ABSTRACT

Assessing the value of canola as a substitute for wheat in crop rotations in Australia depends, amongst other things, on their relative yields. We analyse the impact of nitrogen supply, total water supply and timing of water deficits, and disease on the ratio of canola to wheat yields from both previously published studies and field results collected in the northern region in 1998. In nitrogen responsive situations, the ratio of canola to wheat yield always appears to increase or at least be stable as N rate increases. This increase is exacerbated in situations where wheat “hays off”. As a general rule, the ratio of canola to wheat yield tends to converge towards a value of 0.4-0.6 at high yield levels, and this is confirmed by simulation analysis. Canola and wheat crops grown in the subtropics of northern Australia, a new region for canola production, have ratios of yield similar to that in southern Australia.

INTRODUCTION

Canola is extensively grown throughout the wheat belt of south-eastern Australia and western Australia. However, only small areas are grown in rotation with wheat in northern NSW and southern Queensland. The value of canola as a substitute for wheat in this northern region depends, amongst other things, on their relative yields. Particularly for new growers, there is a need for a rule of thumb to readily assess canola yield prospects, based on a background of experience with wheat yields. Relative yield will depend on many factors, including soil fertility, soil water supply and timing of water deficits, disease and occurrence of frosts and high temperatures.

The aim of this paper is to analyse previously published data in terms of the likely impact of these factors, and to present data on wheat and canola yields from the northern region in 1998. We analyse variation in grain yield (GY) in terms of its two components of total dry matter produced (TDM) and harvest index (HI), to ascertain some physiological basis for the underlying causes. Simulation analysis is used to assess the likely variation in the ratio.

METHODS

Data sources

Yields of canola and wheat crops grown under identical conditions were collated (Table 1). The selected studies covered a range in major agronomic variables such as nitrogen supply, sowing date, and seasonal climatic conditions. The incidence of weeds, pests or diseases was assumed to be minor (except where otherwise stated). Yields were hand-harvested, and adjusted to an oven dry basis.

Simulation studies

In order to assess the likely range of season-to-season variation in the ratio of canola to wheat yields, APSIM with the canola (Robertson et al., 1999) and wheat (Keating et al, 1999; Asseng et al., 1998) modules was configured to simulate the grain yield of canola (cv. Hyola 42) and wheat (cv. Hartog) across the historical climate record at Wagga Wagga (35.2^oS, 147.5^oE). Wagga Wagga, is representative of the high rainfall area of the southern wheat belt, where canola is an established part of cereal-based crop rotations. The model was configured for a 15^th May sowing in each year, a duplex soil of 134 mm plant available water, soil nitrate at sowing was set at 67 kgN ha^-1 and 100 kgN ha^-1 was applied as urea at sowing.

Table 1: Sources of data for the comparison of wheat and canola yields

Author	Location and season	N#	Details of study
Mason 1997	Wongan Hills, Beverly, WA 1994	8	N rates 0-138 kgNha-1
Potter 1997	Lameroo, Pinnaro, SA 1995-6	3	Low rainfall areas of Malee (<350 mm/yr) with boron toxicity problems
Hocking et al 1997	Dirnaseer, NSW 1991-2	10	Sowing date, N rates 0-150 kgNha-1
	Condobolin, NSW 1991-2	10	Sowing date, N rates 0-75 kgNha-1
Scammell 1998	Rutherglen, VIC 1993-5	3
JA Kirkegaard (unpubl.)	Cowra and Cootamundra, NSW 1993-6	7	Varying previous crop

^# Number of treatments

RESULTS

Impact of nitrogen supply

Nitrogen (N) supply and the relative responsiveness of wheat and canola to N supply is one of the major factors influencing their relative yields. Figures 1 and 2 show four published examples where responses of canola and wheat GY to N fertiliser were compared. In the first case (Condobolin, 1991) wheat and canola GY were low (< 1000 kg ha^-1). Wheat responded negatively to increasing N rate, while canola responded positively at low N rates. Consequently, the ratio of canola to wheat GY increased from 0.2 at low N supply to 0.6 at high N supply. This change in the GY ratio with N rate could be explained solely by change in relative TDM, with no clear trend in the ratio of HI with N supply. The second case (Dirnaseer, 1991) also had a negative response of wheat GY to N and a positive response at low N rates for canola. However, wheat GY at 3000-4000 kg ha^-1, was higher than in the first case. As in the first case, the GY ratio increased with N rate, due to parallel increases in the ratio of TDM and HI. The third case (Condobolin, 1992) represents the situation where both wheat and canola GY increase with N rate. In this situation the GY ratio appears stable between 0.4 and 0.5, comprised of a TDM ratio of about 0.5 and a HI ratio of 0.9 across N rates. In the fourth case (Beverly, 1994), wheat GY in unresponsive to N while canola responds to low N rates. Here the ratios of TDM, HI and GY increase at low N rates but then stabilise at higher rates.

The contrasting responses in these four cases highlight a number of points about the ratio of canola to wheat yield in response to N. The ratio always appears to increase or at least be stable as N rate increases. This increase will be exacerbated in seasons where wheat “hays off” and responds negatively to N at high rates. Under haying-off conditions, canola HI will approach or exceed wheat HI, otherwise, it appears that canola HI can be assumed to be around 70-80% of wheat HI. Where the ratio of HI is relatively stable with variation in N supply, it follows that changes in the ratio of GY will be due to variation in the ratio of TDM.

Timing of water deficit

Differences between wheat and canola in the timing of water deficit in relation to crop development would be expected to generate variation in the relative yields of canola and wheat. In addition, the reported deeper rooting of canola than wheat (around 20 cm) (Kirkegaard et al 1997 and unpublished) may indicate a capacity to utilise water from deeper layers although no published data are available. Under a terminal water deficit it may be expected for the ratio to increase, as the earlier maturity of canola relative to wheat would allow drought escape in the former. On the other hand, where water deficit is imposed earlier in the season, the apparently higher sensitivity of canola to water deficit would tend to lower the ratio. Table 2 shows that the ratio of 0.33 in 1994 at Cootamundra, when severe water deficit extended throughout the season produced a much lower ratio than the 0.63 and 0.55 in the higher-yielding 1993 and 1997 seasons.

Table 2: Wheat and canola yields from contrasting seasons at Cootamundra and Cowra in NSW, Australia (JA Kirkegaard, unpubl.)

Location / season	Wheat yield (kg/ha)	Canola yield (kg/ha)	Ratio
Cootamundra 1993	5600	3520	0.63
Cootamundra 1994	1580	520	0.33
Cowra 1997	7500	4100	0.55

Crop growth simulations can give some insight into variation in the ratio of yields from the two crops due to season-to-season variability. Figure 3 shows results of simulations for Wagga Wagga, Australia. Over the 118 years simulated the range in the ratio of canola to wheat yields was 0.39 to 1.11, with a median of 0.49 and a mean of 0.51. There was a significant negative correlation across seasons between wheat HI and the ratio, implying that the yield of canola relative to wheat would tend to be higher in seasons of lower wheat HI (data not shown). Where wheat yields were below 4000 kg ha^-1, the ratio was quite variable, but became less so above 4000 kg ha^-1. The simulations suggest that even under identical agronomic conditions inter-annual variation in seasonal climatic conditions will generate significant variability in the ratio.

Canola v wheat in the northern region

Canola is a new crop to the northern cropping region of Australia, and the yield prospects of the crop relative to wheat, which is widely grown, are relatively unknown. Table 3 shows yields of canola and wheat grown in the subtropics of Australia in 1998.

Table 3: Canola and wheat yields from crops grown in the Australian subtropics in 1998 (JF Holland, MJ Robertson, S Cawley, unpubl.).

Location	Sowing date	Canola yield (kg ha-1)	Wheat yield (kg ha-1)	Ratio canola/wheat
			New South Wales
Gurley	26 May	2800	1440	1.94
Bingara	13 May	2800	2270	1.23
Moree	26 May	3030	1660	1.83
Moree	15 July	1030	1670	0.62
Loomberah	9 June	2810	3370	0.83
Bundella	12 June	2410	4200	0.57
Tamworth	2 June	2520	3010	0.84
			Queensland
Lawes	26 May	3260	6520	0.50
Wallumbilla	12 May	1980	3210	0.62
Muckadilla	12 May	1580	3800	0.42

In a number of cases the yield of canola exceeded that of wheat, caused by leaf and root diseases which depressed wheat yields, as 1998 was one of the wettest on record. In less disease-stricken situations (e.g. Lawes, Bundella, Wallumbilla), the ratio of canola to wheat was similar to that recorded above from sites in southern Australia. It thus appears that in the absence of diseases, canola and wheat crops grown in the subtropics of northern Australia have similar ratios of yield to that found in southern Australia.

Synthesis

A scatter plot of data listed in Tables 1-3 shows that the ratio of yields is highly variable in low yielding situations, as yield level increases, the ratio converges towards 0.5-0.6 (Figure 3). This is also indicated by the simulation results from one location (Wagga Wagga) utilising inter-annual variability to generate a yield distribution of canola and wheat yields. A wide range of abiotic and biotic stresses affect yields in low yielding situations, and the timing and severity of these in relation to crop development and yield formation will contribute to the scatter shown. In more productive situations, differences in phenology and potential biomass production determine relative yields and this will tend to be less variable from situation to situation.

CONCLUSIONS

This analysis has highlighted the following points:

• The yield of canola relative to wheat is highly variable, under sub-optimal conditions, caused by nitrogen deficiency, seasonal climatic conditions, and disease, amongst other things. As a general rule, the ratio is more variable at low yield levels but tends to converge towards a value of 0.4-0.6 in high-yielding situations

• In nitrogen responsive situations, the ratio of canola to wheat yield always appears to increase or at least be stable as N rate increases. This increase will be exacerbated in situations where wheat responds negatively to N at high rates

• In the absence of disease, canola and wheat crops grown in the subtropics of northern Australia have similar ratios of yield to that in southern Australia.

ACKNOWLEDGMENTS

This work was funded in part by the Grains Research and Development Corporation. Our thanks go to co-operating farmers in NSW and Queensland. Shayne Cawthray and Brett Cocks (CSIRO/APSRU) and Bevan Blanch (NSWA) provided assistance in the field.

REFERENCES

1. Asseng, S., Keating, B.A., Fillery, I.R.P, Gregory, P.J., Bowden, J.W., Turner, N.C., Palta, J.A., Abrecht, D.G. (1998). Performance of the APSIM wheat model in Western Australia. Field Crops Res. 57: 163-179.

2. Hocking, P.J., Kirkegaard, J.A., Angus, J.F., Gibson, A.H. and Koetz, E.A. (1997). Comparison of canola, Indian mustard and linola in two contrasting environments. I. Effects of nitrogen fertilizer on dry-matter production, seed yield and seed quality. Field Crops Res. 49: 107-125.

3. Keating, B.A., Meinke, H., Probert, M.E., Huth, N.I, Hills, I. (1999). Nwheat: documentation and performance of a wheat module for APSIM. CSIRO Tropical Agriculture Technical Memorandum (in press).

4. Kirkegaard, J.A., Hocking, P. J., Angus, J. F., Howe, G. N. and Gadrner, P. A. (1997). Comparison of canola, Indian mustard and linola in two contrasting environments. II. Break crop and nitrogen effects on subsequent wheat crops. Field Crops Res. 52: 179-191.

5. Mason, M. G. and Brennan R. F. (1997). Comparison of the nitrogen uptake and yield of canola (Brassica napus) and wheat following application of nitrogen fertiliser. Proccedings of the 11th Australian Research Assembly on Brassicas. Perth, WA, October 6-10, 1997, 110-117.

6. Potter, T.D, Ludwig, I., Kay, J.R (1997). Brassica crops for rotations in low rainfall environments of South Australia. Proccedings of the 11th Australian Research Assembly on Brassicas. Perth, WA, October 6-10, 1997, 128-132.

7. Robertson, M.J., Holland, J.F., Kirkegaard, J.A., Smith, C.J. (1999). Simulating growth and development of canola in Australia. These proceedings.

8. Scammell, G.J. (1998). High yielding crops from legume-dominant pastures. Proceedings of the 9^th Australian Agronomy Conference, Wagga Wagga, 1998, 815-818.

Figure 1: Response of wheat and canola to N fertiliser rate at Dirnaseer and Condobolin, NSW, and Beverly, WA, Australia. Data from Hocking et al. (1997) and Mason and Brennan (1997). (a) Condobolin 1991 (circle), Condobolin 1992 (triangle); (b) Dirnaseer 1991 (diamond), Beverly 1994 (square). Filled symbols are canola and hollow symbols are wheat.

Figure 2: Response of canola relative to wheat across N fertiliser rates at Dirnaseer and Condobolin, NSW, and Beverly, WA, Australia. Data from Hocking et al. (1997) and Mason and Brennan (1997). Data are the ratio of canola to wheat of total dry matter at maturity, harvest index and grain yield plotted as a function of N fertiliser rate.

Figure 3: Scatter plot of canola yield relative to wheat yield plotted as a function of wheat yield. Solid symbols are from datasets listed in Tables 1-3. The outlying points for Gurley, Bingara, and Moree May sowing from Table 3 are not plotted, as they are off-scale. The hollow points are simulated for Wagga Wagga (see previous discussion).