Background, Site Description and Methods Used

Background, Site Description and Methods Used

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. Defering allows growers to better assess seasonal conditions and therefore yield potential before committing to N application.

Research in Central Western NSW in wetter seasons the 1990’s 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
20 May 2011	Post-sowing, pre-emergence	Lorsban
		Endosulfan
		Mouseoff (zinc phosphide)	1 kg
7 July		Mouseoff (zinc phosphide)	1 kg
Ca. 27 July	Post emergence	Intervix Select
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: 50kg/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).

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 spectrphotomoter, 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.

Reference:
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