INFLUENCE OF CHARLOCK CONTAMINATION ON RAPESEED GLUCOSINOLATE LEVEL AND PROFILE

E.J. Booth¹, D.S. Farrington² and S.K. Dhaliwal²

¹Scottish Agricultural College, Craibstone Estate, Bucksburn, Aberdeen, AB21 9YA, UK, e.booth@ab.sac.ac.uk

²ADAS, Woodthorne, Wergs Road, Wolverhampton, WV6 8TQ, UK, derek_farrington@adas.co.uk

ABSTRACT

Charlock (Sinapis arvensis) is known to have a high content of glucosinolates and contamination of rapeseed has a considerable potential for reducing seed quality.  With greater interest in farm-saving seed, the study aimed to consider the extent of charlock contamination of rapeseed, using glucosinalbin (4-hydroxybenzyl glucosinolate), a glucosinolate found in charlock but not in rapeseed, as an indicator.  Samples of rapeseed for farm-saving were analysed for glucosinalbin over 4 seasons.  In a second experiment  samples of seed were collected from fields of both winter and spring sown oilseed rape, where the level of charlock contamination in the fields was also noted.  Samples were analysed for total glucosinolates, glucosinalbin and also examined by microscopy for detection of charlock seed.  It was found that there was a proportion of farm-saved seed samples containing glucosinalbin and that this varied considerably over the years of analysis.  Although indicating that charlock contributed to the high glucosinolate levels of many samples, high levels of glucosinalbin were found in relatively few of these and charlock contamination must be only one of a number of factors influencing glucosinolate content.   It was found that observation of charlock in the field was not always linked with high glucosinalbin content, particularly in winter oilseed rape.  It is suggested that there may be settling of seed, or more likely that charlock pods may shatter and release seed before harvest.  Avoidance of fields with high numbers of charlock plants will help to minimise glucosinolate content of the seed sample.

KEYWORDS  Sinapis arvensis, glucosinalbin, 4-hydroxybenzyl glucosinolate, farm-saved seed, weed seed contamination.

INTRODUCTION

Charlock is a frequently occurring weed of oilseed rape in the UK, and as a Brassica, there are limited opportunities for its control in the crop.  Several studies have shown that charlock has a high glucosinolate content. Work in Australia indicated a mean glucosinolate content of 196 µM/g meal (approximately 137 µM/g seed) for 5 lines of charlock (Salisbury et al., 1991) and seed from populations sampled in Scotland contained between 100 and 145 µM/g seed, whilst the glucosinolate content of seed from a cultivated plot of charlock was recorded as 115 µM/g seed (Booth and Walker, 1993). The similarity in appearance of charlock and oilseed rape seeds makes separation very difficult in the event of seed contamination and these studies show that the contamination of rapeseed with charlock could have significant implications for glucosinolate content and therefore seed quality of the sample.  Calculations have indicated that only a 5% (Salisbury et al., 1991) or even 2% (Booth, 1995) contamination of rapeseed could significantly raise glucosinolate levels.  With the current financial pressure on inputs, there is interest from growers in farm-saving seed and glucosinolate content is of particular relevance as there is a requirement for such seed to contain no more than 18 µM glucosinolates/g seed.  However, there was little information available on the actual extent of  contamination of rapeseed samples by charlock seed and of the effect on glucosinolate content of rapeseed samples.

The glucosinolate glucosinalbin is absent from the glucosinolate profile of pure rapeseed, whereas it is the predominant glucosinolate in charlock seed, allowing the occurrence of this glucosinolate to be used as an indicator of charlock contamination.  The aims of the present work were to investigate occurrence of charlock contamination in harvested rapeseed samples and to indicate the relationship between observations of charlock in the field and actual seed contamination.

MATERIALS AND METHODS

Farm saved seed samples submitted for testing from throughout the UK over 1994 - 1997 were analysed by ADAS for total glucosinolate content using X-Ray Fluorescence.  Samples which were shown to have a glucosinolate content of between 18 and 25 µM were selected for further analysis by HPLC (Anon, 1995) for individual glucosinolates, including glucosinalbin.

In the second experiment undertaken for 1997 material, rapeseed samples were taken from 30 different fields within Scotland, selected at random and including both winter and spring varieties.  The rates of contamination of charlock within the crop were reported by samplers based upon observation of the crop in the field throughout the growing season.  Individual glucosinolate content was assessed using HPLC and charlock seed contamination was also evaluated by visual inspection aided by microscopy.  Charlock was identified, by staff with considerable experience, using the following criteria: (i) charlock has a ‘greasier’ and more shiny coat than rapeseed, (ii) charlock is more spherical than rapeseed and (iii) charlock has no visual radical lobes.

RESULTS 

Individual glucosinolates are deemed to contribute to the calculation of total glucosinolate content when they are found in excess of 1% of the total glucosinolate content.  The results from trial 1 indicate that a significant proportion of farm-saved seed with a high total glucosinolate content contained glucosinalbin.  There was a very large year-to-year variation in the percentage of samples containing glucosinalbin with 1994 and 1997 being associated with a much greater proportion of samples containing glucosinalbin, indicating widespread charlock contamination (Table 1).  Of the samples in which glucosinalbin was detected, this glucosinolate was found at relatively high levels (above 5µM/g seed) in a higher proportion of samples in 1994 and 1995, indicating greater severity of charlock contamination.

With regard to the second experiment, it can be seen that glucosinalbin was not detected for every sample where charlock was observed in the corresponding crop (Table 2).  For winter varieties, glucosinalbin was observed in 5 of the 18 samples when charlock was noted in the field, whereas in 8 cases where charlock was noted, no glucosinalbin was detected.  In 4 of the 5 cases where no charlock was seen, no glucosinalbin was detected and for only one case where no charlock was seen glucosinalbin was detected.  For the spring varieties, in a greater proportion of samples (4 out of 6) where charlock was noted, glucosinalbin was also detected.  In 4 spring rape samples no charlock was recorded and 3 had no glucosinalbin.  Again, only one sample contained glucosinalbin where no charlock was seen in the field. 

Table 1.  Percentage of farm-saved seed containing glucosinalbin.

Harvest year	Total number of samples	Number of  samples analysed by HPLC*	Number of  samples containing glucosinalbin	Number of samples containing over 5µM glucosinalbin/g seed
1997	1445	341	115	3
1996	1722	207	18	0
1995	1398	494	33	5
1994	227	62	45	7

*samples containing 18-25 µM glucosinolates /g seed

Microscopy failed to detect the presence of charlock seed in all but 2 cases.  These 2 samples were also found to contain high levels of glucosinalbin, indicating a high contamination rate of charlock.  A very high population of charlock was also noted by the samplers.  Both samples were from spring varieties, and levels of glucosinalbin detected in winter varieties were much lower.  It was notable that for this season, even for varieties with a high level of glucosinalbin, the total glucosinolate content did not exceed the 18 µM/g seed limit.

Table 2.  Results from glucosinolate analysis and microscopic examination of rapeseed samples taken from fields where various amounts of charlock were observed at time of harvest)

Variety	Total Glucosinolates (µM/g seed)	Glucosinalbin (µM/g seed)	% charlock (microscopy)	Observed charlock in field
Winter varieties
Synergy	12.5	0.01	nil	present
Falcon	4.5	0.01	nil	lots
Rocket	9.7	0.17	nil	present
Falcon	5.5	0.39	nil	present
Falcon	4.0	0.15	nil	present
Synergy	10.8	0.15	nil	none
Arietta	8.7	nil	nil	present
Falcon	7.4	nil	nil	present
Falcon	4.6	nil	nil	present
Synergy	6.9	nil	nil	present
Synergy/Falcon*	4.6	nil	nil	present
Synergy	7.0	nil	nil	lots (patches)
Express/Gazelle*	1.9	nil	nil	present
Falcon	9.0	nil	nil	present
Alpine	6.6	nil	nil	none
Express	7.5	nil	nil	none
Cocktail	16.4	nil	nil	none
Alpine	6.8	nil	nil	none
Spring varieties
Kova	9.5	0.22	nil	present
Kulta	8.6	0.04	nil	present
Starlight	12.9	9.34	present	lots
Starlight	15.6	10.1	present	lots
Starlight	9.9	0.12	nil	none
Starlight	3.9	nil	nil	present
Kulta	5.5	nil	nil	present
Hyola	6.4	nil	nil	none
Acrobat	5.0	nil	nil	none
Aries	9.2	nil	nil	none

(NB mould was found in 2 samples preventing analysis)

CONCLUSIONS

The presence of glucosinalbin in a significant number of the farm saved rapeseed samples analysed by HPLC suggests that charlock seed contamination is occurring and that charlock contamination is contributing to failure of samples to meet the glucosinolate requirements for farm saved seed.  Contamination by charlock seed will raise the overall level of glucosinolate content disproportionately due to the higher glucosinolate content of this weed.  The higher glucosinolate levels also have implications for animal feed due to the anti-nutritive nature of glucosinolates.  As relatively few samples contained high levels of glucosinalbin, other factors are also contributing to the high glucosinolate levels of farm-saved seed samples tested.

The samples considered in experiment 2 here were produced in Scotland which has a  relatively low atmospheric sulphur deposition and low sulphur soils.  Increased sulphur supply is known to raise glucosinolate level of both rapeseed and charlock seed (Booth, 1995).  Therefore, the effect of the observed charlock contamination is likely to be greater in conditions of greater sulphur availability.

The glucosinalbin analysis appears to more consistently reflect the observations of the samplers than of the findings of the microscopist although neither method gives a positive result in all cases where charlock was present within the crop. The results presented here, along with other work which considered a comparison of glucosinalbin analysis and microscopic examination of samples with known rates of charlock contamination (Dhaliwal, 1998) suggest that glucosinalbin analysis may be a more sensitive indicator than microscopy.  Only experienced samplers were used and their observations must be considered as reliable which suggests that the observations of the presence of the weed cannot be used as an indicator as to the level of contamination in the harvested seed.  It is possible that sampling of the seed is not representative due to settling of the smaller weed seed although this is unlikely as the difference in size is very slight.  A more likely explanation is that the charlock has matured earlier than the rape and that the seed has been released and is already on the ground at the time of harvest.  This is also indicated by the greater number of samples of spring varieties containing glucosinalbin, as charlock germinating in the spring when spring rapeseed was sown would be more likely to retain intact pods until harvest than charlock germinating in the winter sown rapeseed crop.  The variation in glucosinalbin content from year to year may imply that conditions at harvest influence shattering of charlock pods and thus seed contamination. 

Although it seems that the occurrence of charlock plants in the field may not necessarily lead to significant seed contamination, it would still be prudent to control charlock to prevent build up of seed and to select seed for farm-saving, particularly from spring varieties, from fields which have low numbers of charlock plants.

ACKNOWLEDGMENTS

The funds to undertake this work were provided by the Ministry of Agriculture, Fisheries and Food and also by  the Scottish Office Agriculture, Environment and Fisheries Department.

REFERENCES

Anon, 1995.  British Standard: Rapeseed - Determination of glucosinolates content - Part 1.  Method using high-performance liquid chromatography.  BS 4289: Part 9: 1993.  BS EN ISO 9167-1:1995.

Booth E J (1994) The influence of environment and management on the glucosinolate content of oilseed rape (Brassica napus L.).  PhD Thesis, the University of Aberdeen, September 1994.

Booth E J and Walker K C (1993) Consequences of Cruciferous weed contamination for oilseed rape quality in the north of Scotland.  6 Aberdeen Letters in Ecology, 12-13.

Dhaliwal S K (1998)  Glucosinalbin study.  Report for M.A.F.F., May 1998.

Salisbury P, Mailer R and Sang J (1991) Potential reduction in Australian canola quality from weedy Crucifer contamination.  Cruciferae Newsletter, Eucarpia 14/15, 128-129.