Nitrogen and water use efficiency of canola and mustard in Mediterranean environment of South Australia

Amritbir Riar¹, Glenn McDonald¹ and Gurjeet Gill¹

¹The School of Agriculture, Food and Wine, The University of Adelaide.

Keywords: canola, mustard, nitrogen, water use, water use efficiency.

Take Home messages

• Promoting early growth prior to green bud stage is important to the yield of canola and mustard
• Despite flowering earlier, mustard did not yield better than canola in 2011
• Canola and mustard used the same amount of water but canola yielded more than mustard in the dry year due to strong sink capacity.
• Nitrogen increased crop WUE without altering total water use.
• Nitrogen application increased yields by improving pre-flowering water use which improves the sink development in the pre-flowering period.

Background and Methodology

Canola has become the third most widely-grown broadleaf crop in Australia and the area grown under dryer regions of southern-Australia has increased.  Mustard (Brassica juncea) is a quick developing crop and is often suggested as an alternative to canola in low rainfall areas.  However, canola and mustard yields can be variable. Introduction of hybrid cultivars with greater early vigour and increased use of nitrogen (N) fertilisers have improved canola yields but there is still a large gap between actual and attainable yields. Canola and mustard have high demands for N and low water and N use efficiency; poor use of water and N may be partly responsible for their sub-optimal performance.  Field studies were undertaken at Roseworthy in 2011 and 2012 to investigate the effect of N management on growth, yield, N and water use efficiency of canola and mustard.  Canola and mustard cultivars, with similar maturity but with differences in early vigour, were grown under different N application strategies, namely three N rates (0, 100 and 200 kg N/ha as granular urea). In 2011, a total of 100 kg N/ha was applied in three identical splits at the rosette stage, green bud appearance and at first flower whereas in 2012, a single application of 100 kg N/ha was applied at the rosette stage. A non-limiting N control treatment was used in both years in which a total of 200kgN/ha was applied in five identical applications at rosette, inflorescence, start of flowering, first pod and pod fill to try to maintain a supply of N throughout the season rather than replicate commercial practice.  Growing season rainfall was 227mm in 2011 and 205mm in 2012 but with marked differences in the rainfall distribution (Figure 1). Soil moisture content at 0-100 cm was measured pre-sowing, at flowering and  maturity using a 4cm hydraulic core.

Figure 1. Monthly rainfall of 2011 and 2012

Results

Grain yield, yield per unit of water use and Agronomic efficiency (AE):

Split applications of N improved the grain yield over the low N control. Highest yield of canola and mustard was obtained in the N non-limiting control (Table 1). In 2012, yield of canola was significantly higher than mustard in all treatments whereas canola and mustard were similar in 2011. In both years cultivars showed significant yield differences although results indicated they used similar amounts of water (Table 2). In 2011, AV Garnet, Hyola 575 CL and the Indian mustard variety Varuna were the highest yielding varieties.  In 2012, the highest yielding varieties were AV Garnet, Hyola 575CL as well as Hyola 555TT.  Comparing the two experiments revealed that mustard yields were more sensitive to the drier conditions in 2012.

Water use efficiency was lower in 2012 than in 2011. AV Garnet and Clearfield canola varieties had consistently higher WUEs than the TT canola and the mustard varieties.  There was no consistent difference in WUE between canola and mustard.  Adding N increased WUE (Table 1) without significantly altering total crop water use.

N use efficiency, measured as agronomic efficiency, was greater in 2011.  In both years agronomic efficiency fell at the higher rate of N.  Canola was more responsive to N and had a higher agronomic efficiency than mustard.  There was a trade-off between N and water use efficiency (i.e. adding N increased WUE but decreased nitrogen use efficiency).

Table 1. Grain yield, agronomic efficiency and WUE of canola and mustard in 2011 and 2012.

^A Agronomic efficiency = net increase in grain yield per kg N applied

Table 2. Grain yield and water use of different cultivars.

Sink development and grain yield:

There was a direct relationship between biomass production and grain yield at green bud appearance and the start of flowering (Figure 2), but canola and mustard showed different relationships.   To produce similar yields, canola needed to produce 36% and 45% less biomass than mustards in 2011 and 2012, respectively (Figure 2a). Whereas at 30% flowering stage, canola required 18% and 24% less biomass production than mustard during 2011 and 2012, respectively (Figure 2b).  These results show the importance of early growth up to flower bud appearance to the yield potential of canola and mustard as this helps develop the sink capacity (i.e. pod number).  It also shows that the canola cultivars used in the experiment have a better sink capacity than mustard which helps canola to perform better when compared with mustard in a dryland environment. Even in a dry season mustard was able to produce higher biomass than canola at early growth stages but failed to convert that into grain yield. It showed that mustard relies more on post flowering development in water limited conditions.

Figure 2: The effect of early dry matter accumulation on yield (a) at inflorescence (b) at flowering.

Water use (WU) and water use patterns

The total water use and the patterns of water use varied considerably between the two years.  In 2011, crops used about 150mm more water than in 2012.  Approximately 70% of the total crop water use occurred prior to flowering in 2011, whereas in 2012, only 20-30% of crop water use occurred before flowering. There was no difference in total water use among the varieties but there was a difference in the patterns of water use (Table 2).  The mustard varieties tended to use proportionately less water prior to flowering than canola, especially in 2012, which reflected their shorter time to flower. 

The N treatment did not alter total crop water use but changed the distribution of water use. (Figure 3).  More water was used during the pre-flowering period as the N rate increased especially in 2011 when more water was available at sowing (Figure 4)  In 2012 total water use differed considerably among nitrogen treatments but not in cultivars. All cultivars showed identical responses to the application of 100kg N/ha at rosette stage.

Figure 3: Pre-flowering and post-flowering water use with different N treatments

In both years crops grown with 200 kg N/ha dried the profile more by maturity compare to no nitrogen treatment (Figure 4). In 2011 water use is limited to 50 cm without nitrogen at flowering (Figure 4a and Figure 4b) whereas with nitrogen added, water use was seen up to 100cm.  In 2012 the soil profile at sowing was much drier than in 2011 and moisture accumulated up to flowering in both nitrogen treatments.  At maturity crops grown with the higher rate of N dried the profile more and especially below 50cm as compared to the control.

Figure 4: Water use patterns in 0-100cm soil profile in 2011 and 2012 (a,b) with no nitrogen (c,d) with 200 kg N/ha.

Conclusion

This study of different canola and mustard cultivars showed the improvement in yield with added nitrogen. Adding N increased subsoil water use which was associated with greater water use efficiency. However seasonal conditions had marked effect on the pattern of water use. Early growth at inflorescence and up to flowering was important to develop a good sink and hence grain yield in canola but greater early growth of mustard did not improve the yield in dry season as it seems that rapid early growth of mustard restricted the sink capacity.

Contact details
Amritbir Riar      
GE07, Waite Campus, The University of Adelaide, Glen Osmond, PBM 1, SA 5064
amritbir.riar@adelaide.edu.au