Index | Search | Home | Table of Contents

Lessman, K.J. 1990. Crambe: A new industrial crop in limbo. p. 217-222. In: J. Janick and J.E. Simon (eds.), Advances in new crops. Timber Press, Portland, OR.

Crambe: A New Industrial Crop in Limbo

Koert J. Lessman

    1. Time
    2. Lack of Support
    3. Premature Commercialization
  9. Table 1
  10. Table 2
  11. Fig. 1


Crambe (Crambe abyssinica Hochst. Ex. R. E. Fries; Brassicaceae = Cruciferae) has been suggested as a promising new oilseed crop for the United States (USDA 1962; Downey 1971; Nieschlag and Wolff 1971; Papathanasiou and Lessman 1966; Papathanasiou et al., 1966; Whitely and Rinn 1963). Interest in crambe lies in the usefulness of its seed oil, the erucic acid present in its seed oil, and in the by-product seed meal residual after oil extraction (Downey 1971; Nieschlag and Wolff 1971; White 1966). Crambe is one of the richest known sources of erucic acid (cis-13-docosenoic), which makes up 55 to 60% of the seed oil glycerides (Lessman and Berry 1967; Mikolajczak et al. 1961).

Traditionally, U.S. companies requiring erucic acid and oil containing erucic acid have been dependent upon rapeseed-growing countries. However, in most of these countries the trend is toward development of rapeseed cultivars low in erucic acid for improved nutritional quality of the oil as a food (Tallent 1972). This has increased the importance of crambe as a domestic source of oil high in erucic acid for industrial purposes (USDA 1972). However, crambe has not become established as a crop.

In order to assure the development of any crop, a sustained program in crop management and genetic improvement is critical. This paper summarizes the status of crambe and efforts underlying its development as an alternative crop for erucic acid oil and protein feed supplement.


Seed stocks of C. abyssinica were first introduced into the United States from Europe in the 1940s by the Connecticut Agricultural Experiment Station (White and Higgins 1966). Evaluation of a number of strains of crambe began in earnest in 1958 and 1962 in Texas and Indiana, respectively. Other states that have evaluated crambe include Iowa, Kansas, Minnesota, Missouri, Montana, Nebraska, North Carolina, North Dakota, Oregon, Pennsylvania, South Dakota and Wyoming (Papathanasiou and Lessman 1966). Since about 1932, crambe has been evaluated in many areas of the world, including Canada, Denmark, Germany, Lithuania, Poland, Russia, Sweden, and Venezuela (White and Higgins 1966).

Crambe is an erect annual herb with numerous branches, growing to a height of 60 to 90 cm and maturing in about 90 days (Whitely and Rinn 1963). It produces a great number of seeds borne singly at or near the terminus of the branches. Seeds weigh 7.0 to 7.5 g/1000, with a hull content of 14 to 20% by weight when grown in Canada (McGregor et al. 1961) and 25 to 40% in the United States (Earle et al. 1966). The single-seeded fruits are spherical. The pod or hull remains on the seed at harvest and is considered a part of the harvested product (Papathanasiou and Lessman 1966). The leaves of crambe are large, oval-shaped, and smooth; flowers are very small, white, and numerous. Flowering is indeterminate, but the early formed pods usually adhere until later ones mature.


The United States has imported an average of 753 thousand metric tons of vegetable oil per year from 1981 to 1983. These have been mainly coconut, palm, palm kernel, olive, castor, and rape (Agricultural Statistics 1984). Almost all fatty acids in domestic seed oils contain 12 to 18 carbon atoms. In contrast, the seed oil of rapeseed (Brassica napus and B. campestris) and crambe contains comparatively large amounts of a erucic acid (cis-13-docosenoic), a fatty acid with 22 carbon atoms:
The composition of the triglyceride, unsaturated oil from crambe seed resembles that from rapeseed, but contains higher levels of erucic acid, 55 to 60% in crambe as compared to 30 to 45% in rapeseed. As harvested, the crambe seed-plus-hull contains 26 to 38% oil, with 32% being about average (Earle et al. 1966). The hull makes up about 30% of the harvested product. Dehulled crambe seed has an oil content of 33 to 54% and a protein content of 30 to 50% (Earle et al. 1966; McGregor et al. 1961). Refined crambe seed oil may be used as is or erucic acid may be extracted from the oil and employed for the synthesis of certain derivatives, such as erucamid, brassylic acid, and pelargonic acid.


Defatted crambe seed meal has value as a supplement in livestock and poultry feeds because of its high protein content and well balanced amino acid content (McGhee et al. 1965; Hesketh et al. 1963; White 1966). Analyses of whole crambe seed, dehulled seed, and dehulled-defatted seed meal are presented in Table 1; amino acid composition of toasted crambe meal is presented in Table 2.

Like most other members of the Brassicaceae (Cruciferae), the seed meal of crambe contains glucosinolates that are associated with unpalatability and goitrogenicity. When the meal is moistened, the glucosinolates are readily hydrolyzed to isothiocyanates by enzymes normally present in crushed or ground seed meal (Hesketh et al. 1963; McGhee et al. 1965; Van Etten et al. 1965, 1969). Isothiocyanates, other enzymatically formed products, or glucosinolates themselves may impart unpalatability or toxicity or both to seed meals.

Kirk and coworkers (1966) found that epi-progoitrin was destroyed in defatted crambe meal by an ammonia-heat treatment. Mustakas and coworkers (1968) improved palatability and reduced toxicity of crambe seed meal by a soda ash process; a process that gained commercial acceptance. The process was satisfactory for yielding a nontoxic protein feed supplement for ruminant animals, but it was not completely nontoxic to nonruminants. By deactivation of the enzyme through moist heat treatment of the whole seed, the glucosinolates are maintained intact during and after the oil extraction process. Meals processed this way are excellent protein supplements in beef cattle rations (Perry et al. 1979; Van Etten et al. 1977). Crambe meal is approved by the FDA for use in such cattle rations. Medeiros et al. (1 978) demonstrated that thioglucosidase was destroyed in intact, moist (14 to 16%) crambe seeds by exposure (38 sec) to microwave treatment. Lessman and Kirleis (1979) also found that thioglucosidase could be inactivated in intact crambe seeds by microwave treatment. Recently, it has been shown that gamma irradiation (50M rad) will inactivate thioglucosidase in crambe seeds, rapeseed and seeds of white mustard (Lessman and McCaslin 1987).


The primary plant breeding challenges to crambe improvement have been and still remain to: (a) Increase seed yield, (b) increase oil production and (c) improve protein meal. Tallent (1972) reviewed the accomplishment of plant geneticists in improving high-erucic acid oilseeds, especially in the removal of glucosinolates. Traditionally, the high-erucic acid oilseeds have been two species of rapeseed, Brassica napus and B. campestris. When used as an edible product, the erucic acid content of rapeseed oil has become a major health concern. Geneticists in Canada and Europe have been directing their attention toward developing lines of rapeseed with low erucic acid content. Rapeseed oil imported by the United States for industrial purposes, requires an oil high in erucic acid. Therefore, the limited plant breeding work in the United States has focused on crambe, since it is one of the richest known sources of erucic acid. Whether the erucic acid oil is produced for edible or industrial uses, the glucosinolates that are characteristically present in crambe seed and rapeseed unfavorably influence the use of their residual seed meals as feed.

C. abyssinica is primarily a self-pollinated plant but some natural outcrossing has been reported (Beck et al. 1975). Best results for obtaining hybridized seed is by hand emasculation and pollination (Meier and Lessman 1973a). The procedure is to remove all flowers that have opened and all pods formed previously on a particular raceme. One to six (usually two to three) unopened flowers are selected that should open within 24 hours. Sepals, petals, and stamens must be removed with care to avoid pistil injury Pollen can be brushed on a stigma with an anther. All other younger flowers on that raceme should be removed, and a small bag supported by a stake placed over emasculated flowers. If flowers are in short supply, instead of removing all of the younger flowers, only those that would open within the next two or three days are removed, as pistils do not remain receptive for more than two days. Hand emasculations and pollinations are best performed on greenhouse-grown plants, as attempts to cross plants in the field are hindered by wind-damage to the bags enclosing emasculated and pollinated flowers.

Mass selection for large and small seed size resulted in the cultivar `Prophet' from PI247310 (C. abyssinica) and 'Indy' from PI249346 (C. hispanica) but `Indy' is now believed to be an ecotype of C. abyssinica, rather than a different species (Meier and Lessman 1973a). 'Meyer' was developed by selection among progenies from the cross of C. abyssinica and C. hispanica type (Lessman 1975). All three cultivars are releases of the Purdue University Agricultural Experiment Station.

Lessman (1975) evaluated 162 lines to detect genotypic diversity for six characters in C. abyssinica. Lines used were from 2000 randomly selected out of PI247310 and PI249346, and reselected in "head-to-row" nurseries on the basis of their progeny performance. The characters evaluated were seed yield, test weight, plant height, oil percentage, days to bloom, and glucosinolate percentage. Cultivars 'Prophet', 'Meyer', and 'Indy' were used as checks. Selection of individual plants on the basis of their progeny evaluation in a "head-to-row" nursery was effective for separating variation in the original germplasm, except for glucosinolate content. Selections tended to differ in their relative average performance for most traits from one year to the next, implying a genetic x environmental interaction among entries. Although statistical differences were detected among lines for all characters except glucosinolate percentage, none of the selections produced higher yields or more oil than 'Meyer' and only four selections were shorter. Many selections required fewer days to bloom, indicating an earlier maturity than 'Meyer'. Several selections were better than 'Prophet in test weight and oil percentage.

Lessman and Meier (1972) reported a lack of adequate genetic variability among initial introductions of crambe for important agronomic traits. Initial stocks appeared quite similar. If needed variability cannot be obtained through hybridization, it will be necessary to acquire germplasm containing natural variations or induce generic variability through mutagenesis.

The inability to detect differences of statistical significance among introductions of crambe for a large number of characters may be due to the following (Meier and Lessman 1971):

Meier and Lessman (1971) estimated that the optimum harvested plot area for evaluating crambe introductions was 6.70 m2 utilizing the regression procedure of Smith (1938), or 5.35 m2, using the modified maximum curvature technique developed by Lessman and Atkins (1968). Long narrow plots, with their greater length in the direction of more variation, are considered as optimum for plot shape. Long plots consistently show significantly less variation than wide plots.

For any breeding program to be successful, genetic variability must be present. Attempts to evaluate crambe germplasm for genetic variability indicated inadequate diversity for needed agronomic improvement (Lessman and Meier 1972, Papathanasiou et al. 1966). This conclusion was based on studies with a limited number of introductions and before appropriate information concerning plot size and shape, as well as cultural practices for testing crambe, were available. Nonetheless, some plant-to-plant variability for branching tendency was noted within certain introductions.

In an attempt to obtain greater generic variability in C. abyssinica, Meier and Lessman (1973a) randomly paired plant selections from PI247310 (C. abyssinica) and PI279346 (C. hispanica) and crossed them reciprocally. The average seedset from these crosses was about 77%, no reciprocal differences were detected. Progeny from these crosses were evaluated for nine characters (Meier and Lessman 1973b). Significant differences were found among F3 and F4 entries for all characters except oil percentage. Ranking the characters according to their heritability estimates, the highest were first bloom and 95 percent bloom; intermediate were test weight, seedling emergence, plant height, and seed size; and lowest were seed yield, oil percent and maturity The superiority of 'Meyer' suggests crossing among diverse germplasm, followed by selection, as a practical step toward improvement in crambe, in addition to straight line selection within and among existing germplasm.


From 1965-1988 no more than 10,000 total acres of crambe have been produced for commercial production of oil and protein meal. Much of this production was used to gain processing data for more efficient oil extraction and refinement of protein meal as a by-product. Crambe is still a "crop-in-limbo." Even with the need for alternative options for U.S. farmers, this potential cash crop has fallen short of attaining crop status. Some major reasons for this are as follows:


Even under the best circumstances, a new crop takes a long time to develop. Soybeans were first introduced in 1765 and until 1940 production had increased to only 5 million acres. It is difficult to hasten a new crop's progress but easy to delay it. Crambe development has been no exception. Crambe research should have been sustained for twenty years. However, this time scale goes beyond the familiar for most people in government and industry and it is almost incomprehensible to the general public. Yet 3000-6000 scientist years of public support have gone into maize and soybean from 1920-1985.

Lack of Support

Sudden infusions of money at unsustainable spending levels cannot be used efficiently. Support for crambe began with approximately four scientist years in 1968 and rose to a maximum of ten during 1970-1974 then declined to less than one during 1978-1986 (Fig. 1). Such variable funding caused discomfort and undue impatience on the part of possible sponsors. A new crop research effort should be conducted in an orderly, noncrisis manner.

During the past 15 to 20 years, the level of sustained public support for agricultural research has declined. Funding has been going as short-term grants which are not appropriate and usually unavailable to a crambe research program. Industry has provided support for established crops on a continuing basis. Such support has not been forthcoming for crambe. Indeed, industry may lack the incentive to support a new crop that might reduce production and increase the cost of established crops to that industry.

Premature Commercialization

Crambe was promoted and grown commercially without enough research and development. During the attempts of the late 1960s and early 1970s it was usually the farmers that lost because management skills were not refined. Consequently, some yields were low and the product price did not allow an adequate return to the grower. More research and development should have preceded the attempted commercialization, particularly in market evaluation. A market for crambe seed was never solidified and no market ever identified for the protein meal by-product.

Disposal of by-products is an important consideration. More than 50% of crambe seeds is oil-free protein meal after oil extraction. Historically this has been given value only as a fertilizer because of the glucosinates present. Even though FDA approval allows limited use in feeder steer rations a market for this by-product was never established.


There are seven essential stages in the domestication of a wild species; (1) Germplasm Collection (2) Germplasm Evaluation, (3) Chemical and Utilization Studies, (4) Agronomic Evaluation, (5) Breeding Program, (6) Production and Processing Scale-up, and (7) Commercialization. Some progress has been made in all stages with crambe but progress is least in the very critical last two stages. It is essential that a market for crambe products be identified. Commercialization cannot be successful without a market to absorb the products at a price that will yield a profit to producers. Yet, ironically efforts in the U.S. have focused heavily on yields and efficiencies of commodities in oversupply while no voice speaks for the development of new or alternative crops. "Maintaining the status quo" is what has been happening in the U.S. Agriculture for the past 15 years. The absence of a commitment to a sustained effort in crambe has been detrimental to development.


Table 1. Analyses of crambe seed and seed meal.

Dehulled seed
Seed meany/
Crude fat33.345.60.4
Protein (N x 6.25)
Crude fiber14.03.14.6
Nitrogen free extract23.218.335.5>
zHull content equals 30%. yDehulled, defatted seed meal.

Table 2. Amino acid composition of toasted crambe meal.

Amino acid G/16 g
Glutamic acid15.5
Aspartic acid6.3

Fig. 1. Total public sector (USDA and Agricultural Experiment Station) funding invested in crambe. Source: United States Department Of Agriculture, Cooperative State Research Service, Current Research Information System, 1986.

Last update August 27, 1997 by aw