Selection for increased seed mucilage content

in yellow mustard


J Philip Raney and Gerhard FW Rakow


Agriculture and Agri-Food Canada, Saskatoon Research Centre, 107 Science Place, Saskatoon, SK, Canada, S7N 0X2




High mucilage content in yellow mustard (Sinapis alba) is becoming a very important seed quality parameter for the mustard seed trade. Because of this, an evaluation of mucilage content was done on the seed harvested from a replicated 1000 entry nursery to see if variation for this seed quality trait exists in our germplasm and if selection for higher contents of mucilage would be possible. The viscosity of a water extract of whole seeds was used as an indicator of mucilage content. Selection for a high protein content was also carried out on this nursery material at the same time. Selected high mucilage, high protein lines were then placed in a three location replicated yield trial for evaluation of mucilage and protein contents, and agronomic performance. In addition, for the past two years, mucilage content has also been included as a criterion for evaluation of entries in the western Canada co-operative mustard trials. The data from the nursery and the multi-location trials are discussed.


KEYWORDS: recurrent selection, viscosity, Sinapis alba




Yellow mustard mucilage contributes to the consistency of prepared mustard and has considerable potential as a food gum additive, which creates additional markets for the commodity (Weber et al. 1974, Siddiqui et al. 1986, Aguilar and Ziegler 1990, Cui et al. 1994, Gerhards and Walker 1997). Some cultivars of yellow mustard contain more mucilage than others. In particular, the cultivar Viscount is considered to be high in mucilage, whereas Ochre is low. At the request of the Canadian Mustard Association the Saskatoon Research Centre began selecting within a yellow mustard breeding population for high contents of mucilage.  Additionally, mucilage content was added to the list of traits evaluated in the 1997 and 1998 western Canada mustard co-operative trials. Because mucilage exhibits different rheological properties depending whether it is extracted from intact seeds or from mustard preparations (Gerhards and Walker 1997), extraction from intact seeds was chosen as the method of isolation. Mucilage content was measured as a function of the viscosity of the extraction solution as requested by members of the Canadian Mustard Association.





Analytical Method


Two-gram samples of yellow mustard seed were placed in 50-ml disposable centrifuge tubes and 40 ml of water was added to each. The tubes were capped and extracted for 15 minutes in a boiling water bath. Afterwards the tubes were shaken on a reciprocal shaker for 30 minutes. The samples were then cooled to 20º before the viscosities of the extracts were measured with a Cannon-Fenske Routine Viscometer (Size 100). Only the number of samples that could be processed within the same day (40) was extracted each day. The kinematic viscosity of the extracts was calculated in centistokes as the efflux time in seconds multiplied by the viscometer constant. The mucilage of the samples was calculated as the viscosity of extract minus the viscosity of water multiplied by the volume of water divided by the weight of the seed (cS*ml/g).


Breeding population and selections


The breeding population of yellow mustard was created in 1989 by making, after an initial screening for suitability in terms of seed colour and adaptability to the western Canada environment, a composite of all germplasm introductions held at the Saskatoon Research Centre. Two cycles of recurrent selection for increased yield were conducted on this population which produced the cultivars AC Pennant and AC Base. In addition, another cycle of selection for increased yield, protein and seed weight produced the lines SA96-HYLD, SA-HWST and SA-LOHP which were each composites of reserve seed from 30 to 40 progeny-tested plants evaluated in 1995. Individual plants were harvested from isolated crossing blocks of these three populations in 1996. In 1997, a total of 1000 of these plants were evaluated in a field nursery consisting of two row plots three metres long replicated twice. SA96-HYLD, AC Pennant and Viscount were entered as checks. Entries were evaluated for agronomic performance, seed yield, protein content, mucilage content, 1000 seed weight and oil content. Twenty-one lines were selected on the basis of seed yield, protein content and mucilage content. The mucilage content of the selections ranged from 41.1 to 64.8 cS*ml/g with an average of 52 (slightly greater than Viscount’s average). Replicated yield trials (1.5 by 6 metre plots with four replicates) of these selections plus checks were grown at three locations in 1998. Due to insufficient reserve seed remaining from each of these lines, seed from the evaluation rows was used to plant these trials. An additional entry consisting of a composite of reserve seed from the 21 selected lines was also entered. Harvested seed from this trial was evaluated for seed yield, protein content and mucilage content.




Western Canada mustard co-operative trials


The results from the 1998 western Canada mustard co-operative trials are shown in Table 1. The results from the 1997 trial were very similar. There were significant differences between entries and as expected, Viscount had consistently the highest amount of mucilage. A significant location effect on mucilage content was observed, but the ranking of lines with respect to mucilage remained consistent across locations. AC Base and SA96-HYLD had a relatively low amount of mucilage. SA96-LOHP and SA96-HWST had similar levels (data not shown).


Table 1. Mucilage content of entries in the 1998 Western Canada Co-op Mustard Trialsa



Variety/Line        WRN         IRR        STD      WAT      SWC        LSB        SCT        SKT      Mean

AC Pennant           49.1        48.8        45.2        62.2        55.5        69.5        58.3        51.2        52.8

Gisilba                   45.2        48.0        40.2        51.2        49.3        71.1        53.3        43.8        48.8

Ochre                    41.6        44.4        34.8        53.0        45.3        58.8        51.0        41.7        45.3

AC Base               44.2        47.2        39.7        55.8        43.2        62.0        47.6        47.1        46.8

SA96-HYLD        42.7        45.3        36.8        51.3        46.8        62.8        47.6        46.4        46.1

Tilney                    46.7        52.7        47.2        63.1        54.0        70.4        53.6        59.7        56.0

Viscount                57.5        61.4        53.7        80.1        72.9        82.4        70.9        69.0        65.6

LSDc                       4.8          5.4          3.9          5.6          6.3          9.1          4.2          4.1          1.8

a.        Data taken from Western Canada 1998 Co-operative Mustard Report compiled by Jo-Anne Relf-Eckstein, Agriculture & Agri-Food Canada, Saskatoon Research Centre.

b.       Warner, AB (WRN); Irricana, AB (IRR); Standard, AB (STD); Watrous, SK (WAT); Swift Current, SK (SWC); Lashburn, SK (LSB); Scott, SK (SCT); Saskatoon, SK (SKT); average of all ten sites evaluated in 1998 (Mean).

c.        Least Significant Difference, by a two sided LSD at 0.05 probability.

1997 field nursery


Significant differences in mucilage content were observed between lines in the 1997 nursery (Fig. 1). Individual mucilage values ranged from 13.8 to 78.6 cS*ml/g. As expected, the mucilage content of SA96-HYLD was at about the mid-point of values observed in the nursery entries. Many of the lines in the nursery contained less mucilage than Viscount, but some lines were higher than Viscount indicating that selection for mucilage contents higher than Viscount is possible from this population.














Figure 1. Histogram of mucilage content in 1997 field nursery


1998 yield trials


Again, there were significant differences between entries in the trials (Table 2). As well, there was a highly significant location effect. As would be expected from the outcrossing nature of this species, the mucilage content of entries whose source seed was from the 1997 evaluation rows regressed towards the average of the 1997 nursery. Still, 20 of 21 entries had higher mucilage contents than AC Base. This indicated that selection for higher mucilage content was successful. The entry (HMUC), created from the composite of the reserve seed of the selected lines, had an average mucilage content of 61.2 cS*ml/g compared to Viscount’s 65.3 cS*ml/g indicating that by using reserve seed much of the selected mucilage gain could be retained.


Table 2. Mucilage content of some entries in 1998 replicated yield trials


Entry                               1997a             SKT1b             SKT2b          Scott, Sk               Mean

SA97-765                         41.1                 48.3                 29.5                 47.5                 41.7

SA97-170                         51.2                 54.3                 36.8                 46.0                 45.7

SA97-202                         48.5                 51.7                 41.3                 52.9                 48.6

SA97-660                         49.9                 53.2                 44.0                 56.2                 51.1

SA97-938                         48.4                 60.0                 41.5                 65.6                 55.7

SA97-297                         59.2                 64.4                 50.0                 62.2                 58.9

SA97-504                         55.3                 63.6                 48.7                 66.6                 59.6

SA97-1009                       64.8                 66.6                 51.2                 75.0                 64.2

HMUC                             52.0                 72.7                 47.4                 63.5                 61.2

AC Basec         Check                               48.6                 35.2                 45.4                 43.1

Viscount           Check       50.0                 71.0                 54.8                 70.0                 65.3

Ochre               Check                               41.0                 32.4                 41.9                 38.4

LSD                                                          12.7                 10.0                   8.3                   5.8

a.        Mucilage content (cS*ml/g) in 1997 field nursery.

b.       Saskatoon Date 1 (SKT1); Saskatoon Date 2 (SKT2)

c.        Replaced SA96-HYLD because of its yield advantage and equivalent mucilage content.


A positive correlation of 0.72 was found between the mucilage data collected in the 1997 nursery and the average mucilage of the selected lines, not including HMUC (Fig. 2). This indicates that mucilage content measured by viscosity may be a highly heritable trait. Two lines have been entered into the 1999 western Canada mustard co-operative trial on the basis of the 1998 replicated yield trials.













Figure 2. Correlation of 1998 average mucilage content with 1997 nursery mucilage content




There are significant mucilage content differences between cultivars of yellow mustard commonly grown in western Canada. We have also found significant differences between lines within our breeding population of yellow mustard. Selection within this population for higher mucilage contents is shown to be possible. By the method of measurement chosen (viscosity of extract) the differences seen may be due to either differences in amounts of mucilage or differences in mucilage composition, which effect its ability to alter viscosity. Further investigation is required in order to determine which is true.




The authors acknowledge the financial support of the Canadian Mustard Association and the Matching Investments Initiative program of Agriculture and Agri-Food Canada for this project.




Aguilar, C.A. and Ziegler, G.R. 1990. Effect of temperature and electrolytes on the viscosity of aqueous dispersions of mustard seed. Food Hydrocolloids 4, 161-166.


Cui, W., Eskin, N.A.M. and Biliaderis, C.G. 1994. Yellow mustard mucilage: chemical structure and rheological properties. Food Hydrocolloids 8, 203-214.


Gerhards, Ch. and Walker, F. 1997. Rheological properties of mustard mucilage isolated from raw and from processed mustard. Nahrung 41, 96-100.


Siddiqui, I.R., Yiu, S.H., Jones, D.H. and Kaláb, M. 1986. Mucilage in yellow mustard (Brassica hirta) seeds. Food Microstructure 5, 157-162.


Weber, F.E., Taillie, S.A., and Stauffer, K.R. 1974. Functional characteristics of mustard mucilage. Journal of Food Science 39, 292-300.