Take home message

• Phosphorus removal in wheat production systems is approximately 10 and 8 kg P/ha/year in the Goondiwindi and Nyngan districts, respectively, based on estimates of long-term average yields and grain phosphorus concentrations.
• Different soils have different phosphorus buffering indices (PBIs) that need to be considered in phosphorus budgeting. Consequently, soils with strong phosphorus sorbing capacity often require increased phosphorus input due to a proportion of applied phosphorus that will be strongly sorbed in the soils and rendered part of the long-term phosphorus pool that is not readily plant available.
• Recent studies show manure and mineral fertilisers (e.g., MAP) are comparable phosphorus sources for crop production despite manure’s having a component of phosphorus in the organic form.
• In soils with a high PBI, there may be an advantage in using manure as a phosphorus source over MAP.
• In low PBI soils, mixing P (i.e., incorporating into the soil) provides yield and P uptake advantages over banding and vice versa in high PBI soils.

Phosphorus budgeting

Phosphorus (P) budgeting aims to maintain soil Colwell P values (0–10 cm) in a range that ensures crop yield is not limited. If the P budgeting process is inadequate to maintain a satisfactory soil Colwell P, then future grain yields will be limited by P supply. A significant proportion of northern and western grain production systems of NSW and QLD have marginal or inadequate Colwell P values (Bell et al., 2012; Dalal, 1997; Hackney et al., 2021; Wang et al., 2007). Where this occurs, P exports over decades have, on average, been greater than P fertiliser inputs. Northern grain production systems research has also identified that P budgeting needs to consider the Colwell P in both the 0–10 cm and 10–20 cm layers (Bell et al., 2012; Singh et al., 2005). In southern NSW, the probability of deep (10–20 cm) P response is very low, where the Colwell P is >38 (mg P/kg soil) in the 0–10cm layer (Uddin et al., 2025). Deep P in northern NSW and QLD is thought to be important because of depleted sub-soil P reserves and surface soil drying in northern production systems that limit crop root access to surface (0–10 cm) applied P (Verburg et al., 2021).

The amount of P removed from the paddock in grain can be estimated by considering the grain P concentration and the grain production per hectare. Phosphorus concentration of wheat grain varies between 2.6 and 3.6 kg P/t (Barati 2024; Norton 2012, 2014), while estimated wheat yields using French and Scholtz (1984) for the Goondiwindi and Nyngan districts over the last 25 years averaged 3.7 and 2.7 t/ha respectively (Figure 1a and 1b). This provides an estimated P removal of 9.6 to 13.3 and 7.0 to 9.7 kg P/ha/yrfor Goondiwindi and Nyngan, respectively. To replace P removal, MAP rates would need to be in the range of 44 – 60 kg MAP/ha/yr for Goondiwindi and 32 – 44 kg MAP/ha/yr for Nyngan, assuming 22% P in MAP.

Figure 1 (a). Phosphorus (kg) exported in grain for locations in NSW, South Australia and Victoria (Barati 2024; Norton 2012; 2014) and (b). estimated P removal using grain yield estimated by French and Shultz (1984) and assuming P removal is 3 kg P/t of grain production. Yield calculations assume no crop was sown if <120 mm of stored soil water for Goondiwindi and <50 mm of stored soil water for Nyngan.

Soil phosphorus sorption

Phosphorus sorption refers to the process by which P in soil solution is adsorbed onto soil particles or chemically bound to soil minerals. This interaction determines the availability of P for plant uptake and affects how P behaves in the soil. Soils sorb P into short-, medium- and longer-term soil P pools. The proportion of P sorbed into the longer-term soil P pool (e.g., aged and crystallised Fe, Al, Mn, Ca and Mg phosphates) should be accounted for in P budgeting to ensure soil P values are maintained at critical Colwell P levels through time. This is sometimes termed the ‘soil P factor’. The amount of P sorbed into the longer-term soil pool increases with increasing P buffering index (PBI) and, therefore, can be considered soil type or PBI dependent. Long-term P sorption can be estimated by monitoring paddocks using known P inputs (e.g., MAP at sowing), estimated P export in grain (grain P concentration [kg/t] x grain yield [t/ha]) and by tracking soil Colwell P (mg P/kg soil) over time. This approach provides a useful estimate of P additions required above P balance (P balance = P inputs – P exports) to maintain a consistent soil Colwell P value over time (e.g., constant Colwell P = P removed in grain plus allowance for soil P factor). A long-term trial at Glenelg in S-NSW found the allowance for long-term P sorption (soil P factor) was in the order of 0.5 – 1.5 kg P/ha/yr (Laycock et al., 2022).

Provided the paddock Colwell P is in a safe range (e.g., 0-10cm is 25 to 33 mg P/kg soil in many Vertosols and 30 to 38 mg P/kg soil in many Chromosols; Sandral et al., 2019; 2020), then the P export in grain and soil P factor (long-term P sorption) can be added together to provide a P budget. Goondiwindi and Nyngan have respective P export estimates of ~10 and ~8 kg P/ha/yr on average, and an estimated allowance of 0.5 to 1.5 kg P/ha/yr is added as the ‘soil P factor’ to provide an estimated P budget of 10.5 – 11.5 and 8.5 – 9.5 kg P/ha/yr (48 – 52 and 39 – 43 kg MAP/ha/yr respectively). This is, of course, a starting point, and the budget is then altered based on the monitoring over time of soil Colwell P values. If Colwell P values are consistently increasing, then the annual P input can be reduced. The best approach is to make small changes (0.5 to 1.5 kg P/ha/yr) when reducing P inputs and monitoring Colwell P values over time. It is also best to have three years of Colwell P data before changing P inputs – unless Colwell P values are quickly declining. In this scenario, it is best to respond quickly with increased P inputs so future yield is not lost (see Laycock 2022).

Table 1. Estimated P inputs required (total P, total MAP) for different yield and grain P concentration scenarios in wheat.

Tracking Colwell phosphorus

A long-term P experiment established and managed by Incitec Pivot Fertilisers at Glenelg in southern NSW showed ~2.5 to 3.0 kg P/ha was required to raise the soil Cowell P value 1 unit. The site starting Colwell P in 2007 was 26 mg P/kg soil and the target Colwell P for 95% of maximum yield in this soil is 34 to 42 mg P/kg soil. The P rate of 30 kg P/ha (Figure 2) rapidly increased the Colwell P to within the preferred Colwell P target range. After one or two years at this high rate (30 kg P/ha), the P rate can be reduced to an estimated rate of between 11 to 14 kg P/ha/yr to achieve the desired Colwell P range. Note that 10 kg P/ha did maintain the soil Colwell P at a constant level over the experimental period, although this Colwell P (open circle, Figure 2) is below the critical P range and produced lower grain yields than treatments in the critical range (horizontal lines, Figure 2) in a number of years. The initial large P input (30 kg P/ha/yr) for two years in this example is known as the maintenance plus capital P phase, and once the adequate P range is obtained, the P maintain phase should be applied (e.g., 11 to 14 kg P/ha/yr).

Figure 2. Annual Colwell P changes for five P rates (0, 10, 20, 30, 40 kg P/ha/yr) applied at the Glenelg experimental site for the unlimited N treatment (120 kg N/ha). The light banded area (horizontal lines) represents the preferred Colwell P range for the Glenelg site. The crop years were wheat (2007 – failed crop), wheat, albus lupins, wheat, wheat, wheat, resown canola, wheat, wheat, canola, wheat, wheat, canola, wheat (2021). The starting Colwell P was 26 mg P/kg in 2007, and the finishing Colwell P for the 10 kg P/ha treatment was also 26 mg P/kg soil in 2021 (Laycock et al., 2022). This treatment (10 kg P/ha/yr) was below the preferred critical soil P range and produced lower yields in a number of years when compared to treatments that had higher P inputs.

Phosphorus buffering index (PBI)

The PBI test measures the soil’s P sorption capacity, which reflects how soluble P is adsorbed onto clay minerals or precipitated within the soil. This process determines the distribution of P between the solid and solution phases. Soils with a high PBI have a greater tendency to sorb P compared to those with a low PBI. As a result, the soil’s P sorption capacity affects the availability of P to plants and is useful in guiding critical Colwell P values (Moody 2007).

Critical Colwell phosphorus range

The critical Colwell P range varies for different soil types/PBI values and crop types. Most farming systems simplify this by targeting the Colwell P range for wheat and making necessary changes for soil type/PBI. Because soil Colwell P values vary from year to year (see Figure 2), it is best to target the 90% critical Colwell P as the minimum base above which you want to maintain Colwell P levels. For instance, Sandral et al. (2020) reanalysed experiments in the BFDC database and identified the critical 90 and 95% range for 44 experiments on Vertosols in NSW to be 24 to 35 mg P/kg soil and Chromosols using 48 experiments to be 34 to 42 mg P/kg soil. Similarly, Bell et al. (2013) found 90% of critical Colwell P in wheat for northern NSW soils to be 24 mg P/kg soil. Because of year-to-year variation in Colwell P (often 6 to 9 Colwell P - see Figure 2), the best practical way of managing Colwell P levels is to maintain them above 90% of critical, which is above 24 (mg P/kg soil) for Vertosols and above 33 (mg P/kg soil) for Chromosols (Figure 3).

Figure 3. Wheat grain yield response to added phosphorus on (a) Vertosols of NSW and (b) Chromosols of NSW using the Mitscherlich equation to estimate 90 and 95% of critical Colwell P. Raw data is taken from the BFDC database. Each dot represents an experiment with 44 and 48 experiments used for the Vertosol and Chromosol, respectively (Sandral et al., 2020).

Phosphorus source

The P source often used in grain production is MAP (21.9% P) or DAP (20% P), while manures are only occasionally used (1.5 to 3.5% P), (Barati M 2024). The main manures used in grain production include chicken and cattle manures, used in cropping rotations, generally at frequency’s ranging from 1 in 4 to 1 in 6 years. In these examples, the question often arises about the value of manures relative to other mineral P sources (e.g., MAP or DAP).

The wheat grain yield response to two P sources (MAP and chicken manure) in two contrasting soils (Kandosol and Vertosol) was tested to determine the effectiveness of MAP and chicken manure (Figure 4). In these controlled environment conditions, MAP and manure resulted in similar yields across soil types at low P application rates; however, at higher P rates, closer to maximum yield, wheat yield response to MAP was highest in the Kandosol, and chicken manure was highest in the Vertosol. MAP performed best on the Kandosol, which had a low PBI (56) and low available aluminium, while chicken manure performed best on the Vertosol, which had a higher PBI (105) and higher levels of soil calcium (Ca). It is possible the MAP was less available in the Vertosol due to greater P sorption, while the organic matter in manure reduced the P sorption. In the higher P-sorbing soils (Vertosol), organic compounds from manure compete with P for soil binding sites and often reduce P-sorption. In addition, the organic acids released from organic matter breakdown tie up Ca, stopping it from forming stable Ca-phosphate compounds and keeping more P available in the soil (Wang and Kuzyakov, 2024). These organic compounds also interfere with the formation of hard-to-dissolve Ca-phosphate crystals, instead encouraging the creation of softer, more easily dissolved forms. Additionally, manure boosts soil biology, increasing the activity of microbes and fungi, which produce enzymes that may help break down organic P into plant available forms (Grossl and Inskeep, 1991), adding another way for plants to access P in manures.

Figure 4. Percent of the maximum grain yield response of wheat (y axis) to chicken manure (solid line) and MAP (dashed line) applied at 10 P rates (x axis) in (a) a Kandosol from southern NSW and(b) a Vertosol from northern NSW (Barati 2024).

Phosphorus placement

A P placement experiment was carried out in a controlled-environment study to determine if subsurface P banding (7cm) compared to incorporation (mixing P in the 0–10cm layer) impacts plant response and P uptake in a Ferrosol and Tenosol with high and low PBI values, respectively (Figure 5). Results showed that both MAP and manure provided an improved grain yield and P uptake when banded in the high PBI Ferrosol while mixing was the better strategy in the low PBI Tenosol.

Figure 5. Percent of the maximum grain yield (a) and P uptake (b) of wheat for banded and mixed P applications in the top 10 cm in a high PBI Ferrosol and low PBI Tenosol. Different capitalised letters within soil groupings indicate significant differences (P=0.05).

Economics of phosphorus sources

The economics of using manures as a P source require a comparative analysis with MAP that considers (i) manure cost per tonne, (ii) per cent P in the manure, (iii) an allowance for water (e.g., 30% water by weight) and (iv) additional transportation and spreading costs associated with the additional volume needed for manures. In Table 2 below, these points are considered with the exception of additional spreading costs. Results indicate that manure is an economically competitive source of P for short haul (approximately 100 km transportation) scenarios.

In the example below, manure is shown to be more sensitive to haulage cost variations than MAP. Consequently, the haulage distance, number of hauls required and percent P in the manure are the key influencing factors in determining the economic value of manure as a P source. Because of the higher concentration of P in MAP and lower volumes of MAP required on-farm relative to manure, the economic value of MAP is primarily driven by product price variation (Table 2).

Table 2. Matrix illustrating the estimated cost of short haul delivery of 4 tonnes of phosphorus as either manure or MAP. Changed variables for manure include the cost of manure ($40, $45 and $50/t of product) and haulage for manure (18, 26 and $32/t). Changed variables for MAP include the cost of MAP ($700/t up to $1500/t). Assumptions include manure provided at 35% moisture and MAP at 10% moisture, and manure P content is 2%. Spreading costs are not included (e.g., $8 to 12/ha for MAP and Manure $18 to $26/ha).

Conclusion

The most important consideration for grain production is ensuring soils are within an acceptable Colwell P range so as not to limit grain yield. Despite the importance of this, many paddocks in western regions of NSW and QLD are below levels required to achieve 90 to 95% of maximum yield (Bell et al., 2012; Dalal, 1997; Hackney et al., 2022; Wang et al., 2007). Phosphorus budgeting for capital plus maintenance levels of P is important where P levels are below the critical range. In paddocks where Colwell P levels are within the critical range, P maintenance levels are important to sustain adequate Colwell P levels.

Phosphorus sources are an important consideration and manure performance is comparable to MAP on a total P basis. In addition, manures are likely to provide additional value in soils with moderate to high PBI values, while MAP is likely the best performer where PBI values are in the low range (PBI <100). Additional P efficiencies are likely in moderate to high PBI soils where P banding is used, while in low PBI soils, mixing (0–10cm) provides the best results.

References

Barati M (2024) Assessing the agronomic phosphorus fertiliser value of animal manures. Submitted doctoral dissertation. Southern Cross University.

Bell M, Lester D, Smith L, Want P (2012) Increasing complexity in nutrient management on clay soils in the northern grain belt–nutrient stratification and multiple nutrient limitations. Proceedings of 16th Australian Agronomy Conference. 14–18.

Bell MJ, Moody PW, Anderson GC, Strong W (2013) Soil phosphorus—crop response calibration relationships and criteria for oilseeds, grain legumes and summer cereal crops grown in Australia. Crop and Pasture Science 64, 499–513. (https://doi.org/10.1071/CP12428)

Dalal RC (1997) Long-term phosphorus trends in Vertisols under continuous cereal cropping. Australian Journal of Soil Research 35(2): 327–340.

French RJ, Schultz JE (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.

Grossl PR, Inskeep WP (1991) Precipitation of dicalcium phosphate dihydrate in the presence of organic acids. Soil Science Society of America Journal 55(3), 670–675.

Hackney B, Wicks T, Hart J, Piltz J, (2022) Cranking up crop yields and livestock production using summer sown pasture legumes – revisiting fundamental soil and agronomy with new technologies to increase production and rotation flexibility.

Laycock J, Mason S, Norton R, and Sandral G (2022) Balancing risk and reward with high phosphorus and nitrogen input costs. Grain Research Update, Wagga Wagga February 2022.

Moody PW (2007) Interpretation of a single-point P buffering index for adjusting critical levels of the Colwell soil P test. Australian Journal of Soil Research. 45(1): 55-62.

Norton RM (2012) Wheat grain nutrient concentrations for south-eastern Australia. Capturing Opportunities and Overcoming Obstacles in Australian Agronomy edited by I. Yunusa. Proc. 16th Australian Agronomy Conference 2012, 14-18 October 2012, Armidale, NSW.

Norton RM (2014) Canola seed nutrient concentration for southern Australia. Ware AH and Potter TD (eds) 18th Australian Research Assembly on Brassicas (ARAB 18). Tanunda, 2014. Proceedings. Australian Oilseed Federation, p 1-6.

Sandral G, Tavakkoli E, Barati B, Pumpa R, Armstrong R, Lester D, Mason S, Norton R, Bell M (2019) Phosphorus and phosphorus stratification. Grain Research Update, Wagga Wagga February 2019.

Sandral G, Tavakkoli E, McBeath T, Armstrong R, Mason S, Wilhelm N, Hailing R, Pumpa R, Fiske K, Barati M and Armstong D (2020) Budgeting phosphorus in medium and lower rainfall zones of southern NSW. Grain Research Update, Wagga Wagga February 2020.

Singh D K, Sale P W G and Routley R R (2005) Increasing phosphorus supply in subsurface soil in northern Australia: rationale for deep placement and the effects with various crops. Plant and Soil 269, 35–44.

Uddin S, Armstrong R, Dunsford K, Wilhelm N, Verburg K, McBeth T and Sandral G (2025) Crop responses to P placement in southern NSW: Residual P effects over five seasons. GRDC Wagga Wagga Updates February 2025.

Wang C, Kuzyakov Y (2024) Soil organic matter priming: The pH effects. Global Change Biology, 30(6), e17349.

Wang X, Lester DW, Guppy CN, Lockwood PV, Tang C (2007) Changes in phosphorus fractions at various soil depths following long-term P fertiliser application on a black Vertosol from south-eastern Queensland. Australian Journal of Soil Research 45(7): 524–532.

Verburg K, McBeath T, Armstrong R, Tavakkoli E, Wilhelm N, McLaughlin M, Haling R, Richardson A, Mason S, Kirkegaard J, Sandral G (2021) Soil water dynamics as a function of soil and climate to inform phosphorus placement strategies. National Soils Conference 2021; Cairns. Soil Science Australia.

Contact details

Maryam Barati
University of Adelaide, School of Agriculture, Food and Wine
Email: maryam.barati@adelaide.edu.au
Ph: 0448 446 224

Date published
February 2025