Budgeting crop nitrogen supply better estimation better decisions

Author: | Date: 18 Feb 2016

 Introduction

Crops require nitrogen (N) fertiliser when the N demand by the crop exceeds the N supply by the soil, provided there is profit after taking account of the grain price and the N cost. Crop N demand and soil-N supply both vary between years and between locations so the best strategy is to commit funds for N fertiliser based on conservative estimates of crop N demand. This means two (or rarely three) N applications during a season. The first application can be at or before sowing, based on a test of soil mineral N. The next is at the start of stem elongation, DC30-31, when the density of shoots provides a good estimation of crop response to N. The rare third application is around the booting stage and depends on a comfortable soil water profile, a confident expectation of rain and a high premium for grain protein. The following equation describes the N balance for wheat. 

Fertiliser N       { [Target grain yield t/ha x ]  –  % N recovery x (Soil mineral N at sowing) }
requirement   = { [target protein % x 2.33   ]                            (+ In-crop mineralisation ) } / % N 
(kg N/ha)          {                                                                                                       } recovery

Crop N demand

The simplest and most frequently overlooked part of a crop N budget is demand, estimated by multiplying the target yield by target protein plus an allowance for the N content of the straw. The equation includes a correction factor of 2.33 that assumes 25 per cent of the above-ground N is contained in straw and also converts protein to N (5.7 units of wheat grain protein per unit of N) and corrects for the units of t/ha and percentage protein. The seed of species other than wheat contain seed protein that contains relatively less N (6.25 units of grain protein per unit of N) so the correction factor should be 2.13. Another part of crop N demand is the amount consumed by animals grazing a vegetative crop. For example, 20 ewes and lambs per hectare grazing a wheat crop for 30 days consume approx. 50kg N/ha

Crop N demand should be updated during the season, taking account of the crop appearance, soil water content and the seasonal rainfall forecast. The French & Schultz equation [Yield = 20(Growing Season Rainfall (GSR)-110), where yield is in kg/ha and GSR is in mm] and crop models provide useful estimates of yield, but local knowledge of the yield and protein history of the paddock is invaluable.

Soil N supply

Soils accumulate available N through mineralisation and may lose some of this by leaching and denitrification. Fertiliser N on the soil surface and urine from grazing animals may also be lost by ammonia volatilisation. There is also transfer of mineral N to organic N by the process of immobilisation, which is not a loss but can lead to reduced mineral N for periods of many months. Table 1 discusses factors that affect N supply.

Table 1: Gains and losses of soil mineral N.
 
 Soil water and temperature
 Mineralisation rises from zero to a maximum when soil water increases from about the lower limit to the drained upper limit.
Mineralisation is zero at mean daily temperature of 5°C (average midwinter on the tablelands) rising to a maximum at about 16°C (mid spring on the plains). The maximum mineralisation rate is about 1kg N/ha/day in the top 10cm for soil containing 1% organic carbon (OC) when soil water and temperature are optimum.
 Soil organic matter  2-3% of soil total N is mineralised during an average year for aar of 500mm in southern Australia. This represents 28-42kg N/ha from a topsoil containing 1% OC.
 Previous legume Grain and annual-pasture legumes fix N biologically and some of that N is removed from the paddock.  Of the amount remaining, 15-30% mineralises and is taken up by the following crop, about 10% for the second, and after that 2-3%. There is more N mineralisation from lucerne residues but it starts slowly, persists longer and more is in the subsoil. Legumes also contribute to yield of the following crop through hydrogen fertilisation. 
 Previous perennial grasses N mineralisation in the first year after cocksfoot and phalaris is greater than after legumes, apparently because subsoil nutrients are recycled to the soil surface. 
 Previous canola
Mineral N accumulation in soil after canola (and mustard) is 40-50kg N/ha more than after wheat during wet summers. 
 Grazing stubbles Grazing increases soil mineral N under stubble. 
 Alternating wet and dry soil Rewetting a dry soil stimulates mineralisation. 
Manure

Maybe a short-term effect, but the added N is usually small – 10t/ha of cattle feedlot manure contains about 120kg N/ha, which is released over decades. The phosphorus (P) content is more significant. 
 Cultivation Soil disturbance gives microbes more access to organic matter and potentially increases mineralisation. However, field measurements show little effect of minimum tillage as practised in Australia.
 Liming acid soil Liming can give a short-lived doubling of N mineralisation rate.
 Summer weeds Dense weed populations (and lucerne) can extract all the soil mineral N. Remineralisation is slow because the soil is dried out. 
 Incorporated straw Incorporated stubble temporarily ties up mineral N until the carbon is lost as CO2. After that the organic N is re-mineralised.
 Flooded topsoil Nitrate is rapidly lost by denitrification from land that is flooded during warm weather. This applies to native soil mineral N as well as fertiliser. Measurements from the Darling Downs in late summer indicate loss of about 2kg N/ha/day.
 Leaching Significant on sandy soil in winter-rainfall areas, especially along the southern coast. Leaching below the root zone unimportant in most of the south eastern cropping region.

The amount of mineral N available at sowing is difficult to predict but can be measured reasonably well with an autumn soil test to 60cm. The amount accumulated during crop growth is more easily predicted but is difficult to measure.  Given the inaccuracies of all soil N measurements, it is best to regard predictions of N requirement as general guides with a range of 20-40kg N/ha.

In practice there are not many tests of soil mineral N (’deep soil nitrate’) in a district during autumn and it is in the interests of advisers to share around the available information and customise it to other paddocks based on the effects in Table 1.

Nitrogen use efficiency

There are many ways to express N use efficiency, most of them misleading. The simplest is yield divided by the amount of fertiliser N applied, which is the most misleading because it ignores the contribution of soil N and the effect of N on grain protein. Another is agronomic efficiency (AE): additional yield from fertiliser divided by the weight of fertiliser, which ignores the effect of N on grain protein. For a cereal with grain protein held constant at 11.5 per cent, the maximum AE is 33kg grain/kg N, but values of 20kg grain/kg N are excellent and a value of about 4kg grain/kg N just covers costs for a grain price of $250/t and fertiliser cost of $1,000/t N.

A better measure is the additional weight of N in grain divided by the amount of N applied, expressed as a percentage.  This still underestimates efficiency because N in grain represents about three-quarters of the above-ground N, but this can vary from about 55 to 80 per cent. Table 2 discusses factors that affect efficiency of N fertiliser. 

On average about half the N fertiliser applied is recovered in the above-ground parts of the crop but the percentages can vary from zero to 100 per cent. The lowest values are in dry conditions when the fertiliser is stranded on the surface or in dry topsoil. Where the N application rate is relatively high (>60kg N/ha), some of the unused fertiliser appears in the following crop. For lower rates, the fertiliser N is probably incorporated into soil organic matter and is released to crops over many years. The highest percentage recoveries measured have been from relatively late applications (mid stem elongation to booting) followed by a wet spring.

Table 2:  What affects efficiency of N fertiliser?

 Crop vigour and health Early sowing and following a break crop increase N efficiency.
 N amount
The higher the N rate the lower the efficiency.
 Time of application No consistent effect on yield, on average, between incorporation before/at sowing and topdressing up to about stem elongation, after which there is relatively more effect on grain protein than yield.
 Form of N fertiliser Nitrate averages about 10% greater efficiency than urea or ammonium.
 Losses  The sooner that fertiliser N is taken up by the crop, the less the immobilisation and losses from leaching, denitrification and volatilisation.
 Rain after application  The more rain the better (within reason) for topdressing on alkaline soil.
 Surface soil alkalinity  The more alkaline the soil, the lower is the efficiency of topdressed urea.
 Grazed crops  Grazing reduces N fertiliser efficiency.

Guidelines for applying fertiliser in-crop

If seasonal conditions are favourable additional N can be applied during crop growth. An accurate indication of crop N status is shoot density measured at the start of stem elongation, DC30. The evidence for this comes from a series of on-farm experiments (Figure 1). The relationship shows mostly positive yield responses when shoot density was less than 500/m2 and mostly negative responses (‘haying off’) when shoot density was greater than 700/m2. Yield responses were more reliable for wheat growing after a break crop than for wheat growing after wheat.

The shoot-number relationship with yield response in Figure 1 applies to medium-maturity wheat varieties. Early-maturing varieties generally produce fewer tillers so the shoot densities that indicate N responses are probably lower than those for medium-maturing varieties. Similarly, late-maturing varieties generally produce more tillers so their critical shoot densities are probably greater. No research has yet been done to quantify the critical shoot densities for early and late varieties.

Figure 1:  Wheat yield response to 40kg N/ha topdressed at the start of stem elongation during three seasons in the Southwest slopes and Riverina, 1987-89. Each data point represents a replicated on-farm experiment with a medium-maturity variety. The seasons are represented by the different symbols; filled symbols represent wheat after a break crop and open symbols represent wheat after wheat.

Figure 1:  Wheat yield response to 40kg N/ha topdressed at the start of stem elongation during three seasons in the Southwest slopes and Riverina, 1987-89. Each data point represents a replicated on-farm experiment with a medium-maturity variety. The seasons are represented by the different symbols; filled symbols represent wheat after a break crop and open symbols represent wheat after wheat.

Shoot density is closely related to the amount of N in the vegetative crop. The percentage foliar cover measured by optical methods such as normalised difference vegetation index (NDVI) provides similar information and can be the basis for a variable application of N fertiliser. It is important to be certain that variation in foliar cover is due to N and not subsoil limitations. Applying additional N could reduce yield where low foliar cover is due to a hostile subsoil.

Novel application methods

Applying a large amount of N fertiliser at sowing leads to excess vegetative growth and increases the risk of haying off in dry conditions as well as lodging and foliar disease in wet conditions. A possible way to minimise these effects is to delay the crop uptake of nutrients from the fertiliser. Granule coatings and nitrification and urease inhibitors delay the release of nitrate in some situations but are expensive. A high concentration of anhydrous ammonia, urea or liquids (aqua ammonia, urea-ammonium nitrate or dissolved urea) in bands between the seed-rows also provides slow release of mineral N to the roots. Generally mid-row bands are between every second seed row so that each seed row is beside only one fertiliser band. The reason for the delayed N uptake is that a high concentration of ammonium in the fertiliser band reduces root activity near the band and delays the microbial conversion of ammonium to nitrate. 

Another application method that is worth considering is side-banding N fertiliser instead of topdressing. With precision guidance, it should be possible to inject urea into the soil beside rows of growing plants. This method should be attractive on alkaline soils where there is a great risk of ammonia volatilisation. It is important to keep urea bands at least 10cm from the rows of plants.

Contact details

John Angus
CSIRO Agriculture, Canberra
02 62465095
john.angus@csiro.au
John Angus Twitter