Monitoring of soil phosphorus potassium and sulphur in the southern region how to get the most out of your fertiliser dollar

Author: Sean Mason (University of Adelaide) and Roger Armstrong (Department of Economic Development, Jobs, Transport and Resources) | Date: 23 Feb 2016


For many grain growers in southern Australia poor seasonal rainfall in 2015 produced poor and even ‘failed’ crops. But what does this previously dry year mean for fertiliser requirements in 2016 and can we obtain any returns from investments in fertiliser made in 2015?

The following outlines current guidelines in order to obtain the best monitoring practices that will accurately determine fertiliser requirements for phosphorus (P), potassium (K) and sulphur (S) in the southern region.


Phosphorus management it is still one of the most difficult nutrients to manage. Various sources of information have suggested that continued application of P fertiliser over several years will produce P levels sufficient to meet most crop requirements and this presents an opportunity to reduce rates of P fertiliser to either replacement levels or even lower. On particular soil types this management practice potentially could drive P levels down to levels that could start to impact on yields. These particular soil types typically have moderate P fixation ability as measured by a phosphorus buffering index (PBI) soil test. Generally in the southern region soils with PBI values above 100. For example, Calcarosols have the ability to fix substantial amounts of applied P.

Due to the inability to efficiently apply sufficient P to during the growing season there is a high emphasis on soil monitoring in order to assess P fertiliser requirements prior to the start of the growing season. The following are brief guidelines that will assist in obtaining the most accurate soil assessment of P, and therefore a fertiliser P recommendation.

Soil sampling

Due the immobility of P, poor soil sampling practices are the primary cause of inaccurate assessment of P availability. In most soil types residual P from previous inputs is restricted to the point of application and results in significant stratification of P both horizontally and vertically (Figure 1). Therefore it is advised that soil cores are taken in locations that best represent the area of soil where the current crop row will be placed, especially in no-till systems.

A example of residual Phosphorus in soil

Figure 1: Distance of P diffusion (shaded area) away from a point source (granule) applied to a low P fixing alfisol (left) compared to a high P fixing Calcarosol (right). (Source: Grijalva et al. unpublished)

Soil testing

PBI (phosphate buffer capacity): This quick and cheap test is often overlooked on a soil report but provides useful information and a guide to the efficiency of applied phosphorus to a particular soil type. Based on soil inherent properties the PBI values represent the ability of a particular soil type to remove phosphate from solution. Typically in the southern region PBI values range from 0 (‘gutless’ sands) to 250 (highly calcareous soils); however Tasmanian Ferrosols can have PBI values up 1000. In the southern region PBI values above 80 are classified as moderate to high. A recent survey of soil test results has shown that low soil test values are typically associated with soils above these PBI values, possibly indicating inadequate replenishment of the P that has been exported as grain in addition to lower fertiliser efficiencies.

Soil P availability: In the southern region either the Colwell P or DGT P methods are used to assess the availability of P in a soil sample to a crop. It is recommended that a PBI measure accompanies the Colwell P value in order to obtain an indication of the critical Colwell P value as these can vary with different soil types (Table 1). DGT has shown to provide an improve estimate of P availability on calcareous soils and preliminary data also suggests this test may also be useful on acidic soil types with high PBI values. Critical values are freely available from accredited lab providers or from the soil quality factsheet.

Table 1: Soil P analysis of four phosphorus responsive trials performed in South Australia between 2014 and 2015.
  Colwell P PBI Critical Colwell P DGT P
  mg/kg   mg/kg ug/L
Site 1 11 39 19 16
Site 2 31 135 32 14
Site 3 15 54 22 14
Site 4 22 97 28 17


Returns from P fertiliser inputs can vary with season, sowing time and soil type. Recent research (SAGIT projects: UA1201 and UA1115) has assessed P requirements of various wheat and barley varieties on South Australian soils with low to high fixing levels (PBI 39-135) (Table 1). With similar starting soil P levels (DGT P) response profiles generated from wheat yield with P inputs is strongly related to the PBI of each soil (Figure 2). Optimal P inputs for sites one and three were 21 and 20kg/ha respectively corresponding to PBI values of 39 and 54. In contrast optimal P values for sites two and four exceeded 50kg/ha. Based on the economic returns simply using the returns from yield minus the cost of fertiliser inputs, economic P rates were comparable between all sites and ranged from 16 to 28kg/ha (Figure 3). The lower effectiveness of fertiliser inputs on higher fixing soils has meant that yield increments per unit of P were not high enough at the greater rates of P to be economical even though optimal yields had not been reached. Importantly, applying enough P to maximise yields may not be the most economic P rate for selected soil types and climatic scenarios. In scenarios where soil P is deficient inputs of fertiliser P can still be economical under current grain and fertiliser prices in drought conditions as outlined by site three.

A line graph showing response from wheat yields with increasing rates of applied Phosphorus

Figure 2: Response curves generated from wheat yields obtained with 5-6 increasing rates of applied P (MAP, TSP) at four sites in South Australia.

A line graph showing net profit of crop yield per hectare and phosphorus rate per hectare.

Figure 3: Net profit ($/ha) with fertiliser inputs calculated by subtracting the fertiliser price at each P rate from the returns generated by the grain yields obtained. Fertiliser cost = $700/t, wheat price = $280/t.

What happens to P in a drought year?

Dry conditions causing low grain yields in 2015 would have resulted in less P removal in grain from the applied fertiliser and residual P pools (assuming that crops were not cut for hay). In theory this provides opportunities in 2016 to utilise more of the residual P left in the soil. In most soil types dry conditions reduces the rate fixation/absorption reactions of fertiliser P products with the soil. However in calcareous soils P efficiency can be reduced in dry conditions due to increased precipitation reactions generated from higher P concentrations immediately around the granule. Capitalising on 2015 residual P fertiliser bands requires careful consideration of plant access. Due to the reduced mobility of P (Figure 1) caused by dry seasonal conditions, crops need to be sown as close as possible to the 2015 sowing line. This will allow the best accessibility of the previous fertiliser band.


Monitoring for soil K levels in the Southern Region is challenging due to the lack of data where soil K test critical levels have been calibrated with field trials. Most of the soil K test validation work has been associated with field trials performed in WA. Critical values for Colwell K are generally below 50 mg/kg but this dataset is dominated by lighter texture soils (Figure 4). Recent glasshouse work (UA00140) utilising soils from western, southern and northern growing regions established critical values of 76mg/kg with a reasonable level of confidence (R2 = 0.70) for wheat grown to GS30 (table 3). Similar critical values were established for exchangeable K (80mg/kg wheat and 79mg/kg canola) with slightly improved accuracy compared to Colwell K, especially for canola. Recent work by NSW DPI (GRDC project: DAN00168) is establishing K responsive trials in the southern region in the hope of refining critical levels for all soil K tests. At present either the Colwell K or exchangeable K test can be used as they extract similar amounts of K on most soil types.

Soil sampling for determination of K availability has routinely used depths of 0-10cm. This is potentially a major source of error due to potential stratification of K throughout the soil profile. Stratification of K within the soil profile is highly dependent on soil type, which controls the mobility of any K applied and location of soil layers naturally rich in K associated with clays. Changes in available K with soil depth are strongly dependent on inherent soil physicochemical properties and soil formation processes as well as management interventions. Significant stratification can occur at the soil surface where fertilizer has been applied but this has also been shown in overseas studies to be amplified by reduced tillage management or by significant depletion of K by crops in subsurface layers. Conversely higher concentrations of K can be found at depth in duplex soils where leaching of K can occur from the sandier top soil only to be captured by the clayey subsoil. Soil pH can also play a role in K leaching with H+ ions being able to displace K+ under acidic conditions where K+ is then subject to movement down soil profiles. In neutral and alkaline soils the displacement of K+ by H+ ions is reduced.

It is therefore recommended that to accurately account for K availability for crop uptake that deeper sampling depths are taken (e.g. 0-10cm, 10-30cm). Research from Western Australia has indicated that in some occurrences non responsive sites on low soil K test surface samples could be explained by significant sources of K in the 10-30cm layer that was freely available to the crop.

If an application of K is found to be needed, the most efficient application method is to band the K (potash) with the seed. This will allow the developing root system to access the source of K but be careful not to over apply as potash (KCl) has a high salt content which inhibit seed germination when placed in close proximity. Broadcasting of K has been shown to be less efficient in moderate to heavier textured soils due to the immobility of K in these soils. An approximation of K mobility can be obtained through a measure of the soils cation exchange capacity (CEC).

Table 2: Defined critical Colwell K values for wheat across a range of soil types. (Source: Brennan et al. 2013).
Soil order Critical Colwell K (mg/kg) Colwell K critical range (mg/kg) Correlation (r)
All (201) 45 41-49 0.43
Chromosols 40 35-45 0.46
Kandosols 49 45-52 0.78
Tenosols 41 32-52 0.27


Monitoring for soil Sulphur (S) is challenging as the plant available form of S, sulphate (SO4-) is highly mobile in the soil (similar to nitrate) where it has the potential to be leached into the subsoil after significant rainfall events. In some soils, clay content increases with depth so sulphur leached from the soil surface could accumulate at depth. Another consideration when performing a soil S test is that significant S mineralisation can occur during the fallow and growing season in association with OM/OC mineralisation. Gypsum layers are common in many of the alkaline soils found in semi-arid areas and consequently soil testing often reveals increasing sulphur concentration down the profile. With this in mind it is recommended that soil samples beyond 10cm are collected in order to fully estimate the S available for crop uptake (e.g. 0-10cm and 10-30cm).

As with K, soil S test calibrations with crop responses to applied S is dominated by Western Australian data which assessed only one soil S test (KCl-40). Critical values were established at 4.5mg/kg for wheat and 7mg/kg for canola (Table 3). This paucity of data from the Southern Region is being addressed by NSW DPI (Conyers et al. GRDC project: DAN00168). To date the team has struggled to identify response of several crops to S applications on some very low S profiles (KCl-40 < 3mg/kg), which suggests current critical values are towards the high end. Project UA00140 also attempted to develop critical values using a combination of soil types from western, southern and northern regions in the glasshouse growing wheat (GS30) and canola (vegetative). The project found that critical values for this particular growth stage were much lower than found in Western Australia which further supports the findings of Conyers et al. Critical values using KCL-40 soil test were 2.6mg/kg and 2mg/kg for wheat and canola, respectively with moderate accuracy (R2 = 0.46 wheat, R2 = 0.57 canola). Similar critical values were obtained for the MCP S soil test method (2.2mg/kg – wheat, 3.6mg/kg – canola) but these critical values were more accurate compared to the KCL-40 method (R2 values of 0.73 and 0.87 for wheat and canola respectively).

In essence S soil levels are easily managed due to the high availability of S products when applied in soil systems as most soil types have a limited ability to form unavailable S complexes. Some management practices are inadvertently supplying significant amounts of S through the application of gypsum to manage sodicity, application of ammonium sulphate as a source of N and even a standard application rate of 100kg/ha of MAP/DAP will supply approximately 1.5kg S/ha.

Table 3: Defined critical levels for KCL-40 soil S test for wheat and canola at various depths of sampling. (Source: Anderson et al. 2013).
Crop (depth) Critical KCl-40 (mg/kg) KCl-40 range (mg/kg) Correlation (r)
Wheat - all (0-10cm) 4.5 3.2-6.4 0.26
Wheat - WA (0-10cm) 4.5 3.5-5.9 0.36
Wheat - WA (0-30cm) 4.6 4.0-5.3 0.63
Canola - WA (0-10cm) 6.8 6.0-7.7 0.32
Canola - WA (0-30cm) 7.1 6.7-7.5 0.63

Why not validate soil test values with paddock trials?

As demonstrated above, the value of applying P fertiliser sources is heavily reliant on soil types and climate. In most cases current P management strategies should be sufficient in maintaining sufficient P levels on soils types with PBI values <80. Soil types with greater ability to remove P from the available pool have a tendency to require greater P amounts then standard replacement rates to achieve maximum yield. In some cases higher rates may not be economical and therefore we recommend establishing on-farm assessments to obtain a greater understanding of economical P rates for these soils. Ideally trials should contain a control (0P), standard rate (replacement) and a P rich strip (2-3 x replacement). Make sure the extra N applied with P sources in balanced in the lower P treatments and continually monitor with soil testing.

The limited data from the Southern Region for assessing soil tests values for K and S with crop responses to applied K and S requires greater reliance on the need to validate your soil test results with on-farm demonstration trials or paddock scale trials. It is recommended that trials are performed that contain a strip of applied K or S across the paddock before including K and S fertiliser as part of a farm nutrient management practice.


Soil monitoring is the starting point in order to diagnose potential nutrient deficiencies of P, K and S. However knowledge around the behaviour of such nutrients is the key to obtaining a representative soil sample for crop nutrient uptake.


Funding for this work was provided through the GRDC Project UA00140 and SAGIT projects UA1115 and UA2101 and their support is gratefully acknowledged.

Useful resources

Soil quality DGT factsheet

Phosphorus after the dry

Fertilizer P management after 2006 drought

Anderson GC, Peverill KI, Brennan RF (2013) Soil sulfur—crop response calibration relationships and criteria for field crops grown in Australia. Crop and Pasture Science 64: 523-530. doi: Article.

Brennan RF, Bell MJ (2013) Soil potassium—crop response calibration relationships and criteria for field crops grown in Australia. Crop and Pasture Science 64: 514-522. doi: Article.

Contact details

Sean Mason
School of Agriculture, Food and Wine
The University of Adelaide, Australia 5005
08 8313 8107
0422 066 635