Index | Search | Home | Table of Contents

Dierig, D.A., T.A. Coffelt, F.S. Nakayama, and A.E. Thompson. 1996. Lesquerella and vernonia: oilseeds for arid lands. p. 347-354. In: J. Janick (ed.), Progress in new crops. ASHS Press, Alexandria, VA.

Lesquerella and Vernonia: Oilseeds for Arid Lands

David A. Dierig, Terry A. Coffelt, Francis S. Nakayama, and Anson E. Thompson


  1. LESQUERELLA
    1. Classification and Domestication
    2. Biology and Adaptation
    3. Germplasm Evaluation and Breeding
    4. Cultivation
  2. VERNONIA
    1. Classification and Domestication
    2. Biology and Adaptation
    3. Germplasm Evaluation and Breeding
    4. Cultivation
  3. CONCLUSION
  4. REFERENCES
  5. Table 1
  6. Table 2
  7. Table 3
  8. Table 4
  9. Table 5

Two oilseed crops, lesquerella and vernonia, with excellent potential for commercialization are presently being developed in the arid Southwest. Lesquerella fendleri (Gray) S. Wats., Brassicaceae, is a source of hydroxy fatty acids, for use in lubricants, plastics, protective coatings, surfactants, and cosmetics. In addition to the seed-oil, the seed-coat gums and seed-meal would also add considerable value. Lesquerella has progressed from small plot research to successful cultivation on larger farm areas. Vernonia galamensis is a source of epoxy fatty acids, for use in formulations of protective coatings and paints with low volatile organic compounds (VOC). This is an important consideration in order for industry to comply with environmental laws that limit VOC emissions. Other uses for vernonia include thermoset resins and coatings, polymer blends, dibasic acids, adhesives, and epoxy composite materials. Breeding and cultural management improvements have been made on these crops and both are suited for production in arid climates. The successful adoption of these new oilseeds can provide high economic return to growers, processors, and industrial users compared to the expenditure for the necessary agricultural research.

LESQUERELLA

Classification and Domestication

All species of Lesquerella are native to North and South America with 83 of 95 known species indigenous to North America (Rollins 1993). Many are concentrated in the southwestern United States and Mexico. Interest in developing lesquerella began in the 1960s when three new hydroxy fatty acids (HFA), lesquerolic, densipolic, and auricolic, were discovered in Lesquerella species (Smith et al. 1961; Gentry and Barclay 1962; Mikolajczak et al. 1962; Kleiman et al. 1972). Most species occurring in the western U.S. contain lesquerolic acid (C20:1-OH) as the primary fatty acid in its seed-oil (Barclay et al. 1962; Dierig et al. 1996). This HFA is very similar to ricinoleic acid, the primary HFA of castor. Most species of the eastern U.S. contain densipolic acid (C18:2-OH) as their primary seed-oil component (Barclay et al. 1962). Only two species are known to contain auricolic acid (C20:2-OH) (Hayes et al. 1995). A closely related genus, Physaria, also contains significant quantities of lesquerolic acid. All three HFA have many potential markets (Roetheli et al. 1991). Commercialization interests are presently focused on L. fendleri because of its growth and yield characteristics (Thompson et al. 1989).

Biology and Adaptation

L. fendleri is a perennial with densely pubescent leaves. The leaf trichomes, fused half or more in length, and the lack of pubescence on the capsule are used as a distinguishing characteristic from other Lesquerella species (Rollins and Shaw 1973). It is found throughout the southwestern U.S. including Arizona, Colorado, New Mexico, Texas, Utah, and northern Mexico. Many of these native populations are found on limestone outcroppings. The chromosome number is 2n = 12, although populations of 2n = 14 and 2n = 24 have been identified (Rollins and Shaw 1973). Seeds are contained in capsules or pods, referred to as siliques, that have a globose shape. Each capsule normally contains between 6 and 25 seeds. Seed weight varies between 0.5 and 1.2 g/1000.

In the field, bees and other insects are observed frequently visiting the flowers. We have estimated outcrossing rates in this species to be between 86% and 89%. Self-incompatibility occurs and is assumed to be a sporophytic multiple allele system as generally found in the Brassicaceae. Self-incompatibility limits production of selfed seed. Male sterility also occurs frequently and is characterized by varying degrees of reduced (vestigial) anthers without pollen production. The genetic inheritance of these traits is being investigated at our laboratory. Interspecific hybridization is not common among native populations found in the western U.S. and difficult to obtain from controlled greenhouse crossing. We have produced punitive hybrids between L. fendleri x L. lindheimeri and L. fendleri x L. gracilis. These three species have the same chromosome number. However, based on general affinities, Rollins and Shaw (1973) place L. fendleri into a different phylogenetic group than L. gracilis and L. lindheimeri. We have not determined if the hybrids are fertile.

Germplasm Evaluation and Breeding

Lesquerella fendleri is a very productive species in regards to seed yield. Unselected accessions yielded between 1000 and 1350 kg/ha of seed in replicated field plots compared to yields in a synthetic population of 1880 kg/ha (Thompson and Dierig 1993). A 25% decrease in seed yield was observed and attributed to inbreeding depression in half-sib family progenies. There are no major barriers to domestication of this plant. The goals of our breeding program are relatively simple and deal with increasing various yield related components that remain stable over different environments.

This species possesses a great amount of variability. Payson (1922) referred to L. fendleri as a remarkably polymorphic species. Even within populations of native stands, there are large amounts of variability. Originally, germplasm available to our breeding program were from relatively few accessions, and did not represent the geographical range of the species (Thompson and Dierig 1993). In 1993, we initiated the collection and evaluation of as many accessions of this species as possible throughout the U.S., and incorporated useful traits into elite lines.

In the future, other species of Lesquerella and/or Physaria may also be domesticated as a source of one of the other HFA or another source of lesquerolic acid for other geographic locations where L. fendleri is not adapted. Over the past three years, we collected germplasm of these genera throughout the U.S. This includes 216 accessions of 31 different species of Lesquerella and three accessions of two species of Physaria. Individual numbers of accessions and species, along with the state of their origin, are listed in Table 1. Many of these have not been previously included in the National Plant Germplasm System. Most accessions have only small amounts of seed from the original collection. Over the 1994-96 growing seasons, we increased seed amounts and evaluated the L. fendleri accessions by planting them in caged field plots supplied with honey bees to prevent cross-pollination between accessions.

Increases in oil and fatty acid contents have been achieved through recurrent selection populations. In 1993, seed from 20 open-pollinated plants selected for high oil content, high lesquerolic acid content, and a third population selected for both high oil and lesquerolic acid yield were field grown. The oil content in 1993 ranged from 22% to 25% in the high oil population. Lesquerolic acid content in 1993 for the high lesquerolic population ranged from 58% to 64%. The high lesquerolic yield population ranged in an index for lesquerolic acid yield (oil content multiplied by lesquerolic acid content) between 15 to 18. These are the base populations which are recurrently selected each year for the three seed-oil traits. Means for subsequent years are expected to be lower since this is an open pollinated population. The ranges of values in following years should be wider because of new genetic recombinations being produced. Seeds with high values are then selected for the next generation. Means and ranges of values for oil characteristics in 1994 and 1995 populations are shown in Table 2. An unselected population was used both years for comparison.

The comparison of population means between years shows modest increases for oil and lesquerolic acid content. However, differences are greater when comparing check and recurrent populations in the same year. Environmental influences are more prevalent between years. An important result of these populations is the extremes of the ranges of values have increased each year.

Within these populations, we are concentrating on selecting for other desirable characteristics such as autofertility and upright growth habit. Autofertility is important since it would eliminate the expense of supplying bees in the field to improve seed-set. Variability for this trait is present within the available germplasm. Upright growth habits would improve combine harvesting efficiency. Yellow seed coat lines are also being developed with the prospect of use in the cosmetic industry. Oil from seed with normal orange-brown seed-coats has a pigmentation that must be removed for their applications. The yellow seed coats may have less or no pigmentation and could eliminate the extra processing step.

Cultivation

Although Lesquerella fendleri is a perennial plant, it is cultivated as a winter annual. Lesquerella does best in production on well drained soils. In the southwestern U.S., planting is done in Oct. by direct seeding with a broadcast planter such as a Brillion used to sow alfalfa. Sowing the seed by aircraft has also worked successfully. The most critical element for a good plant stand is keeping a moist soil surface to allow germination. Emergence occurs 8 to 14 days after planting. Seedlings remain very small until early Feb. when temperatures start to increase. Rapid vegetative development takes place and full soil cover is attained within two months. Seeding rates should be between 6 and 8 kg/ha.

Lesquerella has been successfully cultivated on both raised and flat fields. However, salts may be better managed on raised beds. Flowering begins in Feb. and seeds develop and mature between Mar. and late May. Nitrogen fertilizer applied at rates of 60 to 120 kg/ha resulted in increased dry matter and seed yields (Nelson et al. 1996). A good strategy for irrigation scheduling of lesquerella in southern Arizona is to irrigate about once every 15 to 20 days starting in late Feb. through mid-Apr., then once every 10 days between late Apr. through May.

Irrigation is stopped in mid-May and the plant is allowed to dry until seed moisture has reached about 12% before harvesting. Plants should be harvested from mid to late June to avoid rainstorms. Conventional combines with suitable sieves are used for seed harvesting. Seed losses can be as little as 5% with properly equipped and operating combines. Seed yields of 1800 kg/ha have been obtained in breeding test plots. Large scale field trials have yielded about 900 kg/ha.

Weed control has been a problem with lesquerella production fields. Special use permits have been applied for using Treflan and Goal in Arizona. Permits will be obtained for Fusilade, Balan, and paraquat.

VERNONIA

Classification and Domestication

Vernonia galamensis (Cass.) Less., Asteraceae is an annual native to Africa. Taxa include six subspecies. One subspecies, galamensis, is divided into four botanical varieties (Gilbert 1986). Literature sometimes still refers to this species as V. pauciflora (Willd.) Less. Gilbert (1986) revised the taxonomic treatment and included three taxa that had not previously been described. Eastern Africa appears to be the center of diversity.

Development of an epoxy fatty acid source from plants began in the late 1950s to mid-1960s when Vernonia anthelmintica (L.)Willd., native to India, was seen as a possible candidate. Vernolic acid in this species was discovered and isolated by Gunstone (1954). Excessive seed shattering prevented further development of this species. Collections of V. galamensis and subsequent evaluation (Carlson et al. 1981; Thompson et al. 1994b, c) indicated that this species is substantially better because of the quantity and quality of the seed-oil and better seed retention. The plant is native to equatorial Africa. To develop this crop for temperate zones, the short day-length flowering requirement needed to be altered (Dierig and Thompson 1993).

Biology and Adaptation

V. galamensis is an annual, ranging in plant height from 0.2 m to 5.0 m depending on the subspecies and the geographic location (Perdue et al. 1986; Thompson et al. 1994a). Plantings in cultivated field plots in the U.S. range in plant height between 0.2 and 2.0 m. The two centers of diversity are Kenya and northern Tanzania with only one botanical variety occurring outside eastern Africa. V. galamensis differs from other annual species of Vernonia in leaf form, and /or pappus, and involucre form and size.

The seed head (capitula) contains hermaphroditic, protandrous florets. The range in number of florets per capitula, also representing the number of potential seeds per capitula, is between 50 and 150. The pistil is completely surrounded by an anther sheath, which dehisces as the stigma is emerging through. As the stigma opens and becomes receptive, the under side is covered with pollen by rubbing past this sheath. However, self-incompatibility is prevalent. The chromosome number of all subspecies is 2n = 18. Subspecies readily hybridize among themselves (Thompson et al. 1994a).

Although the arrangement of the flower structure makes controlled crossing and selfed seed production difficult, we have been able to accomplish crossing with a washing process. Central florets are chosen from the seed head. The anther sheath is gently removed from each floret. The remaining stigmata are washed with distilled water from a water bottle. Stigmata at this stage of development are still in an unforked position. We manually separate the lobes of the stigma approximately three-fourths to the base. Stigmata are then washed again and dried after one to three minutes. They can then be pollinated and bagged.

Since V. galamensis is native to equatorial areas, flowering in most of the germplasm is controlled by short day-lengths. As a result, most of the native germplasm cannot be grown in the U.S. for seed production. Plants will remain vegetative during the entire growing season. Plants begin to flower as days become shorter, which is also accompanied by colder temperatures, and senescence occurs. Day-neutral germplasm in one of the accessions, V. galamensis ssp. galamensis var. petitiana, flowers and produces seed during the normal growing season, but lacks other desirable characteristics. Transfer of the day-neutral flowering trait from this variety into var. ethiopica and other germplasm resulted in more desirable plant growth characteristics. These hybrids have been evaluated and selections made over the past four years at various locations across the U.S.

Germplasm Evaluation and Breeding

Short-day flowering response of Vernonia galamensis was the major barrier to overcome for successful seed production in the U.S. Original collections from Africa were not suited for the continental U.S. because of the flowering day length response. The growing season of vernonia is during the long-days of summer months in the U.S. when plants will not flower. An accession of var. petitiana from Kenya was found to be day-neutral. However, these accessions had fewer seeds per flower head, smaller seed weights, and more seed shattering than other germplasm. The day-neutral characteristic was introduced into the working germplasm at ARS-USWCL by hybridizing with short-day flowering plants. Fertile hybrids were retrieved with day-neutral flowering. Flowering response of plants from the two parents and three hybrid lines is listed in Table 3. Daylength was controlled by placing a dark cloth over container grown plants in the greenhouse at Phoenix, Arizona during Aug. and Sept. under natural long-day conditions. The plants under the short-day treatment were kept in the dark for 14 h and then exposed to daylight for the following 10 h. The long-day treatment received 14 h of daylight and 10 h of dark. Plants were six weeks-old when treatments began. The experiment continued until flower-buds were apparent. The parent var. petitiana flowered under both treatments, var. ethiopica flowered only under short-day treatments, and hybrids flowered under both long and short-days (Table 3).

Petitiana is not strictly day-neutral since flowering increases under short-days compared to long-day treatment. The hybrids followed the same trend since flowering significantly decreased under long-days, and when compared to petitiana under this treatment, flowering was less. Our breeding objective now is to increase the quantity of flower heads per plant and maintain other characteristics of the short-day germplasm. These include larger flower-head size, which corresponds to the number of seeds per flower-head, increased seed weight, seed retention, and removing seed dormancy.

In 1994, seeds were planted at five locations in the U.S., and one in Argentina to evaluate oil content, seed weight, and seed yields. Eight hybrid lines and one parent line petitiana were compared. The U.S. locations included Phoenix, Arizona; Fort Stockton, Texas; Medford, Oregon; Columbia, Missouri; Petersburg, Virginia; and Salta, Argentina. The results are summarized by line across the six locations in Table 4 and combined lines results from each location in Table 5. Severe white fly and disease pressure at the Phoenix location resulted in only flowering data being collected at this location. Results showed that location affected all traits with Virginia and Missouri higher in yield, and Argentina higher in 1000 seed-weight and total oil content. Line 29E-OR2-14 and 66C-1-9 had higher yields; 15D-10-12, 29E-OR2-14 and 66C-1-9 had larger seed-weights; and 29E-OR2-14 had higher oil content (Table 4). Significant location x entry interactions for number of plants flowering at 81 days after planting, yield and total oil content were present indicating variability in stability for these traits among the populations. Lower yields, higher oil contents, and 1000 seed weights in Argentina may be due to the harvest method of picking ripe seed instead of combining plants.

Cultivation

Little information is available on cultural management of vernonia since only small research plots have been grown in the U.S. Larger production fields have been grown in Zimbabwe, and other equatorial countries where plants are not limited by long day growing season. Seed yields from 1345 kg/ha in 1985 to 2494 kg/ha in 1987 has been reported by Perdue (pers. commun.).

A planting density study at Maricopa, Arizona was made in 1994 with hybrid lines by varying plant spacing within rows, which were 1 m apart. Plant populations were established at 15,000, 30,000, and 60,000 plants/ha using spacings of 15, 30, and 60 cm, respectively in a randomized complete block design with four replications. Although plants flowered earlier in the 0.60 m spacing, these plants appeared to be the poorest performers toward the end of the season. As plants mature, the stems become brittle. There is an advantage to space plants closer to provide support for other plants. The closer spacing may also force flowering on the top and outside canopy, achieving better plant architecture for harvesting.

CONCLUSION

Significant improvements in lesquerella have been made. Oil and lesquerolic acid contents have increased in recurrent selection populations. These lines should be available for public release in the near future. New accessions of L. fendleri are available for evaluation and breeding. Other species of Lesquerella occurring in the U.S. which have been collected also have commercial potential. These new accessions with the available passport data are being entered into the National Plant Germplasm System. Many of these are becoming increasingly rare and are candidates for endangered species lists.

Hybrid lines of Vernonia galamensis have been identified that will flower under both short and long-days. Further improvements are still necessary for commercialization. Backcross populations have successfully increased the number and size of flower-heads per plant. Comparison of F1 plants to hybrids from later generations for flower head and seed size indicates that some inbreeding depression occurs. High yielding lines adapted to various locations should soon be available.

REFERENCES


Table 1. Species of Lesquerella and Physaria collected from various states from 1993 to 1995.

Species State No. populations
L. arizonica Arizona 3
L. angustifolia Oklahoma 3
L. argyraea Texas 16
L. auriculata Oklahoma 1
L. cinerea Arizona 5
L. densiflora Texas 2
L. densipilia Tennesse 2
L. douglasii Washington 7
L. fendleri Arizona 23
L. fendleri New Mexico 22
L. fendleri Texas 41
L. gracilis var. nutt. Oklahoma 4
L. globosa Tennesse 2
L. gordonii Arizona 7
L. gordonii New Mexico 2
L. gordonii Oklahoma 9
L. grandiflora Texas 6
L. intermedia Arizona 4
L. kaibabensis Arizona 1
L. lasiocarpa Texas 7
L. lescurii Tennesse 2
L. lindheimeri Texas 4
L. lyrata Alabama 2
L. mcvaughiana Texas 1
L. perforata Tennesse 2
L. pinetorum Arizona 1
L. purpurea Arizona 2
L. ovalifolia ssp. alba Oklahoma 3
L. ovalifolia ssp. oval Oklahoma 6
L. ovalifolia ssp. oval Arizona 10
L. ovalifolia ssp. oval Texas 7
L. recurvata Texas 7
L. sessilis Texas 2
L. stonensis Tennesse 3
L. wardii Arizona 2
P. floribunda New Mexico 1
P. newberryii Arizona 2


Table 2. Means and ranges from three recurrent populations of lesquerella harvested in 1994 and 1995. The respective populations include high oil content, high lesquerolic acid content, and high lesquerolic acid yield.

Mean (and range)
Population Year Oil (%) Lesquerolic acid (%) Lesquerolic yield (%)
High oil population
1994z 26.7 (18.0-32.0)
1995y 26.5 (14.0-36.7)
High lesquerolic population
1994x 53.3 (42.1-59.8)
1995 54.5 (42.9-64.2)
High lesquerolic yield population
1994w 14.4 (9.7-19.0)
1995 15.3 (10.8-19.8)
zControl population was 22.1% oil with a range of 19.0 to 24.4%
yControl population was 23.4% oil with a range of 20.0 to 25.6%
xControl population was 53.3% lesquerolic acid with a range of 42.1 to 59%
wControl population was 11.8% lesquerolic yield with a range of 7.9 to 15.1%


Table 3. Response of two parent lines and three hybrid lines of Vernonia galamensis subspecies galamensis to photoperiod.

Mean number of flower heads per plant
14 h day 10 h day
Line No. plants Mean Range Mean Range
Var. petitiana (day-neutral parent) 7 28.1 13-64 60.5 38-101
Var. ethiopica (short-day parent) 9 0 0 4.9 2-5
Hybrid 21J-7-10 7 6.1 1-13 54.0 26-72
Hybrid 5E-10-19 7 15.2 4-41 51.0 15-93
Hybrid 66C-1-9 7 11.7 2-34 55.1 40-72


Table 4. Mean vernonia yield, seed weight, and seed oil content of eight hybrid populations and the AO399 parent from five locations in 1994.

Line Yield
(kg/ha)
Seed wt
(g/1000)
Oil
(%)
14D-2-5 397 2.76 37.1
15D-10-12 435 3.01 38.9
29E-OR2-14 575 2.96 40.2
35A-2-9 434 2.74 38.3
35A-2-10 395 2.69 36.3
48A-10 376 2.68 38.6
66C-1-9 504 3.05 37.7
72A-1-2 432 2.65 36.5
AO399 496 2.62 38.6
LSD 115 0.13 1.4


Table 5. Mean vernonia seed yield, seed weight, and seed oil content of five locations in 1994.

Location Yield
(kg/ha)
Seed wt
(g/1000)
Oil
(%)
Virginia 820 2.82 38.1
Missouri 325 2.76 37.2
Texas 303 2.55 37.8
Oregon 645 2.53 35.2
Argentina 155 3.31 41.7
LSD 86 0.10 1.0


Last update June 16, 1997 aw