receive irrigation and fertilization through a below-ground drip irrigation system. Soil moisture was monitored with gypsum blocks.

The key problem in monitoring growth and yield in such a perennial crop is developing nondestructive methods of biomass estimation. We have measured standing crops each year by randomly sampling 5 plants per row. On each plant, basal circumference and average length of the five longest leaves were measured. Two of the five plants in each row were harvested for fresh-weight determinations. The number of plants harvested from the production plots each year varied between 4 and 8% of the stand, depending on the density. Sampling was done February 1989 (stand age 11 months), November 1989 (20 months), and November 1990 (32 months).

The data on basal circumferences (cm), leaf length (cm), and fresh weights (g) have been used to develop size-biomass relationships. Scatter diagrams show that there is not a particularly good fit between either basal circumference (Fig. 3A) or leaf length (Fig. 3B). However, basal area (BA, in cm2) can be calculated from basal circumference (BC) as: BA = BC2/(4p). I then defined a new variable, SIZE2, as: SIZE2 = (BA)(Leaf Length)/1000. SIZE2 is proportional to the volume of the plant; it is linearly related to fresh weight (Fig. 3C). Plotted on log-log scale the relationship is linear with a very high R2 (Fig. 3D). We used the equation for the relationship shown in Fig. 3D to estimate fresh weights on a large sample of plants in the Production Study in August 1991; we also have used it to estimate fresh weights nondestructively in other studies. This equation works well for plants with a single rosette; a different size-biomass equation probably will have to be developed for plants regrowing with two or more rosettes.

Biomass Production

The development of the standing crops of Hesperaloe funifera at three densities is shown in Fig. 4. During the first growing season, accumulation of aboveground biomass is very slow as plants establish a large crown and an extensive root system. Growth in subsequent years is rapid. Estimated standing crops (leaves only) at the end of the third growing season were 22.1 46.7, and 77.4 Mg fresh weight/ha for the low-, medium-, and high-density plots, respectively.

Standing crop to date is nearly directly proportional to density. Individual plant size is inversely proportional to density but the effect so far is small. Average plant size in the high-density plots at the end of 1990 (year 3) was 2632 g/plant compared to 3535 g/plant in the low-density plots.


Half of each of the 6 plots in the production study were harvested in November 1990. Plants are vigorously regrowing from the 5- to 8-cm stubble left from the harvest, as expected (Fig. 5). Regrowth is occurring both from the harvested rosettes and from new lateral rosettes. The average number of rosettes per plant in these harvested plots is now 2.39, 1.88, and 1.56 in the low-, medium-, and high-density plots, respectively. Since the basic unit of growth is the individual rosette, the effective densities of these plots have increased to 16,000, 25,000, and 42,000/ha, respectively.

Water Requirements

To get a first approximation of this potential crop's water requirement, we examined biomass production as a function of the amount of water applied. While we have tried to balance water applications to rates of soil moisture depletion, our system for monitoring soil moisture is not sophisticated and irrigation schedules are far from optimized. Nevertheless, Hesperaloe appears to have a rather low water requirement for an arid-land crop (Table 2). In the high density plots, a total of 38.5 Mg dry weight has been produced with a total application of 238 cm of irrigation, equivalent to a water requirement of 6.2 cm/Mg dry weight of leaves. In comparison, alfalfa grown in Final Co., Arizona, requires 11.6 cm/Mg dry weight and kenaf grown in Imperial Co., California, requires 13.0 cm/Mg dry weight.

Fertilizer Requirements

Leaf nitrogen levels were measured on leaves from 6 plants at the end of the 1990 growing season. Average N content was 1.42% (dry-weight basis) and did not vary with density level or leaf age. Harvest of 77 Mg fresh weight/ha from the high-density treatment at the end of the third growing season represents a withdrawal of 358 kg N/ha; additional N is removed in the flower stalks, flower parts, capsules, and seeds. The high-density plots were fertilized with only 120 kg N/ha over the first three years. Plants clearly used considerable residual N in our plots and it is likely that growth has been limited by low N to some unknown degree in our study.

Phosphorus contents averaged 0.13% of leaf dry weights, corresponding to a removal rate of 32 kg P/ha over the first three years in the high-density plots.


Hesperaloe produces a fairly large flower stalk. In the 43 randomly selected plants that flowered in 1990, the flower stalk (excluding dehisced flowers and dispersed capsules and seeds) constituted 27% of the aboveground fresh weight. At the end of the 1990 growing season (after the third growing season), percentage of plants in flower ranged from 31% in the low-density plots to 18% in the high-density plots. At the end of August 1991 (in the fourth growing season) percentage of plants in flower in the unharvested portions of our production plots averaged 57% and did not vary among density treatments.

Monthly Growth Rates

We have been monitoring growth rates in a population of 20 Hesperaloe funifera plants on a monthly basis since February 1990. These plants are approaching the end of their second growing season. Number of leaves increases most rapidly between June and October; rates of leaf elongation are greatest during the same period. Basal circumference, however, continues to increase through December. Thus, our estimates for fresh weight, which are based on basal area and leaf length, continue to rise through December of the first year. Estimated fresh weight begins increasing rapidly by May of the second growing season (Fig. 6), consistent with our findings from the production study.

Other Hesperaloe Species

Hesperaloe nocturna has been grown in a small observation plot and several plants of this species have been harvested on a yearly basis. Fibers of this species are nearly as long as those of H. funifera; if H. nocturna can be harvested at yearly intervals rather than the projected 2-year interval for H. funifera, the former species might be a superior crop plant. In October 1989, sufficient seed of H. nocturna was collected from the wild for establishing a series of production plots. We transplanted replicated plots of this species at densities of 6,800, 10,000, 13,500, and 20,000/ha in October 1990. Initial growth appears to be good; these plots will be monitored for biomass production as they mature.


It is difficult at this initial stage of research and development on Hesperaloe funifera to evaluate this plant's potential as a new crop. We estimate that stands of Hesperaloe will need to produce annual yields of 30 to 45 Mg fresh weight/ha (ca. 10 to 15 Mg dry weight) to produce biomass at $40 to $60/Mg fresh weight (N.G. Wright and S.P. McLaughlin unpublished analyses). This would correspond to a feedstock cost of $400 to $600/Mg dry fiber. Leaf standing crop at the end of the third year was approximately 78 Mg fresh weight/ha in the high-density plots, corresponding to an annual productivity of 26 Mg fresh weight/ha. First-year aboveground growth rates, however, were low. We estimate that the standing crop at the end of the fourth year will be between 120 and 140 Mg fresh weight/ha, corresponding to an annual productivity of 30 to 35 Mg fresh weight/ha. The rate of regrowth during the first year after harvest appears to be much higher than the initial growth rate during the first year after stand establishment.

The literature on cultivated Agave species indicates the magnitude of yields that might be possible from perennial rosette plants in the Agavaceae. Agave species are more succulent than Hesperaloe species (12% dry matter vs. 32.5% dry matter, respectively, in the leaves) so that direct comparisons of fresh weights are not meaningful. Nobel (1988) reported the following dry-weight standing crops for 7-year-old stands of cultivated Agave species: A. sisalana, 70 Mg/ha; A. fourcroydes, 80 Mg/ha; and A. tequilana, 90 Mg/ha. These would correspond to annual yields of 10 to 13 Mg dry weight/ha similar to the targets I have set for Hesperaloe. Nobel (1991) reported annual yields of 38 to 42 Mg/ha in special plantings of A. mapisaga and A. salmiana at Tequexquinahuac, Mexico. These yields are comparable to the maximum yields observed in experiment station plants of C3 and C4 crops.

Initial predictions that 4 years would be required to reach a first harvest and that subsequent harvests might be made every three years thereafter were proved wrong. Stands should reach a harvestable standing crop after three years and the amount of regrowth may be enough that subsequent harvests could be obtained every two years.

Hesperaloe funifera, like Agave, is a CAM (crassulacean acid metabolism) plant (Damian Ravetta unpublished data). The photosynthetic pathways of other Hesperaloe species have not yet been determined, but it is of interest to note that in Yucca, Hesperaloe's closest relative, there are both CAM and C, species (Kemp and Gardetto 1982). The very low water requirement of H. funifera is consistent with CAM photosynthesis (Nobel 1991). High water-use efficiency, and hence low water requirement, is an important criterion for potential new crops for arid regions (McLaughlin 1985).

There appears to be a trade-off between flower-stalk production and leaf production in H. funifera, i.e., investment of photosynthetically-fixed carbon into flower stalks may decrease the amount of leaf production. Selection for plants with delayed flowering might result in greater leaf production at the first harvest. However, all rosettes must eventually terminate in a flower stalk. Delayed flowering will only result in improved leaf yields if it is accompanied by formation of an increased number of leaf primordia. Also, flowering is correlated with the production of lateral rosettes and an increased number of rosettes probably will result in larger subsequent harvests. Determining what controls the number of leaves produced by a rosette and what triggers production of flower stalks and lateral rosettes will be critical to improving yields in this species.


Table 1. Average fiber lengths, widths, and cell-wall thicknesses for five genera of Agavaceae from the southwestern United States and northern Mexico.

Genus No. species
Fiber length
Fiber width
Agave 7 1.14 27.0 5.4 16.3 42
Dasylirion 2 0.89 16.9 6.4 4.1 53
Nolina 3 0.94 19.1 4.9 9.2 49
Hesperaloe 5 3.55 14.7 3.6 7.5 241
Yucca 9 2.38 14.3 5.8 2.7 166

Table 2. Water requirement and productivity of Hesperaloe funifera in the high-density treatment (27,000/ha).

Year Water applied
Estimated aboveground
annual productivity
(Mg dry weight/ha)
Water requirement
(cm/Mg dry weight)
1988 42 0.5 ---
1989 46 10.1 4.6
1990 74 14.6 5.1
1991z 76 13.3 5.7
4-year total 238 38.5 6.2
zThrough August, 1991.

Fig. 1. Hesperaloe funifera in the Chihuahuan Desert of central Coahuilla (Fig. 1A) and H. nocturna in desert-woodland transitional vegetation in northeastern Sonora (Fig. IB).

Fig. 2. Projected stand dynamics for Hesperaloe funifera grown as perennial crop. We originally estimated that the plant would require 4 years to reach a first harvest with subsequent harvests every 3 years thereafter (solid line): it now appears that a harvestable stand can be produced in 3 years with reharvests every 2 years thereafter (dashed line).

Fig. 3. Size-biomass relationships in Hesperaloe funifera: scatter diagrams showing the relationships between fresh weight and leaf length (Fig. 3A), basal circumference (Fig. 3B), and the product of leaf length and basal circumference on a linear plot (Fig. 3C) and a log-log plot (Fig. 3D).
Fig. 4. Growth in standing crops of Hesperaloe funifera at three stand densities (6,800, 13,500, and 27,000/ha): March 1988 to August 1991. Bars show ± 1 standard error (SE); for data points without error bars, SE <1 Mg fresh weight.

Fig. 5. High-density plot of Hesperaloe funifera immediately after harvest in November 1990 (Fig. 5A) and in September 1991 (Fig. 5B), showing the rapid regrowth of the harvested stand.

Fig. 6. Estimated standing crops (Mg fresh weight/ha) and growth rates (Mg fresh weight ha-1 month-1) at 4-week intervals in a sample of 20 plants of Hesperaloe funifera.

Last update September 15, 1997 aw