Take home messages
- In 2016 and 2017, mid-row banding in-season consistently increased the proportion of fertiliser N recovered in the grain and straw at harvest compared to other methods of application.
- Mid-row banding increased grain yield compared to the industry standard at one of six trial sites and increased grain protein by 0.7-1.6 per cent (%) at two sites.
- Mid-row banding may offer a range of benefits compared with current methods of nitrogen application, but these should be considered against the associated capital and operational costs.
Background
Nitrogen (N) fertiliser is one of the largest single inputs to modern cropping systems. Variable seasonal conditions, however, make it difficult to predict the optimal time and rate to apply N to the crop. Combined with the logistical considerations that come with large scale grains production, the timing of application can be equally hard to manage. As a result, nitrogen use efficiency is often low with less than half (44%) of the N fertiliser applied to dryland crops being recovered in the grain and straw at harvest, approximately 34% remaining in the soil for subsequent crops and 22% being lost to the environment (Angus and Grace, 2017).
Mid-row banding; where fertiliser is applied between every second crop row has been shown to improve recovery of applied N by wheat in the Wimmera and Mallee (Wallace et al. 2018). In some cases, it has also resulted in increased yield or grain protein compared with other methods of application, but this response has been more variable, and several questions remain. While further testing of mid-row banding is needed to better understand where and when it works, practical considerations such as the ability to handle high stubble loads and narrow row spacings have also arisen. One option to enable mid-row banding in such situations is the use of ultra-high-pressure liquid injection. In this case, liquid fertiliser is applied at pressures of over 50,000 psi, cutting though stubble and injecting the fertiliser into the soil, potentially alleviating some of the issues associated with conventional disc openers.
Aim
To assess the benefits of applying N fertiliser to wheat and canola at either sowing or during
the season by mid-row banding compared to conventional methods.
Paddock details
| Location: | Ouyen and Longerenong |
| Crop year rainfall (Nov-Oct): | Ouyen – 200mm, Longerenong – 284mm |
| GSR (Apr-Oct): | Ouyen – 102mm, Longerenong – 187mm |
| Soil type: | Ouyen – Calcarosol, Longerenong – vertosol |
| Starting soil N (0-1.2m): | Ouyen – 56kg/ha, Longerenong – 101kg/ha |
| Paddock history: | 2017 Ouyen – wheat, Longerenong – barley |
Trial details
| Crop type/s: | Ouyen – Kord CL wheat, Longerenong – 44Y90 canola |
| Treatments: | Refer to methods |
| Target plant density: | Ouyen – 130 plants/m², Longerenong – 60 plants/m² |
| Seeding equipment: | Knife points, press wheels, 30cm row spacing |
| Sowing date: | Ouyen – 17 May 2018, Longerenong – 18 May 2018 |
| Replicates: | Four |
| Harvest date: | Ouyen – 4 December 2018, Longerenong – 26 November 2018 |
Trial inputs
| Fertiliser: | Both sites received Granulock® Z @ 60 kg/ha at sowing N applications: refer to methods |
| Chemical inputs: | Managed as per best practice to keep the trial weed and pest free |
| Irrigation: | To avoid crop failure due to dry seasonal conditions, both sites were irrigated twice between 7 September and 5 October 2018 (coinciding with pre-flowering to early grain filling). Irrigation of 20mm was applied on each occasion, noting that irrigation is likely to be around half the efficiency of rainfall. |
Method
Two replicated field trials were sown using a complete randomised block trial design to assess the effect of a range of N application methods on crop performance. In addition to the 7kg N/ha applied as Granulock® Z at sowing, further N was applied as urea and the methods included:
- mid-row banded at 35-50mm below the surface using a twin disc opener (granular and liquid)
- banded at 25mm below the seed (granular and liquid)
- mid-row banded using ultra-high-pressure injection (liquid)
- streaming nozzle (liquid)
- top-dressed (granular)
Applications were made either at sowing or in-season (Ouyen; GS30, Longerenong; GS2.0) except for banding below the seed which only occurred at sowing. Rates of N were 25kg/ha and 50kg/ha for all methods and dates of application. Additional plots receiving either 15 or 100kg N/ha as granular urea at sowing (banded below the seed at Ouyen and top-dressed at Longerenong) were also added to establish the yield potential of the site.
Additional sub-plots were established within the trial to measure the uptake of fertiliser by the crop in response to application method and timing. By using a ‘labelled’ form of urea, it was possible to directly track the proportion of fertiliser N present in the grain, straw and soil at harvest and thereby measure the proportion of fertiliser lost to the environment.
Results and interpretation
At Ouyen, wheat yield and grain protein responded strongly to additional N application, with a
maximum yield of 1.9t/ha despite the dry seasonal conditions occurring in 2018 (Figure 1). Peak yield occurred at an N rate of 25-50kg N/ha before declining at a rate of 100kg N/ha. This response may in part be due to poor establishment at high N concentrations where there was inadequate separation of seed and fertiliser combined with dry soil conditions. Grain protein continued to increase (peaking at 13%) at N rates of up to 100kg N/ha. At Longerenong, canola yield and oil content also responded strongly to N application. Yield peaked at 1.3t/ha and a rate of 100kg N/ha (top-dressed after sowing to avoid seed damage), while maximum oil content was 38% at rates of 15-25kg N/ha.
Figure 1. Grain yield and protein percentage/oil response to N application rate at Ouyen and Longerenong in 2018. Stats: Ouyen yield – P=0.002, LSD=0.35, CV=8.0%, Longerenong yield – P=0.002, LSD=0.10, CV=6.2%, Ouyen protein – P<0.001, LSD=0.96, CV=1.5%, Longerenong oil – P<0.001, LSD=0.85, CV=1.3%.
Method of application had a significant effect on grain yield when N was applied at sowing, and grain protein when N was applied in-season at Ouyen. Where N was applied at sowing, mid-row banding and banding below the seed increased yield by 0.2-0.3t/ha compared to surface application or ultra‑high-pressure injection (Table 1). Given the dry conditions at sowing it is likely that some of the surface applied N may have been lost due to ammonia volatilisation and/or tied-up by soil microbes present in the shallow soil depths, compared to where N was banded below the soil surface. Where N was applied in-season at Ouyen, mid‑row application either by disc opener or ultra-high‑pressure injection, similar yield was maintained to other methods of application, but it significantly increased grain protein by 0.5-1.6%. This result is consistent with those from a similar trial at Ultima in 2017 and may indicate that banding of N between the crop rows is slowing crop access to applied N pre‑flowering and increasing availability during grain filling.
Table 1. Wheat grain yield and protein response to N application method at sowing and in‑season at Ouyen in 2018. Values presented are the mean of both 25kg and 50kg N/ha rates.
At Longerenong, the method of N application had no significant effect on either grain yield or oil content. While N addition did significantly increase yield (+0.1t/ha) and mid-row banding resulted in the highest average yield, this was not statistically significant compared to any other method of application (Table 2).
There were limited differences at Ouyen and Longerenong when comparing N applied as granular urea versus liquid urea for grain yield. Where N was applied in-season at Ouyen, the use of liquid urea significantly increased yield (+0.2t/ha). While at Longerenong, application of liquid urea at sowing significantly decreased yield (by 0.05t/ha), but otherwise differences were not significant.
Table 2. Canola grain yield and oil response to N application method at sowing and in‑season at Longerenong in 2018. Values presented are the mean of both 25 and 50kg N/ha rates.
After undertaking a total of six trials testing the response of mid-row banding against other methods of N application from 2016 to 2018, several key findings have been identified. Firstly, the effect of mid‑row banding on grain yield seems to vary with season, soil type, crop type and time of application.
At three of the six sites mid-row banding showed potential to increase yield compared with other methods of application, however at only one of these sites was it higher yielding than the current industry practice (banded below the seed at sowing or top-dressed granular in-season).
Secondly, at two of the drier sites in the Mallee (Ouyen and Ultima), in-season mid-row banding significantly increased grain protein in wheat (+0.7-1.6%) compared to all other methods of application without reducing yield.
Thirdly, while results from the ‘labelled’ fertiliser experiment in 2018 are not available at the time of writing, results from 2016 and 2017 indicate that in-season mid-row banding consistently increased nitrogen use efficiency (measured by recovery of applied N in grain and straw at harvest) compared to surface applications.
Commercial practice
While mid-row banding has the potential to increase nitrogen use efficiency when compared to other methods of application in 2016 and 2017, this needs to be considered in the context of longer-term profitability. This is particularly the case when assessing its application in‑season which would require significant investment in machinery and likely increases in operational costs compared to current methods such as topdressing. As a result, increases in either yield and or grain quality will be needed to make a return on such investments.
Conversely, use of mid-row banding at sowing may be less capital intensive and have less of an effect on operational costs. In 2018 this option was shown to result in similar yield and quality when compared with the current standard of banding below the seed with the added benefit of reducing the risk of fertiliser toxicity where N rates are high (as observed for the 100kg N/ha rate at Ouyen). While the ability of ultra-high pressure injection to effectively cut through high stubble loads and handle narrow row spacings is promising, commercially available machinery to mid-row band fertilisers is not yet available and in 2018 it was in some cases found to be less effective than using a disc opener system in wide row, low stubble situations.
References
Acknowledgements
This work was funded by the ‘Improving practices and adoption through strengthening D&E capability and delivery in the southern region’, Regional Research Agronomists program (DAV00143) as part of the Grains Research and Development Corporation (GRDC) and Department of Jobs, Precincts and Regions (DJPR), Bilateral Research Agreement.
The authors also wish to acknowledge the assistance of Birchip Cropping Group and South Australian No-Till Farming Association staff in delivery of the experimental program and the growers who have hosted these experiments.
