Tactical use of nitrogen in canola to manage risk and include break crops in northern Wimmera

GRDC-WALLUP AG GROUP TRIAL
Canola nitrogen rate and timing

Trial location: “Scott’s”, Ailsa Road, Wallup, Victoria
Trial co-operator: Ian Schmidt

Felicity Pritchard and Rob Norton

Key messages
· A trial was undertaken to determine if nitrogen rates for canola can be reduced in the Wimmera without loss of yield. It also aimed to determine how late nitrogen can be topdressed.

· Nitrogen rate is more important than timing.

· A total 80 kg N/ha/tonne of grain yield – the popular ‘rule of thumb’ has stood up as the best treatment over a range of timings. However, results from this trial suggest a lower rate (60 kg) can provide similar yields as long as it is applied early.

· Canola can respond to late applied nitrogen (early flowering) but less efficiently than earlier applied nitrogen in a season with a dry spring.

Background and aims

Nitrogen fertiliser is often the single biggest variable cost to canola growers in the northern Wimmera.

Throughout the drought, canola was commonly dropped from the rotation in the region as it was seen as a risky crop. To counter some of this risk, some growers changed from pre-drilling all nitrogen fertiliser to splitting or deferring all nitrogen fertiliser applications.

Current recommendations suggest deferring N to the eight leaf stage is possible so long as reasonable amounts of N are present in the soil at sowing time, with rates far more important than timing. Deferring application allows growers to better assess seasonal conditions and therefore yield potential before committing to N application.

Research in CW NSW in wetter seasons the 1990s showed N could be applied at flowering and still provide yield response in a good season. The wet season of 2010 created many questions from growers about how late N fertiliser could be applied to canola.

In addition to timing, the ideal rate of N is queried as well as the efficiency with which N is used when applied early or later in crop growth.

Canola seed typically contains 40 kg N per tonne of grain, and a normal nitrogen use efficiency is around 50%. Therefore, current recommendations suggest canola requires 75 to 80 kg N from all sources to produce each tonne of grain. However, more recent trials by IREC/Better Oilseeds and demonstrations by Topcrop State Focus in 2001 suggest that a figure of 60kg N would probably suffice, particularly for higher yielding crops. The exception is in waterlogged situations.

The trial aims to determine the optimal timing of N fertiliser and the rates for the northern Wimmera.

Site details
Paddock history: 2010 Barley. April 2011 stubble was burnt and harrowed. In May 2011, the paddock was prickle chained immediately following trifluralin application.
Sowing date: 20 May 2011.
Sowing rate: 5.6 kg/ha; seed size 212,900 seeds/kg.
Variety: Pioneer 44Y84
Plot size: 1.74 m x 20 m (6 rows)
Row spacing: 290 mm
Herbicide and insecticides:

Date	Crop stage	Product	Rate (per ha)
18 May 2011	Bare earth, incorporated	Lorsban
	Bare earth, incorporated	Treflan®	1.5 L??(Boxhead)
20 May 2011	Post-sowing, pre-emergence	Lorsban	Check with peter
		Endosulfan
		Mouseoff (zinc phosphide)	1 kg
7 July		Mouseoff (zinc phosphide)	1 kg
Ca. 27 July	Post emergence	Intervix Select	boxhead
15 August		Mouseoff (zinc phosphide)	1 kg

Soil type: Wimmera grey cracking clay (vertosol).
Soil pH_Ca: 8.30 (water), 7.70 (CaCl₂).
Colwell P: 22 ppm
EC: 0.19 dS/m
ESP: 1.3%
Soil available N: 3.9 mg/kg nitrate and 0.9 kg/ha ammonium, equating to 34.8 kg/ha available N in top 60 cm at sowing (17 kg/ha N in top 10cm).
Estimated in-crop mineralisation 40.8 kg/ha N.
Fertiliser: 50 kg/ha triple super (20.7% P) at sowing plus varying urea rates.
Plant available water: 45.6mm to 60cm depth; 71mm to 100cm depth in early May 2011.
2011 growing season rainfall: mm (1 April – 31 October 2011).
Windrowed: 5 November 2011
Harvested: 22 November 2011.

Method
Sowing:
The trial was sown dry two days prior to rain into cultivated soil with Incitec Pivot/University of Melbourne small plot cone seeder with trailing press wheels. Urea was drilled below the seed at varying rates in the one pass.

Seed was treated with Jockey and Gaucho. The hybrid 44Y84 was chosen as the trial co-operator was using two other Clearfield varieties in the same paddock, and open pollinated early variety (44C79) and a mid-maturing hybrid (45Y82). The 44Y84 gave the opportunity to showcase a new, early maturing hybrid within the same herbicide tolerance group.

Trial design:
A randomised complete block design with three replicates was used to compare three rates of nitrogen with nine timings, as well as control plots which received no urea.

Potential yield calculations:
Potential yield at sowing was calculated as 2.12 t/ha using the modified French-Schultz equation, assuming a water use efficiency of 9.5 kg/ha/mm for a 19 May sowing, (Robertson and Kirkegaard 2005) which is essentially equivalent to a WUE of 11 kg/ha/mm with a 120 mm intercept.

The calculation included plant-available subsoil water to 100cm depth, rather than 60cm, which is used by fertiliser companies and advisers. Soil moisture was measured in early May. Plant available water (PAW) was determined from APSoil (CSIRO) soil property figures for upper and lower limits that were determined at the nearby property of Trevor McRae. PAW to 60 cm depth was estimated as 46mm; to 100 cm depth PAW was 71 mm.

Fertiliser rates:
The three rates of nitrogen (from all sources) per tonne of potential yield were based on the assumptions of:

1. 80 kg N/ha (used by agronomists)

2. 60 kg N/ha (used by DPI Kerang)

3. 40 kg N/ha (low rate for comparison)

Soil mineral nitrogen at sowing was measured as 35 kg N/ha to 60 cm deep, and it was estimated that in-crop mineralisation would contribute an additional 41 kg N/ha during crop growth, The N demand of the crop, based in the three N efficiencies and the yield potential was adjusted for mineral N at sowing and in-crop mineralisation, so that the urea rates applied were:

1. 204 kg urea /ha (2.12 t/ha * 80 – 35 – 41 = 94 kg N)

2. 112 kg urea /ha. (2.12 t/ha * 60 – 35 – 41 = 51 kg N)

3. 20 kg urea /ha. (2.12 t/ha * 40 – 35 – 41 = 9 kg N)

Fertiliser timing:
Timing of application were pre-drilling and topdressing treatments at stem elongation and early flowering, plus combinations of split applications:
1. All urea pre-drilled (20 May)

2. All topdressed at stem elongation (4 Aug two hours before 6 mm rain, with 13mm over four consecutive days; ideal)

3. All topdressed at early flowering (11 September during cool weather, before forecast showers which did not eventuate - but no other rainfall forecast beyond then. Some pods beginning to form on lower stems.)

4. Half pre-drilled and half topdressed at stem elongation

5. Half pre-drilled and half topdressed at early flowering.

Harvest:

The trial was windrowed and harvested with a plot windrower and harvested with a plot harvester and grain weighed on site.

Grain quality:
Subsamples were cleaned and analysed for oil and moisture content with a NIR spectrophotometer, with four to seven readings recorded per subsample, depending variation in readings. The CV for the different readings was 0.01 in 90% of subsamples; 0.02 for the remainder. The oil content was standardised to 8.0% moisture for data analysis.

Data analysis:
Data was analysed with a general Analysis of Variance using Genstat version 14.2.
Treatments included rate, timing and a third “Any N” treatment, which referred to whether or not the plots received any added N, i.e. control vs treated plots. Treatment structure was specified as rate x timing nested within the Any N treatment. Block structure was specified as plot within replicate nested in replicates.

Results

Observations 4 August – first topdressing

· Excellent establishment and with the crop at eight leaf growth stage with buds visible under leaves and some plants beginning to elongate.

· Pre-drilling treatments clearly visible, with high N rates considerably more lush with darker green leaves than plots with no N pre-drilled.

Observations - 1 September, four weeks after first topdressing

· Pre-drilled plots with high and moderate rates continue to look the best, although these rates which were also topdressed early appear to be catching up. The low rate appears nitrogen deficient, as do those plots without N applied.

	Low Rate 10 kg N	Moderate Rate 50 kg N	High Rate 100 kg N
All pre-drilled
Split half pre-drilled, half stem elongation
Split half pre-drilled, half early flowering
All Stem Elongation

Effect of rate

· The addition of at the top rate enabled the presowing estimated yield to be achieved.

· Addition of nitrogen fertiliser compared with control plots increased yield by an average 35 per cent.

· Increasing rates improved yield, with highest yield of 2.12 t/ha achieved with the highest rate. The lowest rate was no different to the control.

· Oil content was declined with higher nitrogen rates. The effect of N rate on oil content was highly significant (p<0.001). However, more importantly, oil yield increased with nitrogen rate due to higher grain yields.

Table 1: Increasing nitrogen rate increased canola yield (t/ha) but reduced oil percentage (t/ha). Oil yield increased with nitrogen rates.

Rate	Mean yield	Oil % standardised to 8.5% moisture	Oil yield (t/ha)
High	2.12	41.9	0.89
Moderate	1.85	42.8	0.79
Low	1.50	43.7	0.66
Control (no added N)	1.36	43.9	0.60
LSD* (5%) when comparing high, low and medium rates	0.15	0.52
LSD when comparing rates with control	0.25	0.90

*NB: The LSD is the least significant difference. Here, when comparing two nitrogen rates, the difference between yields needs to be 0.15 t/ha or more to be significant. Otherwise, the differences may well be a chance finding.

Effect of timing
Timing alone did not affect yield or oil content (p>0.05).

Interaction timing x rate
There was no extra yield response at the moderate and high rates of N, providing the N was supplied – at least in part – by stem elongation. Nitrogen top dressed after stem elongation resulted in lower yields than earlier applications. Pre-drilling all nitrogen produced highest yields for high nitrogen rates only.

Topdressing at early flowering alone was too late to maximise yields, although when rates were high, there was still a yield response - compared with the control.

Table 2: Nitrogen rate and rate x timing interaction affected canola yield.

Rate	Timing
	All predrilled	All topdressed stem elongation	All topdressed early flowering	50:50 split; predrilled & topdressed stem elongation	50:50 split; predrilled & topdressed early flowering	Control (no added N)	LSD (5%)
High	2.11	2.30	1.92	2.10	2.16	1.36	0.33
Moderate	1.80	2.11	1.63	2.03	1.69
Low	1.60	1.37	1.67	1.55	1.32

Low rates of nitrogen sometimes produced higher oil content, with some delayed applications increasing oil content further (Table 3). In contrast, this trend was not evident with the moderate and high rates, where delayed N often reduced oil content compared with pre-drilled N.

Table 3: Nitrogen rate and rate x timing interaction affected oil content (%).

Rate	Timing
	All predrilled	All topdressed stem elongation	All topdressed early flowering	50:50 split; predrilled & topdressed stem elongation	50:50 split; predrilled & topdressed early flowering	Control (no added N)	LSD (5%)
High	42.6	41.8	41.2	42	41.9	44.0	1.16
Moderate	44.1	42.5	42.1	43.1	42.4
Low	43.0	43.2	44.0	44.0	44.2

Oil content tended to follow yields inversely.

Figure 1: Yields and oil content were inversely related. Each dot represents the mean oil content and yield of each treatment, shown in Tables 2 and 3. Despite this, higher grain yields gave higher oil yields.


Discussion

Rates more important than timing

The results reinforce the importance of nitrogen for canola. It also reinforces previous recommendations that canola rate is more important than timing, provided the crop does not become nitrogen limited in its early growth stages.

While the trial found that ‘rule of thumb’ high rates (80kg N/ha/t) was the best rate, growers can achieve equal yields with lower (moderate) rates (60kg N/ha/t) so long as the fertiliser is supplied early enough during crop growth, including a topdress by early stem elongation.
A high yielding irrigated trial conducted by VICC under irrigation found yields of canola no different between the 80 kg/ha/t nitrogen ‘rule of thumb’ rate and a lower 60 kg/ha/t rate. It is possible that lateral movement of nitrogen between plots may have confounded that trial.

VICC canola variety trials under irrigation at Kerang have also yielded above their theoretical nitrogen-limited potential when based on the rule of thumb discussed in the introduction. One explanation may be higher growing season mineralisation rates under irrigation in the warm Mallee region than experienced in the dryland Wimmera. The calculation used to estimate in-crop mineralisation is very rough and needs to be researched further to be more useful to growers. (that is, in-crop mineralised N in kg/ha = organic carbon % x 0.15 x GSR).

Alternatively, or in combination with differences in mineralisation, the rooting depth of canola is assumed to be between 60 cm and 100 cm, with mineral N supply taken to the former and water supply to the latter depth. On these vertosols, roots of canola have been measured down to 1.8 m (Norton and Wachsmann, 2006) so that additional N and water could have been accessed during growth.

Topdressing nitrogen fertiliser during early flowering can still provide a yield response. In 2011, the efficiency of late topdressed fertiliser was lower than earlier applications. The late topdressing occurred before dry conditions. Research at wet sites by Central West Farming Systems in the 1990s showed that canola respond to nitrogen topdressed during flowering as long as the spring was wet.

Crop demand during the season

The site had about 40 kg nitrogen at sowing. The soil organic matter content suggests that around 40 kg nitrogen would be mineralised during the season, based on the rough calculation for estimating in-crop mineralisation.

By the four to six leaf stage, the crop would have extracted between 50 and 70 kg N/ha (about 1 t/ha biomass with 6% nitrogen content). At that stage the control would have been nitrogen stressed, and possibly also the low nitrogen rate at sowing.

By stem elongation, the canola would have around 2 t/ha biomass, with a lower nitrogen content of around 5%. This means the crop would have required around 100 kg N/ha. None of the low nitrogen treatments would have been able to meet this nitrogen demand and so would be nitrogen stressed.

The application of early ‘splits’ helped the moderate rate catch up a little, while the nitrogen applied at the higher rate would have kept the crop growing without nitrogen stress.

By first flower, the better growing crop is around 4 t/ha of biomass with around 4% nitrogen, so that it has demanded 160 kg nitrogen by now. If there was no additional nitrogen supplied, this would meet a 2 t/ha yield potential.

It seems that the predrilled treatment is using the N more efficiently – a placement effect largely, but all the high N treatments – even the late top dressing – still look pretty good. The late topdress recovered significantly compared with the control.

Some Rules of Thumb
Please treat these as guides only, the biggest variable is the efficiency of N use which is a culmination of many complex interactions between soil, plant and climate. The values below are largely taken from the PhD work of the co-author (Norton, 1993) which was conducted in the Wimmera in the late 1980’s:

· Read your crop –even at early flowering if the lower leaves are yellow and dropping – N supply is likely to be limiting.

· Each 25 mm of rainfall has the potential to add an extra 250 kg canola per hectare. To achieve this extra yield an additional 10-20 kg N would be required.

· Each added tonne of grain yield will reduce seed oil content by around 2%

· Response

o With 50 kg N at seeding, and an extra 50 kg N will increase biomass at flowering by 50%.

o With 50 kg N at seeding, and an extra 100 kg N will increase biomass at flowering by 100%.

o With more than 120 kg N at seeding, then there will be little biomass at flowering response to extra N.

· Canola does manage to regulate its growth and yield so that harvest index (proportion of grain yield to total biomass) does not fall under moderate to high nitrogen supply. Frost and high temperatures can cause harvest index to fall though.

· Although N is important, other nutrients – especially S - should be considered.

Conclusion

· The original rule of thumb for nitrogen rates in canola of a total 80 kg N/ha/tonne of grain has stood up as the best treatment over a range of timings. However, results from this trial suggest a lower rate (60 kg) can provide similar yields as long as it is applied early.

· Canola can respond to late applied nitrogen (early flowering) but less efficiently than earlier applied nitrogen in a season with a dry spring.

Acknowledgments

· Ian and Graham Schmidt, trial co-operators;

· Peter Howie, University of Melbourne/Incitec Pivot, for sowing, spraying between plots and harvesting the trial and assisting with windrowing;

· Ashley Purdue, Victorian DPI, for windrowing;

· Mark Schumann, NuSeed for providing use of aspirator and NIR spectrophotometer.

References
Norton RM and Wachsmann N, (2006). Nitrogen use and crop type affect the water use of annual crops in south-eastern Australia. Australian Journal of Agricultural Research, 57, 257-267.

Robertson M and Kirkegaard J (2005) Water use efficiency of dryland canola in an equi-seasonal rainfall environment. Australian Journal of Agricultural Research, 56, 1373-1386

Norton RM (1993). "The effect of applied nitrogen on the growth, yield, water use and quality of rapeseed (Brassica napus L.) in the Wimmera." PhD. Thesis, La Trobe University, Bundoora, Vic. 308 pp.