HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Diederichsen, A. Articles by Duguid, S. D. Search for Related Content PubMed Articles by Diederichsen, A. Articles by Duguid, S. D. Agricola Articles by Diederichsen, A. Articles by Duguid, S. D. Related Collections Other Oil Crops Plant Genetic Resources Seed Technology

CROP BREEDING, GENETICS & CYTOLOGY

Variation of Mucilage in Flax Seed and Its Relationship with Other Seed Characters

^a Plant Gene Resources of Canada, Saskatoon Research Centre, Agriculture and Agri-Food Canada, 107 Science Place, Saskatoon, SK, S7N 0X2, Canada
^b Morden Research Station, Agriculture and Agri-Food Canada, 101 Route 100 Unit 100, Morden, MB, R6M 1Y5, Canada

^* Corresponding author (diederichsena{at}agr.gc.ca)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
A four-location trial in western Canada with 16 North American flax (Linum usitatissimum L.) cultivars was conducted in 2001 and 2002 to investigate the genotypic and environmental influences on mucilage in the flax seed coat. The viscosity of a water extract from intact flax seed was used as quantitative descriptor of the seed-coat mucilage (mucilage indicator value, MIV). Thousand-seed weight (TSW) and oil content in the seed were also observed. The MIV in the 16 cultivars ranged from 90.6 to 246.1 centistokes (cSt) mL g^–1, TSW ranged from 5.21 to 6.91 g, and oil content ranged from 447 to 491 g kg^–1. Significant differences were found among the cultivars for all characters. A screening of 1689 recently regenerated accessions from 61 countries of the Plant Gene Resources of Canada (PGRC) flax collection showed MIVs ranging from 22.1 to 343.4 cSt mL g^–1. The ranges for TSW and oil content were 3.53 to 11.50 g and 314 to 457 g kg^–1, respectively. The group displaying the highest MIV included registered North American cultivars. There was no association of MIV with geographic origin, TSW, or seed oil content. Recent North American linseed cultivars varied considerably for MIVs, and breeding for this trait without impact on seed oil content or TSW should be possible.

Abbreviations: cSt, centistokes • MIV, mucilage indicator value • PGRC, Plant Gene Resources of Canada • TSW, thousand-seed weight

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
MUCILAGE IN THE SEED COAT of cultivated flax (L. usitatissimum L. subsp. usitatissimum) is a potential product derived from cultivated flax in addition to the existent products of seed oil and fiber for textile or technical use. A separate recovery of mucilage and seed oil is possible because mucilage is located in the seed coat and can be extracted independently from the seed oil. High content of mucilage in the seed coat is of interest for the food industry because flax mucilage can be used as a stabilizer or a thickening agent similar to guar [Cyamopsis tetragonoloba (L.) Taub.] gum, locust bean (Ceratonia siliqua L.) gum, or gum arabic from Acacia senegal (L.) Willd. or A. seyal Delile (Mazza and Biliaderis, 1989). The relaxation or mild purgative effect of flax mucilage in human nutrition has been known since ancient times and the health benefits of the seed oil, rich in unsaturated fatty acids, has also been noted (Heeger, 1956; Hammer, 2001; Westscott and Muir, 2003). On the other hand, flax with low mucilage content is better suited for poultry, swine, and fish feed, since it can be given in large quantities without interfering negatively with the animals' digestion (Bhatty, 1993). Flax seed markets in the animal feed sector are already established for hens to produce eggs with high -linolenic acid content. For these reasons, it was suggested more than 10 yr ago that flax with either high or low mucilage content may need to be developed (Bhatty, 1993).

The existence of genetic diversity regarding mucilage content and properties is necessary to improve this trait by breeding. So far, a limited amount of flax germplasm has been evaluated for mucilage content or chemical and physical properties of the mucilage. The genotypic influence on mucilage content and rheological properties was reported in several studies: Bhatty (1993) investigated eight Canadian flax cultivars; Wannerberger et al. (1991) evaluated 23 European cultivars, Cui et al. (1996) studied 12 North American lines, Cui and Mazza (1996) reported results from four cultivars, and the most recent study by Chornick et al. (2002) was based on seven Canadian cultivars. The most comprehensive germplasm screening by Oomah et al. (1995) found mucilage contents between 3.6 and 8% with an overall mean of 6.2% in 109 accessions. Bhatty and Cherdkiatgumchai (1990) reported location had a greater influence on flax meal composition than cultivar. Flax mucilage is located in the epidermal cell layer of the seed coat and consists of polysaccharides which can be separated into a neutral and two acidic polymers (Heinze and Amelunxen, 1984; Warrand et al., 2003). The mucilage parameters are not correlated with the seed oil or protein content (Oomah et al., 1995).

To our knowledge, breeding programs with a focus on mucilage in the seed coat do not exist, although a considerable amount of research has already been done in this area. It is useful for flax producers to know the potential of current flax cultivars for producing seed with low or high mucilage content. If diversity for this character existed in current cultivars, then current cultivars could easily be used to serve a potential demand for high or low mucilage viscosity. If such a market demand comes into existence, information about the diversity of this character in the flax world collection of PGRC will be very useful. Therefore, the objectives of this study were: (i) to investigate genetic variation and environmental influence on mucilage in flax cultivars adapted to western Canadian growing conditions; (ii) to screen the PGRC world collection for differences in mucilage, and (iii) to consider the mucilage in relation to seed weight and oil content. We present in this paper the results of a replicated field trial with a limited number of cultivars in combination with the results of nonreplicated observations of the same characters in a large germplasm collection. The results of the replicated field trial give some orientation when interpreting the results of the screening of the PGRC flax collection and such approach seemed useful for investigating the variation in a large germplasm collection, which does not easily allow for replicated field trials.

	MATERIAL AND METHODS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Plant Material, Locations, and Experimental Design for the Multilocation Trial
A replicated field trial was conducted with 14 Canadian flax cultivars, one flax cultivar from the United States, and one Canadian breeding line (hereafter all called cultivars; Table 1). For eight cultivars, information about the mucilage was available from a preliminary screening conducted by PGRC in 1999. They were selected to represent a broad range (low to high) of MIVs. The multilocation trial was conducted in 2001 and 2002 at four research sites of Agriculture and Agri-Food Canada (AAFC) in western Canada: Morden, MB; Indian Head, SK; Melfort, SK; and Saskatoon, SK. The two experimental years were characterized by drought conditions at Saskatoon and Melfort. The 2002 growing season was generally warmer than 2001 at all four locations (Table 2). The experimental design was a randomized complete block design with three blocks, each of which represented a replication. Plot size varied at the locations ranging from 3.6 m² at Indian Head to 6.8 m² at Melfort; details of the plot sizes are presented in Table 2. The seeding rate was 5.65 g m^–2 at all locations, which corresponds to the seeding rates applied in farming practice in western Canada. The seeds were planted in four equidistant rows. Harvest was done with plot combines. Subsamples of the cleaned seed which were free from cracked seed and randomly taken from the single plot yields were used for determination of MIV, oil content, and TSW, respectively.

Mucilage and Other Seed Characteristics
The seed-coat mucilage content was quantitatively described by measuring the viscosity of a hot-water extract from intact flax seed as suggested by Bhatty (1993). The mucilage extraction was conducted by placing flax seed (1 g or most often 2 g, ± 0.001 g) in 50-mL disposable polypropylene centrifuge tubes (no. 05-539-9, Fisher Scientific, Leicestershire, UK). After the addition of 40 mL of deionized water, the tubes were sealed and brought to 100°C by placing them in a boiling water bath for 15 min. Thereafter, the tubes were shaken for 30 min. and then centrifuged (6000 rcf) for 1 h at room temperature to clarify the extract. The kinematic viscosity measured in cSt of the liquid fraction was determined at 20°C using calibrated Canon-Fenske viscometers (ASTM size 150). For calculation of the MIV, a standardization for the water volume (mL) used for seed extraction and the exact amount (g) of seed used was conducted as shown in Eq. [1].

[1]

This MIV allowed for comparison among the samples independent of the amount of water used for extraction and the amount of seed used for extraction.

Thirty accessions from the PGRC world collection produced, in addition to the soluble mucilage, an insoluble gelatinous precipitate which could not be removed by centrifugation or filtration and, therefore, influenced the viscosity measurements. Because of the inseparable precipitate, it was impossible to conduct a quantitative measurement of the water extract viscosity for these accessions using the applied methodology, and these accessions were excluded from the summarizing overview of the remaining 1689 accessions considered in this study.

Thousand-seed weight was calculated after weighing 300 seeds. Oil content was determined on air-dried (40°C, 2 d), intact seed by continuous-wave nuclear magnetic resonance spectroscopy (NMR; model 4000, Oxford Instruments, Oxon, UK) based on a seed sample of 10 g according to the recommendations of the American Oil Chemists Society (Firestone, 1998).

Statistical Data Analysis
An ANOVA was conducted for MIV, TSW, and oil content for the multilocation trial using the GLM procedure of the SAS computer program (v. 8.2, SAS Institute, Cary, NC). Cultivar was considered as fixed effect; and location, year, and blocks as random effects. The blocks (replications) were nested in location and year. Missing values from three plots occurred due to lost harvest bags and one additional oil analysis was missing. The missing values were substituted by expected values before the ANOVA. The LSD differences were adjusted accordingly as suggested by Baker (2000, personal communication). For each cultivar in the multilocation trial, a coefficient of variation of the seed characters across all locations and years was calculated to quantify the environmental variation. Correlations among the seed characters were calculated based on the mean values for each cultivar. The data from the screening of the world collection was summarized using descriptive statistics (minimum, maximum, mean, standard deviation, coefficient of variation). The results of the nonreplicated measurements of the 1689 accessions have a preliminary character when considering single accessions, but the large number of different germplasm accessions allowed recognizing trends within the entire genepool of cultivated flax or among larger subgroups of all accessions.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Multilocation Trial
The ANOVA for MIV, TSW, and oil content indicated significant effects of cultivar for all three characters (Table 3). The mean values for MIV of the 16 cultivars across all locations ranged from 90.6 cSt mL g^–1 (AC McDuff) to 246.1 cSt mL g^–1 (AC Emerson) (Table 4). The ranges for TSW and oil content were 5.21 g (‘Flanders’) to 6.91 g (line 84495) and 447 g kg^–1 (line 84495) to 491 g kg^–1 (‘Macbeth’), respectively (Table 4). The differences among the locations were not significant for all characters and only oil content showed significant differences between the years (Table 3). Nearly all interactions between the sources of variation were significant. The results for the four locations in the different years did not allow recognizing any clear associations of higher or lower MIV with the respective weather conditions or location effects. For example, the warm and wet growing season 2002 resulted in low MIVs at Morden (mean: 148.2 cSt mL g^–1) but in high MIVs at Indian Head (mean: 221.2 cSt mL g^–1). Also, the medium drought at Saskatoon in 2001 resulted in low MIVs (mean: 169.5 cSt mL g^–1), but under severe drought in 2002, the MIVs at this location were high (mean: 216.6 cSt mL g^–1). The mean values for each location of the three characters, and the mean values for the two experimental years had much narrower ranges than the ranges of mean values which are due to the genetic differences among the cultivars shown in Table 4. This underlines the genetic influence on this character.

Screening of the PGRC World Collection
The MIVs observed in 1689 accessions from the flax world collection ranged from 22.1 to 343.4 cSt mL g^–1 (mean ± SD = 134.1 ± 51.1 cSt mL g^–1). The 15 accessions of the world collection with the highest and lowest MIV are listed in Tables 5 and 6, respectively; the screening results for all accessions can be found in the PGRC Internet accessible database http://www.agr.gc.ca/pgrc-rpc, verified 13 Oct. 2005). The group displaying the highest MIV included registered North American cultivars. The variability of this character was high (coefficient of variation: 38.1%) compared with the variation in TSW found in the same accessions, which ranged from 3.53 to 11.50 g (mean: 6.0 ± 1.20 g, coefficient of variation: 19.9%), or the even less diverse oil content, which ranged from 314 to 457 g kg^–1 (mean: 382 ± 17.3 g kg^–1, coefficient of variation: 4.5%). The histogram for frequency of quantiles of MIVs was similar to the distribution histogram of TSW in showing skewness toward the lower values (Fig. 1 ). The distribution histogram for oil content was more symmetric around the mean value. There was no clear association of higher or lower MIVs with germplasm from certain regions, whereas TSW and oil content showed some geographic differentiation (Table 7). Heavier seed and higher oil contents were observed in flax from India, while flax from Africa (mostly Ethiopia) and Central Europe had lighter seed and had lower oil contents.

View larger version (19K): [in this window] [in a new window]   Fig. 1. Standardized histograms for distribution of mucilage indicator value (n = 1689), 1000 seed weight (n = 1660) and oil content (n = 1683) of flax accessions from the flax collection at Plant Gene Resources of Canada.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

A screening of the PGRC world collection identified genetic resources with lower and higher values (range: 22.1–343.4 cSt mL g^–1) for MIV than those found in the North American cultivars (range of mean values: 90.6–246.1 cSt mL g^–1). On the basis of the CV as quantifier, the variation for MIV within the North American cultivars (CV = 31.6%) and within the world collection (CV = 38.1%) is comparable with the high variation for the rhamnose to xylolose ratio in the water-soluble polysaccharides, for which Oomah et al. (1995) found a coefficient of variation of 45%. However, the variation of the MIV in our experiment exceeds the variation found for mucilage content reported as carbohydrate content with a coefficient of variation of 12.9% by Oomah et al. (1995).

The information gained in the multilocation trial about the genetic influence on variation in mucilage properties gave some basis to judge the relevance of the screening results of the nonreplicated measurements in the world collection. The environmental influence on MIV was higher than on TSW or seed oil content, but the wide range of variation of this character should allow for selecting germplasm from the world collection for this character also based on the nonreplicated observations. At least these measurements indicate which accessions out of the 1689 might deserve further investigation when looking for germplasm with high or low MIV.

The fact that the MIV was not (multilocation trial) or only weakly (world collection) correlated with seed weight and oil content and that there was also no clear association with seed color or geographic region of origin of the accession is possibly due to the complex chemical makeup of mucilage. Cui et al. (1996) reported that yellow-seeded accessions had higher viscosities than brown-seeded accessions, and the yellow-seeded accessions 84495 and ‘Omega’ were also included in their experiments. In our study, however, we found two brown-seeded cultivars, AC Emerson and AC Watson, had the highest MIVs. As Oomah et al. (1995) have shown, different chemical substances affect the water extract viscosity. Therefore, direct measurements of defined substances might be better suited to detect associations with certain geographic origin or other seed qualities, as Oomah et al. (1995) described. One might expect that seed weight should have significant influence on the MIV: Large seeds, which have greater seed size, would have less seed surface than small seed in an equal weight of seed. If the thickness of the mucilage layer on the seed coat would be equal in all genotypes, the amount of surface in a given weight would be critical.The wide range of TSWs in the world collection (3.53–11.5 g) should make such an effect visible, while the range for seed weight was perhaps too low in the cultivars of the multilocation trial (5.21–6.91 g) to make such tendency obvious. However, the correlation found in the world collection was not strong (Table 8). On the other hand, an association of small seed size with lower oil content became obvious in the world collection, but was not observed in the multilocation trial. One reason might be that the weight proportion of seed coat, which contains no oil, is larger in small seed.

The MIV measured in our study was clearly more influenced by genotype than by environment. Environmental influences on mucilage were described by Dorrell and Daun (1980), who reported that extreme humidity during the harvest period resulted in significantly lower mucilage content. None of the experimental locations in the present study experienced such unfavorable weather conditions during or immediately before harvest. However, the effects of environmental factors influencing this character could not be specified from our experiment. The differences found between the locations and the years were not significant. The soil conditions at the locations were quite similar (Table 2), but more experimental years would be required to understand the influence of the climatic conditions on this character and the interactions between genotype and environment.

Measurement of the water extract viscosity as a method for assessment of the mucilage content can be problematic. One concern is the chemical complexity of the mucilage. This may explain why the variation of the MIV in the North American cultivars and also in the world collection considerably exceeded the variation reported from direct quantitative measurements of the mucilage content by others. Another concern is that the water insoluble gelatinous precipitate observed consistently in accessions excluded from this investigation may still have occurred occasionally, albeit in low amounts, in the accessions included, and may have interfered with the viscosity measurements. This insoluble gelatinous precipitate, which is by definition not considered as mucilage, contributed to the mucus emerging from a flax seed after extraction with boiling water. This is offered as an explanation of the high coefficient of variation for the MIV found in line 84495 in the multilocation trial. The substances forming such a precipitate may also be of relevance when using or consuming flax seed, and these accessions deserve closer investigation. The physiochemical properties of flax mucilage warrant further studies because of the complex nature of these properties and their relationship to the chemical constituents of the flax mucilage.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Registered North American linseed cultivars displayed a wide range of diversity regarding the seed-coat mucilage. Mucilage indicator value and TSW or oil content were not associated. Therefore, it should be possible to breed for differing mucilage in flax without impacts on the latter two traits.

	ACKNOWLEDGMENTS

 
Funding for this project was received from the Saskatchewan Flax Development Commission, Saskatoon, and the Matching Investment Initiative of Agriculture and Agri-Food Canada, Ottawa. Thanks for technical help to Ms. L. Dunkley (Saskatoon), Ms. L. Jones-Flory (Saskatoon), Mr. D. Kessler (Saskatoon), Mr. P. Kusters (Saskatoon), Mr. M. Sandercock (Morden), and Mr. G. Serblowski (Saskatoon). We thank the directors and farm managers at the four experimental locations (Melfort, Morden, Indian Head, and Saskatoon) for support.

Received for publication February 15, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Bhatty, R.S., and P. Cherdkiatgumchai. 1990. Compositional analysis of laboratory-prepared and commercial samples of linseed meal and of hull isolated from flax. J. Am. Oil Chem. Soc. 67:79–84.
• Bhatty, R.S. 1993. Further compositional analysis of flax: Mucilage, trypsin inhibitors and hydrocyanic acid. J. Am. Oil Chem. Soc. 70:899–904.
• Chornick, T., L. Malcolmson, M. Izydorczyk, S. Duguid, and C. Taylor. 2002. Effect of cultivar on the physiochemical properties of flaxseed mucilage. p. 7–13. In Proc. 59th Flax Institute of the United States, 21–23 Mar. 2002, Fargo, ND. Flax Inst., Fargo.
• Cui, W., and G. Mazza. 1996. Physiochemical characteristics of flaxseed gum. Food Res. Int. 29:397–402.[CrossRef]
• Cui, W., E. Kenaschuk, and G. Mazza. 1996. Influence of genotype on chemical composition and rheological properties of flaxseed gum. Food Hydrocolloids 10:221–227.
• Diederichsen, A., J.P. Raney, Y.B. Fu, and K.W. Richards. 2002. Diversity in the flax collection at Plant Gene Resources of Canada. p.138–143. In Proc. 59th Flax Institute of the United States, 21–23 Mar. 2002, Fargo, ND. Flax Inst., Fargo.
• Dorell, D.G., and J.K. Daun. 1980. Effect of weather damage on the grade and quality of flaxseed. Can. J. Plant Sci. 60:799–805.
• Firestone, D. (ed.) 1998. American Oil Chemists Recommended Practice Ak-3–94. Oil content of oilseeds by nuclear magnetic resonance. Official methods and recommended practices of the AOCS. 5th ed. Am. Oil Chemist's Soc., Champaign, IL.
• Hammer, K. 2001. Linaceae. p. 1106–1108. In P. Hanelt and Institute of Plant Genetics and Crop Plant Research (ed.) Mansfeld's encyclopedia of agricultural and horticultural crops (except ornamentals). Vol. 2. Springer, Berlin.
• Heeger, E.F. 1956. Handbuch des Arznei- und Gewürzpflanzenbaues. (In German.) Deutscher Bauernverlag, Berlin.
• Heinze, U., and F. Amelunxen. 1984. The development of mucilage in seeds of Linum; electronmicroscopic and chemical analysis. (In German, with English abstract.) Ber. Dtsch. Bot. Ges. 97:451–464.
• Mazza, G., and C.G. Biliaderis. 1989. Functional properties of flax seed mucilage. J. Food Sci. 54:1302–1305.
• Oomah, B.D., E.O. Kenaschuk, W. Cui, and G. Mazza. 1995. Variation in the composition of water-soluble polysaccharides in flaxseed. J.Agric. Food Chem. 43:1484–1488.[CrossRef]
• Wannerberger, K., T. Nylander, and M. Nyman. 1991. Rheological and chemical properties of mucilage in different varieties from linseed (Linum usitatissimum). Acta Agric. Scand. 41:311–319.
• Warrand, J., P. Michaud, L. Picton, G. Muller, B. Courtois, R. Ralainirina, and J. Courtois. 2003. Large-scale purification of water-soluble polysaccharides from flaxseed mucilage, and isolation of a new anionic polymer. Chromatographia 58:331–335.
• Westscott, N.D., and A.D. Muir. 2003. Chemical studies on the constituents of Linum spp. p. 55–73. In A. Muir and N. Westscott (ed.) Flax, the genus Linum. Taylor & Francis, London.
• Zhukovskij, P.M. 1971. Kul'turnye restaenija i ich sorodii–Sistematika, geografija, citogenetika, imunitet, ekologija, proischodenie, ispol'zovanie. (In Russian) Kolos, Leningrad.

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Diederichsen, A. Articles by Duguid, S. D. Search for Related Content PubMed Articles by Diederichsen, A. Articles by Duguid, S. D. Agricola Articles by Diederichsen, A. Articles by Duguid, S. D. Related Collections Other Oil Crops Plant Genetic Resources Seed Technology