PERFORMANCE OF AUSTRALIAN CANOLA QUALITY BRASSICA JUNCEA BREEDING LINES

W.A. Burton¹, S.J. Pymer¹, P.A. Salisbury^1,2, J.T.O. Kirk³ and R.N. Oram³

¹Agriculture Victoria, Victorian Institute for Dryland Agriculture, Private Bag 260, Horsham Victoria 3401 Australia

²Institute of Land and Food Resources, The University of Melbourne, Parkville Victoria 3052 Australia

³CSIRO Division of Plant Industry, GPO Box 1600, Canberra ACT 2601 Australia

email: wayne.burton@nre.vic.gov.au

ABSTRACT

Indian mustard (Brassica juncea) breeding lines developed at the Victorian Institute for Dryland Agriculture (VIDA), Horsham and CSIRO, Canberra have been tested as a more drought and heat tolerant alternative to canola quality B. napus.  These lines are being targeted at the drier half of the southern Australian wheatbelt, where limited alternative break crops are available for use in rotation with cereals.  The first canola quality (double low) B. juncea evaluated were late maturing and not suited to low rainfall environments.  However, recently developed early maturing canola quality B. juncea have shown significant yield advantages over early maturing B. napus varieties in these areas.  Yield increases to the present level of about 1.0 – 1.5 t/ha in these low rainfall environments have been achieved by shortening both the time to flowering and the duration of flowering and by increasing pod length.  The best canola quality B. juncea lines have glucosinolate concentrations between 2-20 mmol/g of seed, less than 2% erucic acid in the seed and oleic acid levels ranging from 45-55%.  Early flowering, early maturing lines with high yield and good agronomic characteristics and quality have been identified.  The breeding program at VIDA is now selecting high yielding, early flowering, double low lines with high levels of oleic acid in the oil, so that canola quality B. juncea seed can be used interchangeably with B. napus canola seed in the food and feed industries. 

KEYWORDS: glucosinolates, high yield, double low, erucic acid, early flowering, oleic acid

INTRODUCTION

Brassica juncea (Indian mustard) has many advantages over B. napus (canola) in low rainfall cropping districts of Australia, including more vigorous seedling growth, quicker ground covering ability, greater tolerance to heat and drought and enhanced resistance to the disease blackleg and to pod-shattering.  The potential benefits of developing canola quality B. juncea are recognised by a number of northern hemisphere countries, particularly Canada, where there are major breeding programs focussed on its development.  The Indian mustard breeding lines developed at VIDA, and CSIRO, Canberra have been evaluated over the past three years as a more drought- and heat-tolerant alternative to canola quality B. napus in the drier part of the southern Australian wheatbelt.  The VIDA breeding program is aiming to select canola quality (double low) B. juncea lines which are high yielding, early flowering and large seeded with good early vigour and agronomic acceptability.  A major objective is to select genotypes which reach the end of flowering earlier than the new early maturing B. napus cultivars and have comparable or enhanced yields.  An additional aim is to raise oleic acid levels to 60-65% to match those in canola.  This paper summarises progress in the development of canola quality B. juncea for Australia. 

RESULTS AND DISCUSSION

In 1996/97, over 12,000 single plant selections were taken from 72 F₃-F₅ families in the field at Horsham (in the Wimmera); in 1997/98 and again in 1998/99 over 1,500 selections were taken from 45 F₂-F₄ families at Horsham and at Beulah (in the southern Victorian mallee).  Early flowering and maturing selections were quality tested (using near infra-red spectroscopy [NIRS] analysis) for oil and protein content, total glucosinolate concentration and fatty acid composition.  The highest quality selections were then preliminary yield tested at Beulah before going into multi-location trials across Australia the following year. 

1996/97 Selections

Of the selections analysed in 1996/97, <3% had glucosinolate concentration <20 µmol/g seed and <12% had glucosinolate concentration <30 µmol/g seed.  Selections with glucosinolate concentration <30 µmol/g of seed came from only 30 of the 72 families.  The oleic acid levels of the selections ranged from 15-60% (Figure 1).  Several selections with low glucosinolate concentrations (<20 µmol/g of seed) and moderate levels of oleic acid (50-54%) were identified.  The lines with higher levels of oleic acid (up to 60%) tended to have oleic acid contents that were genetically unstable, producing progeny with reduced levels of oleic acid the following year.  These lines were also late, tall, low yielding and often had high glucosinolate concentration.  These selections have been crossed with early selections having better agronomic traits and seed quality. 

Figure 1.  Scatter diagram of glucosinolate concentration (µmol/g of seed) against oleic acid content.

1997/98 Selections

Promising yield results were obtained from Beulah in 1998 (Figure 2).  Extremely dry conditions were experienced in 1998, with less than 250 mm rainfall during the growing season.  The best earlier flowering double-low B. juncea lines with shorter flowering duration and longer pods were close to double the yield of the early flowering B. napus variety,  Monty (Figure 2).  A number of these high yielding, early flowering lines have been sown into multi-location replicated trials across Australia in 1999. 

Figure 2.  Yield  (t/ha) of selected canola quality B. juncea breeding lines compared to early season canola cultivars Monty and Rainbow at Beulah in 1998. 

In Canada, yield increases of up to 15% of canola quality B. juncea over B. napus and B. rapa have been recorded (D. Potts, pers comm).  Yields to date of Australian canola quality B. juncea have not been as good as the early B. napus cultivars across a number of environments.  There is evidence from previous years that in low rainfall environments the time to the beginning and end of flowering is negatively related to yield in both B. napus and B. juncea (Oram et al., 1997).  The selection of genotypes which are early flowering and which reach the end of flowering at least as early as the early currently grown B. napus cultivars, has therefore been a major priority for the breeding program. 

The oil and protein content, glucosinolate concentration and fatty acid composition of the seed of the high yielding lines preliminary yield tested at Beulah during 1998 (selections taken in 1997/98), are shown in Table 1.  These lines have been included in multi-location trials across Australia in 1999.  Seed samples taken from yield plots at Beulah had higher oil and protein contents when compared to B. napus (Table 1).  They also had much lower oleic acid levels than either B. napus or the single plants selected from the same families of B. juncea taken at Horsham (Table 2). 

Table 1.  Oil and protein content, glucosinolate concentration and fatty acid composition of high yielding canola quality B. juncea lines compared to B. napus cv. Monty from the 1998 Beulah yield trial. 

	Oil Content %	Protein Content %	Glucosinolate µmol/g seed	Oleic Acid %	Linoleic Acid %	Linolenic Acid %
B. juncea
JN1	36.79	30.28	28.26	43.88	31.31	15.46
JN2	36.75	30.36	22.81	41.58	32.17	17.70
JN3	37.40	28.48	25.87	43.43	31.75	16.26
JN5	38.37	27.06	20.03	44.67	31.35	15.39
JN9	36.04	28.82	18.42	43.90	28.42	16.56
JN25	37.71	27.57	23.67	43.38	31.48	16.15
B. napus
MONTY	35.38	25.50	7.11	60.84	19.63	11.19

1998/99 Selections

The oil and protein content, glucosinolate concentration and fatty acid composition of seed from some of the most promising selections taken in the 1998/99 season are shown in Table 2.  Seed from single plants at Beulah tended to have much lower oleic acid levels than those taken from Horsham, this is possibly due to the high temperatures and low soil moisture at grain filling.  The best low glucosinolate (0-20 µmol/g of seed), higher oleic acid (45-55%) and earlier flowering and maturing lines will be reselected with some being grown in yield trials at Beulah in 1999.  The highest yielding and best quality lines from these trials will then go into multi-location replicated trials across Australia in 2000. 

Table 2.  Oil and protein content, glucosinolate concentration and fatty acid composition of five of the highest quality single plant selections from Horsham and Beulah in 1998. 

	Oil Content %	Protein  Content %	Glucosinolate µmol/g seed	Oleic Acid %	Linoleic Acid %	Linolenic Acid %	Erucic Acid %
97-015J3*814^a	40.09	23.57	15.34	54.73	21.04	11.58	0.69
97-006J3*801^a	38.15	23.82	19.54	53.18	23.11	13.17	0.66
96-006J4*10-816^b	37.04	28.64	13.16	44.72	33.3	13.89	0.47
96-024J4*4-807^b	38.61	27.97	13.91	45.07	29.37	17.01	0.48
96-006J4*10-808^b	38.92	26.11	18.40	46.47	31.59	13.40	0.49

^a Selections taken from Horsham (Irrigation) in 1998, ^bSelections taken from Beulah in 1998

Oram et al. (1997) discussed the difficulties associated with increasing oleic acid content in B. juncea.  However, this work has shown there is variation for oleic content in B. juncea and, with further crossing and selection over several generations, or by the use of anti-sense technology, it should be possible to increase oleic levels in B. juncea to those in canola.  Recently, Potts et al. (1999) discovered a high oleic variant in B. juncea and have high oleic B. juncea lines currently being increased for commercial release.  Stoutjesdijk et al. (1999) have also down-regulated d-12 desaturase and so increased oleic acid levels to 65-70% in Australian adapted material. 

CONCLUSIONS

Seed from the best canola quality B. juncea lines from the VIDA and CSIRO Brassica breeding programs have glucosinolate concentrations between 2-20 µmol/g, less than 2% erucic acid and oleic acids levels between 45-55%.  Early flowering, early maturing lines with high yield and good agronomic characteristics combined with canola oil quality have been developed.  Increased oleic acid levels and the selection of high yielding, early flowering lines are now the main priorities for the breeding program at VIDA.  Potential sources of high oleic acid incorporated into Australian high yielding, early flowering lines with good adaptation will enable an industry to be developed in the drier wheatbelt of Australia. 

ACKNOWLEDGEMENTS

This research project has received financial support from the Grains Research and Development Corporation.  The assistance of Mr David Robson, Mr Darryl Pearl, Ms Rachael Sullivan for the field operations and Dr Tadeusz Golebiowski, Ms Claire Greenwood, Ms Tina Pallot and Ms Audrey Leong for quality analysis is acknowledged. 

REFERENCES

Oram, R.N., Walton, G. Marcroft, S., Potter, T.D., Burton, W.A., Robson, D.J., Salisbury, P.A., Easton, A.A. and Kirk, J.T.O. (1997).  Progress in breeding canola-grade Indian mustard.  In: Proc. 11th Australian Research Assembly on Brassicas, Perth, pp 79-83.

Potts, D.A., Rakow, G.W. and Males, D.R. (1999).  Canola-quality Brassica juncea, a new oilseed crop for the Canadian prairies.  In: Proc. 10^th International Rapeseed Congress, Canberra, Australia, September 23-29, 1999 (in press). 

Stoutjesdijk, P.A., Hurlstone, C., Singh, S.P. and Green, A.G. (1999).  Genetic manipulation for altered oil quality in Brassicas.  In: Proc. 10^th International Rapeseed Congress, Canberra, Australia, September 23-29, 1999 (in press).