Back to nitrogen basics – soil testing and nitrogen budgeting fundamentals

Take home messages

• Nitrogen (N) fertiliser rate decisions based on soil test data and a formalised decision process are more profitable than fixed rates or decisions based on ‘gut feel’.
• This article goes back to basics on N budgeting and is designed to help young agronomists make better N fertiliser recommendations.
• The article simplifies a lot of complex topics and should just be a starting point for learning about N management in southern Australian farming systems.

Background

Nitrogen (N) management has a big impact on farm profit, and being able to effectively advise growers on fertiliser N inputs is an extremely important skill for agronomists to have. Fertiliser N makes up a large component of variable costs of cropping and return on investment in N fertiliser is not certain, which makes decisions on N application rate risky. The GRDC Riskwi$e investment aims to help growers and advisors make better decisions where risk is a factor, and the initial focus is on N management decisions. Previous research has shown that choosing an N rate that uses soil test data and a formal decision-making process is far more profitable than ‘gut feel’ or flat application rates.

This article is a back to basics guide to N fertiliser budgeting and is targeted at early career agronomists to help improve decision making. It tries to provide enough science background to assist effective management but simplifies a lot of complex topics. If you want to further understand the complexities, please read the GRDC publication ‘A nitrogen reference manual for the southern cropping region’. Some of the complexities around yield uncertainty are also expanded on the article written by Peter Hayman and Barry Mudge in these proceedings, and the articles are complementary.

Why is N so important for crops?

Nitrogen (N) is the nutrient required in the largest amount by grain crops. Grain crops require about seven times as much N as they do phosphorus (P), which is the next highest nutrient required by mass. Plants use N to make amino acids and proteins, and it is an important constituent of chlorophyll (the pigment that makes most plants look green) and RuBisCO (the enzyme used in carbon fixation), which are both important components of photosynthesis. N deficiency makes plants look less green because they contain a lower concentration of chlorophyll. N deficiency means that plants cannot photosynthesise (turn carbon dioxide into dry matter) as well as N sufficient plants, which means they cannot grow as much. Because crop grain yield is determined by the amount of growth that occurs during the critical period of yield determination, which occurs about 30 days before the start of grain fill in most crop species, N deficiency during this time causes large reductions in grain yield. Conversely, an oversupply of N can in some cases reduce grain yield in cereal crops. This yield reduction is often referred to as ‘haying off’, but the exact mechanisms of yield loss due to excessive N uptake are not known.

N taken up by crop plants is translocated to grains during grain filling to form proteins, and this is why N deficient cereal crops have low grain protein. Wheat or barley with protein of less than 11.5% is likely to have been N deficient and not achieved the best yield that it could.

Because most plant N is translocated to grain, large amounts of N (about 20kg/ha N per 1t/ha cereal yield or 40kg/ha per 1t/ha canola or legume yield) are exported from paddocks in grain at harvest. Sustained crop production requires inputs of N from fertiliser, legumes or organic wastes that equal or exceed offtake in grain. Most N (55 to 70%) is supplied to crops from the soil, but fertiliser can be used to supplement soil N supply, though this can usually only provide 30 to 45% of plant N uptake at the most. It is best to view fertiliser as a replacement for soil N that has been removed in grain, rather than the primary resource of N for crop growth.

Commercially, N deficiency can have a big impact on farm profitability. Ensuring that crops have sufficient N to achieve water limited potential yield, but not excessive amounts of N, can often mean the difference between profit and loss for a farm business.

What do we need to know to calculate a fertiliser N rate

N budgeting is an effective way to calculate a fertiliser N rate. To calculate a fertiliser rate using an N budget, we need to know the likely N requirement (or demand) of the crop we are managing, and the amount of N that is supplied by the soil. In basic N budgeting, fertiliser N requirement is calculated as the difference between the crop N demand and the soil N supply. When soil N supply exceeds crop demand, no additional fertiliser is required. When crop N demand exceeds soil N supply, additions of fertiliser N are required to avoid N deficiency and yield loss.

Estimating crop N demand

N is different to most other nutrients in that N requirement is proportional to grain yield. To effectively manage N, we need to understand crop yield. The yield concepts of Fischer (2015) are helpful in achieving this and are described below.

Farm yield (FY) – the yield achieved by growers in their fields. This can be measured individually in a sub-field unit (e.g., 3 ha), single field or is often aggregated up into larger areas.

Potential yield (PY) – the measured yield of the best cultivar, grown with optimal agronomy and without manageable biotic (for example, weeds, pests and diseases) and abiotic stresses, under natural resource and cropping system conditions representative of the target area. This is determined by solar radiation and temperature and is a useful benchmark in irrigated production systems and regions with very high rainfall.

Water-limited potential yield (PYw) – the yield obtained with no other manageable limitation to the crop (as for PY) apart from the water supply. This is a more useful metric in rainfed or dryland regions of crop production that commonly occur in Australia, and what we will use in the example at the end of this article.

Economic yield (EY) – the yield attained by growers with average natural resources when economically optimal practices and levels of inputs have been adopted while facing all the vagaries of weather. This metric recognises the law of diminishing returns; as inputs required to achieve high yields (fertilisers, biocides) are increased, returns decrease to the point where they become unprofitable. Economic yield is estimated as 80% of PYw, and this is what we use in the example below to calculate N demand.

In the case of wheat and barley, a robust rule of thumb is that 40kg/ha of N (soil mineral N and fertiliser) is required per 1t/ha of EY to ensure N sufficiency. In canola, approximately 80kg/ha of N supply is required per 1t/ha of grain yield. Only about half of this N supply is taken up by the crop and translocated to grain, the rest remains in the plant residues or soil.

Estimating economic yield

Because N is either applied at sowing and/or in crop, EY is unknown at the time that N application rate needs to be decided. In irrigated systems, or environments with very consistent rainfall, estimating N demand is easy because EY does not vary. In southern Australian dry land systems, EY varies enormously with the regions’ highly variable August–October rainfall. It is this variability that makes it difficult to get N fertiliser rate ‘right’.

There are lots of ways of estimating EY, including with complex crop simulation models like APSIM and its commercial web interface Yield Prophet®. In water limited environments like most of SA, a simpler way is to use known relationships between crop evapotranspiration (water use, WU) and an upper limit of grain yield. This relationship between crop water-use and grain yield was first described in SA by French and Schultz (1984) in their seminal work on water-use efficiency (WUE) and has most recently been updated by Harries et al. (2022) based on commercial crops in WA , where water limited potential yield for different crops can be calculated as follows:

Wheat PYw = (WU – 45)*25

Barley PYw = (WU – 50)*24

Canola PYw = (WU – 80)*15

This method of calculating PYw has a lot of assumptions and simplifications, but it is robust enough to make it useful for N management. EY is simply calculated as 0.8*PYw.

Measuring evapotranspiration is difficult in commercial crops, but it can be estimated from rainfall records, assuming that 25% of rain that falls during the summer fallow period (Nov–Mar), and all the rain that falls during the growing season (Apr–Oct) is used by crops for evapotranspiration (none leaches, runs off or is left behind by the crop). Therefore:

WU (mm) = (0.25*Nov–Mar rain) + Apr–Oct rain

If a decision on N rate is being made in April for the purposes of ordering urea after soil test results are available, Nov–Mar rainfall is known, but Apr–Oct needs to be estimated based on historic records. If a decision on N rate is being made in late July, Nov–Mar and Apr–Jul rainfall are known, but an estimate of likely rainfall needs to be used for Aug–Oct. Because Apr–Oct and Aug–Oct rainfall vary so much, basing a decision on the full range of possible outcomes, rather than assuming an average, can greatly improve decision making (please see the article by Peter Hayman and Barry Mudge in this proceedings). However, long term experiments have shown that simply using median rainfall for future months when estimating WU results in highly profitable N rate decisions is better than ‘gut feel’.

Estimating N supply

Crops take up most of their N (55 to 70%) from the soil. N in the soil exists in two major pools – mineral N and organic N. Most N in cropping soils (tonnes per hectare) sits in the organic pool, which is not available to plants. It is contained in soil organic matter (SOM) with carbon (C) and other elements. Nitrogen cycles from the organic to the mineral pool through the process of mineralisation, and from the mineral pool back to the organic pool by the process of immobilisation. Both processes are the result of soil microbial activity. Mineralisation requires wet and warm soil, and in southern Australia mostly happens in summer when crops are not growing. Some mineralisation happens when the crop is growing (in crop mineralisation) and is highest in wet springs. In some systems N mineralisation and immobilisation are approximately equal resulting in no net change in mineral N availability to the crop.

Mineral N includes nitrate (NO3) and ammonium (NH4) which can both be taken up by plants but are much less abundant than organic N (tens to hundreds of kg per hectare). Nitrate is most readily taken up by plants and is usually the most abundant form of mineral N in the soil. It is also the form of N most readily lost to the environment by leaching and denitrification. Nitrate and ammonium are what we measure in soil tests to estimate soil N supply. Mineral N that we measure in a soil test at the start of the growing season has either not been taken up by the previous crop or has mineralised from soil organic matter or crop residues during the summer fallow period.

Taking effective soil tests

To estimate N supply, effective soil tests are required. These are best taken in the month or so prior to sowing, and before any fertiliser N is applied to the paddock. Mineral N is spatially variable, and estimates are highly prone to error. Multiple cores are taken in a single paddock to try and reduce error.

How many cores should I take

Accuracy of the estimate of soil mineral N increases with the number of cores that are taken. Most operators take 6–8 cores within a production zone and bulk them before sending for analysis. This typically gives a reasonable probability (~80%) of being within 20kg/ha N of the true mean. This is usually good enough for commercial N management. If bulking cores, it is extremely important to mix the soil very well before sub-sampling. There is a big advantage in not bulking cores and analysing them separately, as this avoids having to mix cores and helps to understand the paddock variability better. However, it increases the number of samples and the costs of analysis.

Where should I take cores

In a uniform paddock with a uniform yield map (I’ve been told they exist), a transect across the paddock is the best sampling strategy, avoiding any headlands, areas within 60m of trees, or unusual features. In paddocks with obviously variable soil types or topography, or variable yield maps with consistently high and low yielding zones, it is best to divide them into different production zones and sample and manage them separately. It is best practice to GPS locate your sampling points and return to the same location each year.

How deep should I take cores

Ideally, cores should be taken to maximum rooting depth of crops, which is usually at least 1.5m in SA and can be much deeper, particularly for canola. However, taking cores this deep is practically difficult, and most N is concentrated in the surface layers of soil, so an acceptable compromise is to sample to 0.6–1.0m depth. If you do not sample to full rooting depth, you are underestimating soil N supply.

How should I segment my cores

A big increase in accuracy of soil tests can be achieved by segmenting cores into different layers. This is because mineral N concentration decreases greatly at depth, so segmenting avoids mixing soils with very high and very low concentration, which is prone to sampling error in the laboratory. If sampling to 60cm, as a bare minimum, segment 0–10cm, 10–60cm and bulk the cores within these segments and analyse separately. It is even better to segment 0–10cm, 10–20cm, 20–40cm, and 40–60cm.

How should I handle my cores

Soil samples need to be always kept cool and arrive at the lab as quickly as possible or mineralisation will occur and inflate the amount of mineral N in the sample. Keep samples in an esky in the field, and transfer to a fridge or cool room before sending via express post or courier. Send samples early in the week so they are not stuck in transit over the weekend. You can tell if mineralisation has occurred if NH4 concentration is higher than about 2mg/kg. If this is the case, ignore NH4 in the estimate of soil N supply and just use NO3.

Calculating N supply

The number you get back from the soil test is nitrate and ammonium concentration in mg/kg. To convert this to kg/ha of N, you need to multiply by soil bulk density and the depth of the soil that was sampled. An example of how to do this is provided below.

In-crop mineralisation also supplies N to the crop, but I prefer not to include this in estimates of N supply because it is highly variable (can be negative in a dry year) and difficult to estimate, and allowing for it in N budgets contributes to mining of soil organic N.

Putting it all together

The following is an example of calculating a fertiliser N rate using the N budgeting approach and the estimate of PYw estimated by Harrieset al. (2022). It is for a hypothetical wheat paddock in the MRZ of SA that is sown on time and has no other agronomic constraints. The calculation of the N rate is being made at the end of July. Rainfall for the previous season and year to-date is in Table 1, median rainfall for the future months of August, September and October are in Table 2.

Table 1: Recorded rainfall for the summer fallow period and first 4 months of the growing season for the example wheat paddock in the medium rainfall zone of SA.

Table 2: Long term median rainfall for the final three months of the growing season for the paddock location using data from the nearest BoM website for a representative BoM weather station.

Crop water use (rounded up to the nearest mm) is then calculated as:

WU (mm) = (0.25*Nov–Mar rain) + Apr–Oct rain
WU (mm) = (0.25*(72 + 43 + 18 + 3 + 23)) + (37 + 35 + 60 + 25) + (49 + 42 +32)
WU = 320mm

The rainfall numbers in the equation above in bold are measured rainfall for Apr–Jul for the current year (Table 1). The numbers in italics are median values for Aug, Sep and Oct taken from long term rainfall records (Table 2).

Wheat PYw is then calculated as:
Wheat PYw (kg/ha) = (WU-45)*25
Wheat PYw (kg/ha) = (320-45)*25
Wheat PYw = 6875 kg/ha

Wheat EY is then calculated as:
Wheat EY (kg/ha) = PYw*0.8
Wheat EY = 5500 kg/ha

This can be converted into t/ha by dividing by 1000.

Wheat EY = 5.5 t/ha

Crop N demand is then calculated as:

N demand (kg/ha) = 5.5*40
N demand = 220 kg/ha N

Based on the assumption of median rainfall for Aug–Oct, this crop will require an N supply of 220 kg/ha to not be N deficient and achieve economic yield.

The soil test results for the paddock are in Table 3 assuming a segmented soil test to 1m.

Table 3: Soil test results for the paddock from cores taken prior to sowing.

Soil mineral N at each depth is calculated as:

Mineral N (kg/ha) = (NO3 in mg/kg + NH4 in mg/kg)*bulk density (mg/m3)*depth increment
(where 1 decimetre, dm = 0.1m = 10cm)

An estimate of different soil bulk densities is provided in Table 4. For this example, we will assume a loam with bulk density of 1.3 mg/m3.

Mineral N 0–10 cm = (12 + 2) * 1.3 *1     = 18 kg/ha
Mineral N 0–10 cm = (6 + 1) * 1.3 *3        = 27 kg/ha
Mineral N 0–10 cm = (3 + 1) * 1.3 *3        = 16 kg/ha
Mineral N 0–10 cm = (1 + 1) * 1.3 *3        = 8 kg/ha

Total mineral N for the soil profile down to the sampling depth is calculated by summing all the depths:
Total mineral N to 1m = 18 + 27 + 16 + 8
Total mineral N to 1m = 69 kg/ha

Table 4: Range and average of bulk density for soils of different texture class.

Crop N supply is calculated as:

N supply (kg/ha) = total soil mineral N + N in fertiliser applied to-date (80kg/ha MAP in this example at 10% N)
N supply (kg/ha) = 69 + 8
N supply = 77 kg N/ha
N fertiliser requirement is calculated as:
N fertiliser requirement (kg/ha) = crop N demand – soil N supply
N fertiliser requirement (kg/ha) = 220 – 77
N fertiliser requirement = 143 kg/ha N

To calculate a urea rate, divide this number by 0.46 (which is the proportion of urea that is N) = 311 kg/ha urea.

Final words

This is a basic and simplified approach to N budgeting, but evidence has shown that it is effective at calculating N fertiliser rates that are highly profitable over the long term. Many growers and agronomists use other methods of N budgeting that are just as valid and just as effective.

Questions often get asked about the validity of the approach when crops achieve high yield when a soil test indicates that the crop should have been N deficient. The reason for this can be found in the simplifications used in the technique which can underestimate soil N supply. More soil N is available below sampling depth, soil tests can easily be out by 20–40kg/ha N or more if not done well, and in wet springs, crop mineralisation can supply the crop with >80kg/ha N in soils with high soil organic matter. Continually growing N deficient crops not only reduces profitability, but also mines soil organic matter, releasing CO2 into the atmosphere, damaging soil structure, and reducing the soil’s ability to supply N to a crop.

The future

Researchers in Riskwi$e are evaluating new ways of deciding N rates, including N banks (which still require a soil test) and data from header mounted protein and yield maps, and more information about these will be available in the future from the GRDC Riskwi$e investment.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the author would like to thank them for their continued support.

References

Fischer RA (2015) Definitions and determination of crop yield, yield gaps, and of rates of change. Field Crops Research 182, 9-18.

French R, Schultz J (1984) Water use efficiency of wheat in a Mediterranean-type environment. I. The relation between yield, water use and climate. Australian Journal of Agricultural Research 35, 743-764.

Harries M, Flower KC, Renton M, Anderson GC (2022) Water use efficiency in Western Australian cropping systems. Crop and Pasture Science 73, 1097-1117.

A nitrogen reference manual for the southern cropping region

Contact details

James Hunt
School of Agriculture, Food and Ecosystem Sciences, Faculty of Science
Level 5, Building 184, Royal Parade
The University of Melbourne, Victoria 3010 Australia
0428 636 391
james.hunt@unimelb.edu.au