Previous PageTable Of ContentsNext Page


Felicitas Katepa-Mupondwa, Gerhard Rakow and Phil Raney

Saskatoon Research Centre, Agriculture and Agri-Food Canada, Saskatoon, SK, Canada, S7N 0X2.


Canola-quality strains of Sinapis alba have been developed, which are devoid of erucic acid, p-hydroxybenzyl and benzyl glucosinolates. These strains were developed through successive single plant selection following a cross between low erucic acid and low glucosinolate content plants. In addition, oil content has been increased (>4%) by crossing canola-quality plants with a high oil content (36-38%) line. The fatty acid profile of canola-quality strains indicates the potential to produce premium oil with very high oleic acid and low linolenic acid.

KEYWORD: erucic acid, p-hydroxybenzyl glucosinolate, canola


The rationale for developing oilseed Sinapis alba for production in western Canada has been previously presented by Rakow (1997). Oils from Brassica seed crops are characterised by high levels of long-chain monoenoic fatty acids, eicosenoic and erucic acid (Downey and Rakow, 1987). The erucic acid content in S. alba seed oil is 25-35%. Inheritance studies (Drost, personal communication) indicate that the low erucic acid trait in S. alba is highly heritable and controlled by a single gene, with partial dominance of high over low erucic acid content. Previous studies (Ecker and Yaniv, 1993; Tang et al., 1996) concluded that erucic acid content was under additive gene action, with partial dominance for the enhancing alleles. The predominant glucosinolate in S. alba is p-hydroxybenzyl glucosinolate and is present at rates of 120-150 μmole/g seed in western Canadian condiment mustard cultivars. Genetic studies (Drost, personal communication; Tang et al., 1996) indicate that the low p-hydroxybenzyl glucosinolate trait in S.alba is highly heritable and controlled by a single recessive gene.

The reduction of erucic acid and glucosinolates through breeding has been successfully employed in Brassica (Downey, 1964; Kirk and Oram, 1981; Stefansson et al., 1961). The objective of this research is to develop strains of S. alba with zero erucic acid, zero p-hydroxybenzyl glucosinolate and increased oil content.


High oil content canola-quality S. alba was developed by introgressing genes from three germplasm pools into one population: i) a low erucic acid line (BHL-926) developed at the Saskatoon Research Centre by Drs. D. Woods and K. Downey in the 1970's; ii) a low glucosinolate line developed in Poland (Kryzymanski et al. 1991); and iii) a high oil content (36-38 % oil) line developed in Sweden (Olsson, 1974). The initial cross to combine low glucosinolate and low erucic acid content was made in 1992, between a low glucosinolate plant, 92-6669 [zero p-hydroxybenzyl glucosinolate, low total alkenyl glucosinolate (7.65 μmole/g seed), high erucic acid content] and a low erucic acid content line BHL-926 (less than or equal to 1 % erucic acid, high glucosinolate) (Raney et al., 1995). In 1994, F4 progeny of this cross (<4 μmole/g seed p-hydroxybenzyl glucosinolate and < 1% erucic acid) were crossed with the high oil content line from Sweden. F5 half seed analysis indicated that variation for zero erucic acid was absent. Thus F1 lines of the cross (F4 x High oil line) were crossed with a high oleic, zero erucic acid selection from BHL-926. In 1995, selections were made among F2 plants for canola-quality plants with zero p-hydroxybenzyl glucosinolate and zero erucic acid content. Selected F2 plants were crossed to the high oil line for a second cycle of gene introgression for high oil content. F3 plants from this cross were evaluated for oil content in a replicated field trial in 1996, with two checks, the high oil line (39.5% oil) and AC Pennant (32% oil). Twelve lines with the highest oil content (35.9-36.8 %) were analyzed for fatty acid profile, and plants with the highest oleic acid content (>73%) were open pollinated to produce F4 seed. Oil and protein content of the 43 OP F4 lines were evaluated in a replicated field trial in 1997. Selections from these F4 lines were crossed with a low erucic acid content plant containing zero benzyl glucosinolate (WD97-148), which was selected from the original cross between 92-6669 (low glucosinolate) and BHL-926 (low erucic acid).


Following the initial cross between BHL-926 (low erucic) and 92-6669 (zero p-hydroxybenzyl glucosinolate) 60 S. alba plants with no p-hydroxybenzyl glucosinolate and less than 1% erucic acid were selected from 1000 F3 open pollinated progeny (Raney et al., 1995). The simple inheritance patterns of both low erucic acid and low p-hydroxybenzyl glucosinolate content, facilitates rapid genetic fixation of these traits. Variation for zero erucic acid was not found among F5 plants derived from the initial cross. Instead these plants contained low levels (>0 to 2%) of erucic acid. This observation suggests the possibility that alternative alleles for low erucic acid content may exist. The cross to introgress zero erucic acid into the selected low erucic acid content germplasm resulted in F2 selections which had zero erucic acid and zero p-hydroxybenzyl glucosinolate content (Tables 1 and 2). The observed high oleic acid and low linolenic acid content in this experimental material indicated the potential to produce premium quality canola oil from S. alba through conventional breeding methods. A concurrent selection program for reduced erucic acid content in BHL-926 revealed that it is possible to select for a fatty acid profile with very high oleic acid (>80%) and very low linolenic acid (<4%). Erucic acid is strongly negatively correlated with oleic acid, suggesting a genetic interdependence between the two fatty acids (Ecker and Yaniv, 1993; Tang, 1996).

Selecting for zero p-hydroxybenzyl glucosinolate was accompanied by increased levels of minor glucosinolates, particularly benzyl glucosinolate (Table 2). In 1997, 43 selected canola-quality plants had zero hydroxybenzyl glucosinolate, and low total glucosinolate content (6.3 μmole/g seed) however, they had elevated levels of benzyl glucosinolate (5.1 μmole/g seed). The zero benzyl glucosinolate trait was successfully introgressed from a plant which was observed to possess this characteristic in an adjunct genetic study. Two cycles of high oil gene introgression from the high oil line resulted in canola-quality germplasm with increased oil content (33.7%), exceeding the standard check variety AC Pennant by 4.2% (table 3). This increase in oil content is especially noteworthy since some genetic dilution for high oil content was expected following the cross to the low oil content zero erucic acid selection (from BHL-926) in 1995. The reduction in protein content with increasing oil content was consistent with the negative correlation reported between these traits.


Developing canola-quality S. alba strains is readily accomplished with conventional plant breeding. The fatty acid profile of canola-quality strains indicates that there is potential to produce a premium oil with very high oleic acid and low linolenic acid content.

Table 1. Fatty Acid Content* (%) of canola-quality Sinapis alba (Mean of 43 Plants), a Swedish high oil line and the cultivar AC Pennant.

Sinapis alba

Oleic Acid

Linoleic Acid

Linolenic Acid

Erucic Acid






High Oil





AC Penn.





Table 2. Content* of p-hydroxybenzyl, benzyl, and total glucosinolates (μmole/g seed) in canola-quality Sinapis alba (Mean of 43 Plants), a Swedish high oil line and the cultivar AC Pennant.

Sinapis alba



Total Glucosinolate

Canola –Q.




Hi Oil




AC Pennant




Table 3. Oil and Protein Contents of Canola-Quality Sinapis alba (Mean of 43 Lines), AC Pennant and a Swedish High Oil Content Line, Saskatoon, 1997.

Sinapis alba

Oil (%seed)

Protein (%meal)

Canola quality



Svalof high oil



AC Pennant




Strategic support for this project has been provided by the Canada-Saskatchewan Agri-Food Innovation Fund. Mr. Todd Olson is gratefully acknowledged for his technical assistance.


1. Downey, R.K. 1964. Can. J. Plant Sci. 44:295.

2. Downey, R. K., and G.F. Rakow, 1987. Rapeseed and Mustard. In Fehr, W. (Ed). Principles of cultivar development. MacMillan Publishing Co. New York.

3. Drost, W. J. 1999. Personal Communication.

4. Ecker, R. and Z. Yaniv. 1993. Euphytica 69:45-49.

5. Kirk, J.T.O and R.N. Oram. 1981. J. Austral.Inst.Agr.Sci.47:51-52.

6. Krzymanski, J., T. Ratjska, I. Byczanska and K. Krotka. 1991. Proc 8th Int. Rapeseed Cong., Saskatoon, Canada.

7. Olsson, G. 1974. Hereditas 77:197-204.

8. RaneyP., G.Rakow, and T.Olson. 1995. Proc 9th Int. Rapeseed Cong. Cambridge, UK.Vol12:416-418.

9. Rakow, G. 1997. GCIRC. 75-78.

10. Stefansson, B.R., F.W.Hougen, and R.K.Downey. 1961. Can.J.Plant Sci.41:218-219.

11. Tang, JiHong, J. Brown, J. B. Davis and J. H. Tang. 1996. Cruciferae Newsletter. 18:90-91.

Previous PageTop Of PageNext Page