Determining production gains from N,P,K inputs on Australian dairy farms

Year of study: 2012

Project owner: Future Farming Systems Research Division, Department of Primary Industries Victoria

Contacts: Cameron J. P. Gourley, Sharon R. Aarons

Project details


The cost of manufacturing fertilisers, principally associated with the cost of fossil fuels, is expected to rise substantially in real terms in coming decades and consequently fertiliser use will become a larger part of dairy farm operating costs. Many farms have already high soil fertility levels with limited production gains from further fertiliser expenditure. In this study we investigated nitrogen (N), phosphorus (P) and potassium (K) whole-farm balances and existing soil nutrient levels on 41 contrasting dairy operations across Australia.

Hypotheses/conceptual models

There are opportunities on many dairy farms to reduce or exclude fertiliser inputs.

Basis of Study

Total N, P and K inputs onto dairy farms, mainly in the form of feed and fertiliser, are usually much greater than the outputs of P, K and Nin milk, animals, and crops(VandeHaar and St-Pierre 2006). These surpluses tend to increase as farms intensify and stocking rates increase. Excess P and K on dairy farms can result in increasing soil P and K levels beyond agronomic requirements (Weaver and Reed, 1998). Unlike P and K, N is not significantly buffered by soils, and where N is applied in high concentrations such as indung, urine or fertiliser, losses through volatilization and leaching can be high (Rotz et al.,2005). The challenge of optimising the production potential and profitability of nutrient inputs in animal agriculture while reducing negative environmental effects is faced by most industrialized countries (Steinfeld et al. 2006) including Australia. Animal agriculture is now commonly recognised as the dominant contributor of nutrient inputs to water bodies because of its extensiveness (Department of Water 2010).


Summary of key messages/findings

The results suggest that there are opportunities on many dairy farms to reduce or exclude fertiliser inputs.

We found that there are large excesses of N, P and K on a broad range of dairy operations across Australia and that nutrient surpluses and wi thin farm nutrient distribution was related to stocking rate and milk production per ha . While there was a positive relationship (P<0.001) between milk production and N fertiliser inputs, there was no overall relationship between P and K fertiliser inputs and milk producti on.

The quantity of nutrients imported onto these diver se dairy operations varied markedly, with feed and fertiliser generally the most substan tial imports. Elevated N, P and K surpluses on dairy farms were related to milk produ ction and stocking rate per ha. While N fertiliser input was a driver of milk production fr om home-grown pasture and crops, there was a high degree of uncertainty around production gains. In contrast, the lack of a relationship between P and K fertiliser inputs and milk production from home-grown pasture and crops reflected the high soil P and K l evels measured on these farms. Within farm soil nutrient heterogeneity is substantial, ir respective of the intensity of the dairy operation.

Higher soil nutrient levels of P, K (and N) are dri ven by paddock stocking density, proximity to the dairy, frequency of effluent applications, a nd feeding strategies. These results support recent assessments in Australia that sugges t that excess nutrients may be cycling in some agricultural systems and placing considerab le pressure on the environment (OECD 2008). Soil testing to determine available soil P and K is therefore an important tool to manage P and K build-up and maintenance.

What recommendations were made about soil/land management and soil health?

We recommend comprehensive soil testing to guide fe rtiliser decisions on dairy farms and more sophisticated quantification of nutrient flows through feed and manure deposition to determine manure nutrient loading rates across the farm for highly stocked dairy operations. The relatively small costs associated with a strate gic and on-going soil sampling program are likely to be returned many times over through t he potential savings in unnecessary fertiliser expenditure. Where applications are war ranted, appropriate rates and blends of fertiliser should ensure profitable increases in pa sture and crop productivity.

What challenges or opportunities exist to apply findings?

In the future, more sophisticated approaches will likely be required which quantify nutrient flows through the continuum of feed, milk productio n and manure; increases the capture and storage of excreted manure; determines nutrient loading rates in areas across the farm; and ensures that fertiliser and manure are ap plied under favourable soil and climatic conditions for optimum plant uptake.

Location and site details

Site details

Forty-one commercial dairy farms representing a bro ad range of geographic locations, soil types, milk production, herd and farm size, and rel iance on irrigation, were selected for this study (Gourley et al. 2012). Farm area ranged from 47 to 496 ha, cow numbers per farm ranged from 59 to 1930 cows, and number of paddocks per farm ranged from 14 to 141.

Management history

Pre-trial management: NA

Trial management: NA

Experimental design

Treatments: NA

Trial design/layout: NA

Soil sampling method

Soil sampling Soil nutrient levels were determined by sampling al l paddocks used for pasture and crop production on each farm, as well as areas with high animal densities (holding areas, feeding areas, sick paddocks, bull paddocks). Thirt y 0-10 cm soil cores were collected along a diagonal transect across each area, bulked, dried (40&ordm;C) and ground (2 mm sieve) and analysed for pH (0.01M CaCl2), Olsen extractabl e P, Colwell P and K, and P buffering index (PBI). More than 1860 paddock samples were in cluded in the analysis.

Production measurements

Plant and/or animal production measurements

Whole-farm nutrient balances were determined for N, P and K as described by Gourley et al. (2012). Milk production from home-grown feed wa s determined for each farm as the difference between total annual milk production and that produced from imported feed (Heard et al. 2011).

Cost and value of production: NA


Farm-scale nutrient balances

Whole-farm N surplus (the difference between total nutrient imports and total nutrient exports) ranged from 47 to 601 kg N ha-1 yr-1 with a median of 226 kg N ha-1, and N use efficiency (the ratio of total 466 nutrient exported in product divided by total nutrient imported at the farm scale) ranged from 14 to 50%, with a median of 26%. We also found a strong correlation between total N imported and milk production per ha (Fig. 1a) while N surplus was also strongly related to milk production (Fig. 1b) with the slope of this linear relationship (0.0121; SE = 0.0015) providing an estimate of the productivity N surplus, equivalent to 12.1 g N litre-1 milk produced. Whole-farm P surplus ranged from -7 to + 133 kg P ha-1 yr-1, with a median of 28 kg ha-1. Phosphorus use efficiencies ranged from 6 to 158%, with a median of 35%. Potassium balances ranged from 13 to 452 kg K ha-1, with a median value of 74 kg K ha-1; K use efficiency ranged from 9 to 48%, with a median value of 20%.

Figure 1: Relationships between milk production and (a) whole-farm nitrogen inputs and (b) nitrogen surplus for 41 contrasting dairy farms across Australia. Unshaded symbols represent organic dairy farms.

Soil P and K levels

There was a large range in soil P and K levels from grazed pasture paddocks (Figure 2). Olsen P levels ranged between 3 and 189 mg kg-1 and the Colwell K levels ranged fro m 14 to 3400 mg kg-1. Only 20% of the paddocks sampled had soil P or K values below the recommended agronomic optimum (Olsen P of 20 mg kg-1 and Colwell K of 180 mg kg- 1), while 50% of the paddocks sampled were at least 1.5 times the recommended agronomic optimum. At the high fertility end, 20% of paddocks sampled had Olsen P or Colwell K levels at least 3 times the agronomic requirements. Areas with high animal densities, such as calving paddocks, feed pa ds, holding areas and ‘hospital’ paddocks, had subs tantially elevated soil nutrient levels when compared to the overall p asture paddocks. In contrast, low intensity areas such as ‘other animal’ and areas with trees had much lower fertility level s. Production gains from fertiliser inputs Milk production from home-grown feed increased with increasing N fertiliser input, although there was a high variation and uncertainty around milk production gains (Fig. 3a), suggesting substantial improvements in the utilization of N by grown pasture could be achieved on many farms. The relati onship improved slightly when additional N from the fixation of atmospheric nitrogen by pasture legumes was include d (Fig. 3b).

Figure 2: Box and whisker plots of soil P and K levels for individual paddocks for dairy farms grouped into annual milk production classes (litres ha-1). Mean soil test levels for each production class are below the plots.

In contrast there was no relationship between P and K fertiliser applications and milk production attributed to home-grown feed (Fig. 3c and d). This lack of relationship is supported by the generally high levels of soil P an d K measured. Under these conditions, additional pasture and crop production from the application of P and K fertiliser would not be expected and therefore neither would an associated increase in milk production from home grown feed. Moreover, the milk production from farms with low or no P or K fertiliser inputs but with adequate levels of soil P or K, suggest that these soil reserves can be utilized without a resulting decline in milk production.

Figure 4: Relationships between milk production from home-grown feed and (a) N fertiliser input, (b) N fertiliser plus N fixation, (c) P fertiliser input and (d) K fertiliser input.

How results have been reported (e.g. leaflet, technical report, journal article)

CSIRO Animal Production Science
Internationally Peer Reviewed Scientific Journal PaperDairy Australia Final Report

Soil Science Australia
2012 Joint Australia and New Zealand Soils Conference p.465

How can a copy of any relevant reports be obtained?

Future Farming Systems Research Division,
Department of Primary Industries Victoria,
Ellinbank Centre, 1301 Hazeldean Rd, Ellinbank, Victoria 3821, Australia


Intellectual property ownership

DPI Victoria

Approval to publish project summary in trial directory



Gourley CJP, Dougherty W, Weaver DM, Aarons S, Awt y IM, Gibson DM, Hannah MC, Smith AP and Peverill KI (2012) Farm-scale nitrogen, phosphorus, potassium and sulphur balances and use efficiencies on Australian dairy farms.

Department of Water (2010) Vasse Wonnerup Wetland and Geographe Bay Water Quality Improvement Plan.

Heard JW, Doyle PT, Francis SA, Staines MVH, Wales WJ, 2011. Calculating dry matter consumption of dairy herds in Australia: the need to fully account for energy requirements and issues with estimating energy supply. Animal Production Science 51, 605-614

OECD (2008) Environmental performance of agriculture in OECD countries since 1990, Paris, France, (verified 25th November 2011).

Rotz CA, Taube F, Russelle MP, Oenema J, Sanderson MA, Wachendorf M (2005) Whole-farm perspectives of nutrient flows in grassland agriculture. Crop Science 23, 2139-2159.

Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosa les M & de Haan C (2006) Livestock’s long shadow: environmental issues and options. Food and Agriculture Organisation, Rome.

VandeHaar MJ, St-Pierre N (2006) Major advances in nutrition: relevance to the sustainability of the dairy industry. Journal of Dairy Science 89, 1280-1291.

Weaver DM, Reed AEG (1998) Patterns of nutrient status and fertiliser practice on soils of the south coast of Western Australia. Agriculture, Ecosystems and Environment 67, 37-53