Managing N in Variable Paddocks – Application of N Bank Strategies Part 2

Take Home Messages

• Variable rate applications of nitrogen (N) to a 125kg N/ha N bank in 2023 resulted in equivalent or greater yields in comparison to a theoretically applied blanket application approach.
• Applying the N Bank theory with the grid soil sampling technique resulted in fairly uniform yields across both demonstration paddocks in 2023. Longer term analysis is required to validate if this is an economical approach for applying this N management strategy.
• Grid soil sampling for nitrogen applications was still a profitable option in 2023, but is better suited to more costly inputs, such as gypsum or lime, and where time between sampling is greater than one year.

To look at strategies that quantify and optimise the use of sub-paddock level soil sampling and the application of variable rate (VR) technology for nutrient management to maximise productivity while minimising mining of key soil nutrients.

To further validate the N Bank nitrogen management strategy.

Australian wheat yields are only half what they could be for the rainfall received (Hochman et al. 2017). This is also likely to be true for other non-legume crops (barley, canola and oats). Nitrogen (N) deficiency is the single biggest factor contributing to this yield gap. Alleviating nitrogen deficiency has the potential to increase national wheat yields by 40 per cent (Hochman and Horan 2018). 

On farms with no legume pastures, most of the crop nitrogen supply must come from nitrogen fertiliser. Grain legumes do not provide enough nitrogen to support the yield of subsequent crops at the intensity at which they are currently grown. With nitrogen fertiliser a costly input and a lack of accurate seasonal forecasts for rainfall at the decision-making time, it can be difficult to match nitrogen supply to crop demand. Growers therefore tend to be conservative with their nitrogen inputs.

To simplify the approach, novel nitrogen management strategies have been explored, including nitrogen banks. The nitrogen banking approach aims to maintain a base level of soil mineral nitrogen that is topped up each season to a target nitrogen bank level that is most appropriate for a given location to achieve the water-limited yield potential in most seasons. Through previous work undertaken by Hunt et al. (2022) this strategy has been identified as a promising approach for managing nitrogen in crop production areas with low environmental losses (such as leaching and denitrification).

In 2022, two paddock scale demonstration sites were established (Angel and Fleay 2022) and maintained during 2023. The first paddock is adjacent to BCG’s pilot nitrogen bank trial at Curyo and the second is on a typical soil type and rotation for the Southern Mallee region, at Warne.

Two demonstration sites were established near Curyo and Warne in the southern Mallee, north-west Victoria in 2022. At the beginning of the 2022 season, grid soil sampling and electromagnetic induction (EM) mapping was undertaken by Precision Agriculture to develop a baseline soil map of the paddocks. The paddocks were divided into a 2ha grid with soil sampling undertaken in February 2022 across the grid at 0–30cm and 30–60cm depths. Lab analyses were used to calculate soil mineral nitrogen (nitrate and ammonium) in the soil profile, along with identifying other soil characteristics including areas of sodicity and high chloride levels in some of the sub-soils. The paddocks were again soil sampled in February 2023 by Precision Agriculture using the same approach except no EM mapping was undertaken. The results from each year were then used to calculate the urea required to achieve a uniform N Bank across the paddock. Given the environment and work from the James Hunt-led N Bank trials, it was identified that 125kg nitrogen would be suitable for both sites as a target nitrogen strategy (note: in 2022 the Warne site received starter fertiliser at sowing and a blanket application of 40kg of N/ha so the N bank of the site had ~170kg N/ha. In 2023 the target was reset to 125kg/ha). This report also compares the economic analysis of variable rate applications to theoretically applied zone and traditional transect sampling of the paddocks for the 2023 season only. To determine the paddock zones, elevation maps from the EM survey conducted in 2022 and host growers’ long-term knowledge were used to divide the paddock into two zones. The soil mineral nitrogen (which is the sum of the nitrate and ammonium nitrogen in the soil) for these areas was calculated using an average of the grid data that landed within each zone. For the transect sampling approach, a diagonal line was drawn across each paddock and where the line intersected with a grid sampling data point (up to six) the data points were averaged to give the soil mineral nitrogen for the paddock. When calculating the yields for both the zone and paddock scale data a 2ha square area was overlaid on the yield map. The square with the same soil mineral nitrogen as was theoretically calculated for the different approaches was then selected and the yield for that 2ha area averaged to provide yield data. Yield data was cleaned (erroneous data points removed) by USDA ARS Yield Editor and analysed using Ag Leader Technology SMS Advanced, version 22.00.

At Curyo, the theoretical paddock scale strategy of soil sampling resulted in 0.2t/ha less yield in comparison to the hypothetical zone and grid sampling average (Table 3). This approach also resulted in less nitroge being applied in season. This shows using a VR approach does not necessarily result in less inputs and thus cost savings, but it can allow for the better placement of nutrients to where they are needed most to generate a better financial return (Table 3). Nonetheless, soils are inherently variable, therefore soil analysis results can be substantially different, depending on the sampling locations within the paddock and zone. The theoretical zone approach at Warne in 2023 resulted in 0.1t/ha more yield in comparison to the hypothetical paddock scale approach and the grid sampling method (Table 4). Consistent with Curyo, the paddock scale method however resulted in the least amount of urea applied in season but equivalent average yield to the grid sampling approach (Table 4). The lower application of urea from the paddock scale approach is possibly due to both paddocks being variable in their soil mineral nitrogen content. The overall paddock average is likely to have smoothed out some of the extreme areas resulting in the under and over application of urea. Grain quality, specifically protein, was not measured in both paddocks. For future research it would be recommended to also capture this data as the different soil sampling approaches and nitrogen applications would likely have an impact on grain quality and economic return.

Applying the N Bank approach with the grid soil sampling technique resulted in 79.1 per cent of yield in the Curyo paddock and 65.9 per cent in the Warne paddock, a difference of less than 1.0t/ha (Figure 3 and Figure 6). Considering the N Bank approach is a long-term strategy, some yield variation is to be expected after only two years of implementation. Further monitoring over several years is needed to properly assess this nitrogen management strategy at the paddock level. Note this also assumes that nitrogen is the main driver impacting yield here.

In short, it depends. Knowledge of a paddock’s variation and the factors creating the soil and production variability is needed before the best soil sampling method can be chosen. For example, if high and low yielding areas of a paddock are observed and the characteristics driving these differences understood, then zone soil sampling is an appropriate option. Paddock Paddock scale is the traditional method of soil sampling which typically involves taking multiple cores either randomly or along a transect across a paddock. A composite sample is then sent off to a soil laboratory for analysis. Whilst this approach can be cost effective it ignores the inherent variability of soils and inefficiencies are likely to occur when making fertiliser applications (Condon 2019). For example, if an alternative transect had been sampled at Curyo as demonstrated in Figure 7 a 12kg N/ha difference in average soil mineral nitrogen would have occurred. A blanket application of 35kg N /ha across the paddock instead of 22kg N/ha would have been applied, resulting in an over application of nitrogen in some parts of the paddock and under application in others

Figure 7. Grid sampling maps for soil mineral nitrogen at Curyo, 2023. Left: transect used in this article. Right: alternative transect.

Zone Sampling In recent years there has been increased adoption of a zone soil sampling approach. Zones are typically identified by production areas (high or low biomass, high or low yielding) or soil properties (colour, texture and slope). With the increase in PA technologies, zones can also be identified using elevation maps or EM survey mapping work. Once the zone has been identified, each zone can be sampled multiple times to produce separate composite samples for each zone. All these tools to identify zones are low to no cost which can help make sure nutrients or soil amelioration strategies can be provided where it is needed most (Condon 2019). In-paddock sampling and laboratory analysis are however more expensive than the paddock scale approach. Grid Sampling Grid soil sampling is considered the most accurate and robust method for determining spatial variability in soils and mapping the spatial variability present at the sub paddock level. The cost effectiveness of this practice, however means it is important to know when it is suitable to use, and when it becomes just another cost. For longer term inputs such as lime or gypsum, grid sampling in combination with EM mapping and development of refined VR application to this level can be cost effective as sampling may only need to occur every few years in line with the life of the amelioration. For more mobile and seasonally variable nutrients such as nitrogen, sampling needs to occur more frequently, ideally annually. There is also increased time involved with grid sampling as well as handling/preparing samples for postage and data return. Commercially available grid soil sampling services exist to remove the time and labour factor involved, but these come at a cost. Grid sampling for nitrogen might still have a place, however as the paddock demonstrations at Warne and Curyo have shown. In paddocks with variable nitrogen, grid sampling could be justified as a way to better target resources, particularly if an N Bank strategy is being adopted. More research is needed to confirm this.

Angel K., and Fleay B., 2023, 2022 BCG Season Research Results ‘Managing N in variable paddocks: application of N bank strategies part 1’ pp 71–77. Condon J., 2019, GRDC Update Paper ‘Effective soil sampling – high and low-cost options to gain soil fertility information for management’, <https://grdc.com.au/resources-and-publications/grdc-update-papers/tab-content/grdc-update-papers/2019/02/effective-soil-sampling-high-and-low-cost-options-to-gain-soil-fertility-information-for-management> [accessed on 24 January 2024]. Hunt J., Kirkegaard J., Maddern K., and Murray J., 2022, GRDC Update Paper, ‘Strategies for long term management of N across farming systems’, <https://grdc.com.au/resources-and-publications/grdc-update-papers/tab-content/grdc-update-papers/2021/02/strategies-for-long-term-management-of-n-across-farming-systems> [assessed on 18 January 2024]

This project is supported by BCG, through funding from the Australia Government’s National Landcare – Smart Farms Small Grants: Soil Extension Activities Program.

Thank you to DeAnne Ferrier for providing a technical review of this article.

We thank the Barclay and Warne families for hosting these paddock demonstrations