Agri-environmental indicator - risk of pollution by phosphorus

Data from May 2017. Most recent data: Further Eurostat information, Main tables and Database. Planned update: November 2018.

This article provides a fact sheet of the European Union (EU) agri-environmental indicator risk of pollution by phosphorus. It consists of an overview of recent data, complemented by all information on definitions, measurement methods and context needed to interpret them correctly. The risk of pollution by phosphorus article is part of a set of similar fact sheets providing a complete picture of the state of the agri-environmental indicators in the EU.

Table 1: Gross phosphorus balance, 1990, 1995, 2000, 2005 and 2010-2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Figure 1: Gross phosphorus balance, averages 2000–2004 and 2010–2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Table 2: Phosphorus inputs, 1990, 1995, 2000, 2005 and 2010-2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Table 3: Share of mineral fertilisers input in total phosphorus inputs, 1990, 1995, 2000, 2005 and 2010-14 (%)
Source: Eurostat (aei_pr_gnb)
Table 4: Share of gross manure input in total phosphorus inputs, 1990, 1995, 2000, 2005 and 2010-2014 (%)
Source: Eurostat (aei_pr_gnb)
Figure 2: Nutrient input per ha of utilised agricultural area, average 2010-2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Figure 3: Share of the different phosphorus inputs in total phosphorus input, average 2010–2014 (%)
Source: Eurostat (aei_pr_gnb)
Figure 4: Share of different livestock in manure phosphorus production, average 2010–2014 (%)
Source: Eurostat (aei_pr_gnb)
Table 5: Phosphorus outputs, 1990, 1995, 2000, 2005 and 2010-2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Figure 5: Nutrient output per ha of utilised agricultural area, average 2010-2014 (kg P per ha of utilised agricultural area per year)
Source: Eurostat (aei_pr_gnb)
Figure 6: Share of the different phosphorus outputs in total phosphorus outputs, average 2010–2014 (%)
Source: Eurostat (aei_pr_gnb)
Map 1: P retention classes in EU-25 (including influence of erosion and artificial drainage) (Bomans et al, 2005)

The indicator provides an indication of the potential surplus of phosphorus (P) on agricultural land (kg P per ha per year) and is measured by the following indicators:

Main indicator:

  • Potential surplus of phosphorus on agricultural land (kg P per ha per year).


  • Vulnerability to phosphorus leaching/run-off (to be developed).

Main statistical findings

Key messages

  • In all countries, with the exception of Latvia, the average gross phosphorus balance (GPB) per ha of utilised agriculture area (UAA) between 2010 and 2014 was lower than between 2000 and 2004.
  • The average GPB 2010-2014 was slightly higher in EU-15 (2.3 kg P per ha per year) than in EU-28 (2.0 kg P per ha per year).
  • Due to methodological issues or missing data, balances have been estimated by Eurostat for Belgium, Bulgaria, Denmark, Greece, Spain, Croatia, Italy, Cyprus, Latvia, Lithuania, Luxembourg, Malta, Austria, Romania and Slovakia.
  • The quality and accuracy of the estimated gross phosphorus surplus per ha depends on the quality and accuracy of underlying data and coefficients used. As methodologies (especially with regards to the coefficients) and data sources used in countries vary, the balances are only consistent within a country across time. The gross phosphorus balances are not consistent across countries, which means that data cannot be compared between countries. An update of the handbook on gross nutrient balances is being discussed with experts in the Member States (MS), to improve the coherence and transparency of data and methodologies used across countries. The next data collection will take place in 2017. 
  • The gross phosphorus balance can only indicate the potential risk to the environment while the actual risk depends on many factors including climate conditions, soil type and soil characteristics, management practices such as drainage, tillage, irrigation etc. The risk of phosphorus pollution is only partially determined by the phosphorus balance of a particular year. It is often more strongly determined by the cumulative P balance of the past. Therefore it would be more accurate if results of the main indicator could be interpreted in relation to the subindicator "vulnerability to phosphorus leaching/run-off". However, this subindicator is still not developed.  


Analysis at EU level

The GPB for the EU-28 decreased from an estimated average of 5.2 kg P per ha per year in the period 2000-2004 to 1.9 kg P per ha per year in the period 2010-2014. For the EU-15 the GPB dropped from on average 6.4 kg P N per ha per year in 2000-2004 to an average of 2.3 kg P per ha per year in 2010-2014. The average GPB per ha in 2010-2014 was highest in the Mediterranean islands Cyprus and Malta, above the EU average in Norway, Denmark, Croatia, Belgium, the Netherlands, Portugal, Lithuania, Luxembourg, Finland, Spain, Poland, the United Kingdom, Slovenia, Switzerland, Ireland and Latvia while the balance was negative for Sweden, Slovakia, the Czech Republic, Romania, Hungary, Italy, Bulgaria and Estonia (Figure 1).

Figure 2 shows that many of the countries with a high P surplus also have a relative high P input per ha per year (for instance Malta, the Netherlands and Cyprus). A high input does not necessarily lead to a high P balance, as the balance is also determined by the output. Belgium and Denmark for instance had a higher P input per ha per year than Norway, but due to a much higher P output per ha per year, the estimated balance per ha of Belgium and Denmark was lower than that of Norway.

The inputs of the gross phosphorus balance consist of mineral fertilisers, organic fertilisers (except manure), manure input, and other inputs like atmospheric deposition and seeds and planting material. Mineral fertilisers and manure accounted for more than 93 % of the P input in EU-28 between 2010 and 2014. Other organic fertilisers such as compost, sewage sludge, industrial waste accounted for little more than 5 % of total P inputs (Figure 3). Data on other organic fertilisers are however lacking in many countries, therefore the significance of these fertilisers in these countries and at EU level could be underestimated. Other inputs (atmospheric deposition, seeds and planting material) represented only 1 % of total phosphorus input.

Figure 3 also shows large differences between countries in the use of inputs. The choice of fertiliser type (manure, mineral fertilisers, other organic fertilisers) has different impacts on the environment. Phosphorus is not available to the plant in the same rate for different types of manure and fertilisers. Through the use of energy during production, P fertilisers contribute to GHG emissions and fossil fuel depletion. Mineral P fertilisers are produced from natural resources which have a limited availability and which are depleting. Other potential fertilisers like urban wastes often include health hazards (to both plants and humans) and procedures commonly used to reduce these hazards - such as composting - tend to reduce the fertiliser value (leaching/volatilisation).

Manure production is determined by the amount and type of livestock in the country. The agri-environmental indicator livestock patterns shows the livestock density in EU-28. Malta, the Netherlands, Belgium, Denmark, Cyprus and Ireland have the highest livestock densities and also have the highest rates of manure input per ha (over 14 kg P per ha per year). Bulgaria, Estonia, Latvia, Lithuania and Slovakia have the lowest livestock densities and also belong to the countries with the lowest rates of manure input per ha (less than 6 kg P per ha per year). The countries with the highest livestock densities also belong to the countries with a high share (over 60 %) of manure in total nutrient input.

The share of mineral fertilisers in total P inputs was larger than 50 % in Lithuania, Latvia, Poland, Croatia, Bulgaria, Hungary and Slovakia, while the share in Belgium, Denmark, the Netherlands, and Malta was below 15 %. These figures should however be interpreted with caution as the P input of manure depends largely on the excretion coefficients used to convert animal numbers to P manure production. For many countries coefficients were not available and have been estimated by Eurostat.

The removal of P with harvest and grazing of crops and forage per ha varies between crops and countries as can be seen in Figure 5. Per ha, grassland and cereal production had the highest P removal rates, though great variation exists between countries due to differences in yields and P content per tonne of product. The P content per tonne of product depends among others on farmer practices like fertilisation, irrigation, mowing vs. grazing etc. It should however be noted that coefficients may also vary among countries and crops due to differences in methodologies used (see also the section on data and methodology).

The dominant shares of total P output in the EU-28 in the period 2010-2014 were represented by the uptake of P with cereal production (39 %) and permanent grassland (23 %), see Figure 6. P output depends on cropping patterns, yields, farm management practices (tillage, irrigation), climate etc. The agri-environmental indicator cropping patterns shows that permanent and temporary grassland covered 41 % and cereals 32 % of the UAA in EU-28 in 2010. There are however significant differences between countries. In some countries (for example Ireland, Luxembourg, Slovenia, the United Kingdom, Norway and Switzerland) grassland dominates the UAA, whereas for instance in Bulgaria, Hungary and Denmark cereals are the dominating crops. Permanent crops are significant in Mediterranean countries.

Analysis at country level

As explained in the notes on the data and methodology used, the current balances are not comparable between countries due to differences in definitions, methodologies and data sources used.

Since 1990 P inputs per ha have been continuously decreasing for Belgium, Denmark, Germany, Greece, France, the Netherlands, Finland, Sweden, the United Kingdom, Norway and Switzerland. In the Central and East European countries the total P input decreased significantly during the economic transition between 1989 and 1995. In some of the Central and East European countries a trend towards recovery can be noted, for example Poland, while in some of other countries such a trend cannot be noted and P inputs remain well below the level before the transition period, e.g. the Czech Republic, Latvia and Romania (Table 2).

On average, the main P input at EU-28 level 2010-2014 came from manure, see Figure 3. The P input from manure was lower than the P input from inorganic fertilisers in Lithuania, Latvia, Poland, Croatia, Bulgaria, Hungary, Slovakia and Spain. In many of the Central and East European countries manure P input decreased significantly after the significant decreases in livestock populations during and after the transition. Manure input in the gross phosphorous balance is based on manure production, manure withdrawals and manure import where the manure production is calculated from livestock numbers and excretion factors. The average P excretion per animal per year depends on animal characteristics (e.g. production level, race etc.) and farming practices (for instance animal diets). Many countries take the average level of milk production per cow into account in the estimation of excretion coefficients (a higher level of milk production per cow in general means a higher feed intake and therefore a higher P excretion). Only a few countries update their excretion coefficients regularly to take into account changing farming practices. Most countries use fixed coefficients; this means that for these countries changes in farming practices other than reducing animal numbers are not taken into account. The Netherlands is one of the countries which has updated its excretion coefficients regularly. They are estimated annually based on a model which takes into account animal feed requirements, statistical data on livestock numbers, animal products, feed production, grazing systems, animal housing and other data.

The estimated P output depends mainly on the areas and yields of crops and fodder. In most countries the coefficients used to determine the nutrient contents of crops are constant over time. The yields of crops and fodder are influenced by farming practices (like pesticide and fertiliser use, irrigation etc.) and by soil type and weather. Weather conditions vary significantly from year to year and these fluctuations can also be seen in the estimated P outputs and surplus. In 2003 and 2007 large parts of Europe were hit by extreme weather conditions[1] causing significant drops in crop production. In many Member States a peak in the nutrient balance can be noted for these years. Variations in the P surplus between years should therefore be interpreted with care. Climate and weather conditions are beyond the control of the farmer. To dampen the effect of weather conditions on the balance, the results presented with regards to the nutrient balance are mostly presented not referring to a particular year but as an average for a certain period.

Vulnerability to phosphorus leaching/run-off

The subindicator needs to be developed. Currently only limited data are available.  The vulnerability to P leaching, or P-sensitivity, refers to the combined risk of phosphorus loss to the groundwater or to the surface water by combinations of low sorption capacity, high erosion risk and increased risk of drainage. In 2005 the European Commission commissioned a study Addressing phosphorus related problems at farm practice from which Map 1 is derived. It shows the P-retention capacities of soils in Europe; the capacity of the soil to retain phosphorus by sorption and by resistance to erosion. Sorption of P in the soil is here defined as a process whereby readily soluble phosphate is changed to less soluble forms by reacting with inorganic or organic compound of the soil so that P becomes immobilized.

Red colour in Map 1 corresponds to class 1: very weak P retention capacities, dark green corresponds to class 5: very strong P retention capacities. As can be seen from this map, very dense clusters of soils with very low P retention capacities (class 1) can be found in Finland, Scotland, the north of England, the north of Ireland, Northern Ireland, the Netherlands, the north of Belgium and the Baltic states. Scattered clusters of class 1 soils can be found in the north of Germany, Poland and Greece. Also clusters of soils with low P retention capacities (class 2, orange colour) are detected. These clusters are found in the northwest of Spain and in the Alps. Phosphorus can easily be leached from highly organic soils (e.g. peat) and from sandy soils with low retention capacity. Small amounts are lost from most soils, but when the soils become phosphate saturated, leaching will be enhanced. P accumulation in soils might increase concentrations of dissolved and colloidal P in drainage. Calcareous soils on flat land were found to be the least sensitive to P surplus.

Areas at risk of encountering potential phosphorus excess were indicated in the study by estimating phosphorus balances and combining them with the proportion of vulnerable soils. Manure transfers were not included in these balances, and the mineral phosphorus input was assumed linearly proportional to arable land area. The soil sensitivity was determined for the entire NUTS 2/NUTS 3 region or Member State, not taking into account that sensitive soils (i.e. easily erodible or with a low sorption capacity) are often considered marginal to agriculture. Areas (NUTS 2/NUTS 3 regions or Member States) with high phosphorus surpluses (pressure) and at the same time a high proportion of soils prone to erosion and/or low P-sorption capacity are most vulnerable. The Netherlands and Slovenia display the highest rate of vulnerable soils and high balance surplus. In the Netherlands sensitive soils are prone to leaching, in Slovenia, erosion and run-off are the main agents of phosphorus loss.

Data sources and availability

Indicator definition

Potential surplus of phosphorus on agricultural land (kg P per ha per year).


Main indicator:

  • Potential surplus of phosphorus on agricultural land (kg P per ha per year).


  • Vulnerability to phosphorus leaching/run-off (to be developed).

Links with other indicators

Data used and methodology

Main indicator: gross phosphorus balance

The methodology of the phosphorus balances is described in Eurostat/OECD Phosphorus Balance Handbook. The phosphorus balance lists all inputs and outputs into and out of the soil and calculates the gross phosphorus surplus as the difference between total inputs and total outputs. The gross phosphorus surplus per ha is derived by dividing the total gross phosphorus surplus by the reference area. The reference area of the current version of balances uploaded in Eurobase is the UAA.

The inputs of the phosphorus balance are:

  • inorganic fertilisers,
  • organic fertilisers (excluding manure).
  • Gross manure input, which is calculated from:
  • manure production (phosphorus excretion);
  • manure withdrawals (manure export, manure processed as industrial waste, non-agricultural use of manure, other withdrawals);
  • change in manure stocks;
  • manure import.
  • Other phosphorus inputs, which consist of :
  • seeds and planting material;
  • atmospheric deposition.

The outputs of the gross phosphorus balance are:

  • Total removal of P with the harvest of crops (cereals, dried pulses, root crops, industrial crops, vegetables, fruit, ornamental plants, other harvested crops).
  • Total removal of P with the harvest and grazing of fodder (fodder from arable land, permanent and temporary pasture consumption).
  • Crop residuals removed from the field.

The P inputs and P outputs are estimated for each item of the balance from basic data multiplied with coefficients to convert the data in P content. Basic data (fertiliser consumption, crop production, livestock number, agricultural area) are mostly derived from statistics. Coefficients are mainly estimated by research institutes and can be based on models, statistical data, measured data as well as expert judgements.

Due to methodological issues or missing data, balances have been estimated by Eurostat for Belgium, Bulgaria, Denmark, Greece, Spain, Croatia, Italy, Cyprus, Latvia, Lithuania, Luxembourg, Malta, Austria, Romania and Slovakia.

Data on fertiliser consumption is available in many countries from country specific data sources (surveys, trade/production statistics). For countries where such data was not available, data from Fertilizers Europe have been used. Due to different data sources used (farmer surveys vs trade/production statistics) and inherent problems of data sources used (for instance inclusion of non-agricultural use in statistics based on trade and production) the quality of data cannot be sufficiently verified. There is need for a common methodology to estimate fertiliser consumption by agriculture to ensure reliable and consistent estimations and comparability across countries. 

Data on crop production, livestock number, and agricultural area are available in all countries from harmonised and regulated European statistics (FSS, animal production statistics, livestock registers, crop production statistics). Data on livestock numbers, crop production and reference area have been checked with Eurostat statistics. These data can be considered to be of good reliability.

Excretion coefficients delivered by the country have been used together with the livestock numbers reported by the country to estimate manure production. In the case where the country did not deliver excretion coefficients: the P excretion coefficients vary among countries due to differences in farming practices etc. If N coefficients were available for the country the average P:N relation of available countries has been multiplied with the N coefficient to estimate the P coefficient. Though the N:P ratio may vary between countries due to differences in farming practices it is assumed this variation is less than the variation in P excretion coefficients between countries. If N coefficients were not available P coefficients have been based on other country coefficients. Verification and validation of the coefficients requires expert knowledge. A European uniform methodology and validation procedure to estimate excretion coefficients would be necessary to ensure reliable and consistent estimations and comparability across countries. At present the estimation of coefficients varies between countries in methodology (expert judgement vs complex models based on statistical data) and in updating procedures (to take into account the effect of mitigation actions (other than reducing the level of production) in the estimations, coefficients need to be updated regularly to reflect these changes in farmer practices (mitigation actions)). Because manure input is a large part of the balance, caution should be taken when comparing the data (especially the level of surplus) between countries as these are largely depending on coefficients used which are based on varying methodologies.

Data on manure withdrawals, manure stocks and imports were not available in most countries. Available data on manure withdrawals for non-agricultural use show that this input is significant (more than 5 % of total manure input) in some countries (Belgium, the Czech Republic, the Netherlands), while non-significant in other countries (the United Kingdom, Switzerland).

Almost no country had data available on changes in manure stocks. Under normal circumstances it can be assumed that the change in manure stocks is on average zero.

Data on manure imports were only available from Switzerland, Austria and Belgium. Manure imports were insignificant (less than 1% of total manure input) for these countries. Trade of manure is mainly from high surplus countries to neighbouring countries. It is known that there are exports from for instance the Netherlands to Germany. It is likely that countries with a high nutrient surplus like Belgium and Denmark may export manure as well.

Data on crop production is available from reliable statistics for most of the significant crops cultivated except grassland. The harvest and grazing of grassland is estimated in some countries based on feed requirement models (for example in the Netherlands), while in other countries expert judgments are used. As grassland production (from temporary and permanent grassland) is estimated on average at 34 % of total P output in the EU-28 in 2010-2014, the estimation of grassland production determines to a large extent the outcome of the balance. It is necessary to develop a uniform European method to estimate grassland production to ensure reliable and consistent estimations and comparability across countries. For permanent crops a part of the P uptake is stored in the plant (which is not harvested) and is therefore not at risk to run-off, leaching or volatilisation. This is not taken into account in the balance estimations. Some countries do not have coefficients (used to convert the production in nutrient content) available and reliability varies due to the method used (scientific research vs expert judgments). Verification and validation of these coefficients requires expert knowledge. A sufficient European structure to verify and validate these coefficients is missing. Improvements of the methodologies to calculate more reliable coefficients used in countries are therefore of great importance.

Data on seeds and planting material were not available in all countries. The quality of available data depends on the data sources and assumptions made, as well as the quality of coefficients. Available data show this input is less than 2 % of total P input.

Atmospheric deposition of P is currently, according to the Nutrient Budgets handbook, not included in the practical implementation.

Reference area has a large impact on the final balance per hectare. After several discussions with countries and other stakeholders during the process of the last revision of the Nutrient Budgets handbook it was agreed to define the reference area as the total UAA. An important issue still remains unsolved, however. According to the handbook the balance should in principle relate to the potential fertilised area, excluding very extensive unfertilised areas, to make comparisons useful and to identify the potential risks of agricultural production to the environment in respect to nutrient surplus and deficits. After numerous discussions with stakeholders it is still unclear how to deal with extensively managed areas, for instance the non-herbaceous rough grazing areas mostly in Mediterranean regions, or mountain meadows in some other countries. The balance is an average for a country (or region) which can have both extensive and intensive agriculture. When extensively or sparsely used rough grazing or mountain areas are included, results of the balance for the country or region may be diluted, although in parts of the region the nutrient use can be quite high. This is still an open issue to which a solution needs to be found.

Climatic conditions have a big impact on the balance through the impact on yield and therefore N output. Climate and weather conditions are beyond the control of the farmer. To dampen the effect of weather conditions on the balance the results presented in this fact sheet with regards to the nutrient balance are presented not referring to a particular year but as an average for a certain period.

Subindicator: Vulnerability to phosphorus leaching/run-off

The methodology of the sub-indicator is still under development. Several factors influence the risk of P leaching and run-off such as soil erosion, drainage and sorption capacity. The retention map from the study commissioned by the European Commission in 2005 Addressing phosphorus related problems at farm practice was created based on the Soil Geographic Database of Europe. Pedotransfer rules were used to define areas at potential risk, i.e. with a low sorption capacity, high erosion rates and increased risk of accelerated drainage. Because of the unclear effect of the factor drainage and the lack of reliable data for the EU-25, efforts were focused on sorption capacity and erosion risk. Five vulnerability classes for phosphorus retention capacity were determined, whereby soils with sandy texture, poor drainage and wet water regime or high water table, with low pH and with low content of sesquioxides and/or soluble salts are most susceptible. The sorption vulnerability class was corrected through iteration for erosion, runoff and drainage processes based on soil physical properties such as soil texture, erodibility and slope.

Another factor which influences the risk of P leaching and run-off is the application of P to the soil in the present and in the past. A high value of soil test P (STP) is an indication of a strong accumulation of P in the past. Unfortunately, a database of STP as a risk indicator is not available in most EU countries.

A workshop organized by the Diredate project in Sevilla on 28 September 2010 with experts on the issue identified data needed to build the sub-indicator:

Data needed Description Possible source of data
Soil P status The release of P into a mobile water phase (surface and subsurface) depends largely on the degree of P saturation in soils, which typically is related to soil test P. Similarly, the release of colloidal P is also related to soil test P. LUCAS survey.
Soil type Soil texture and mineralogy affect the risk of erosion and macropore transport as well as colloid mobilization. P retention capacity depends on the contents of active P sorbents and pH. Partial data from the LUCAS survey (because LUCAS only characterizes the superficial texture of the soil).
Tillage and cropping practice The vulnerability to erosion and surface runoff is affected by plant and residue cover, surface roughness, tramlines and traffic-induced soil compaction AEI 21 - Soil erosion, and use of JRC models.
Fertiliser and manure use and practices There is a risk of direct P loss from fertilisers and manures when they are not incorporated into soils during periods with high precipitation. Also, historic P applications affect soil P enrichment and hence the soil P status and degree of P saturation. SAPM
Climate Total rainfall and its distribution over the year affect the hydrological pathways and the vulnerability to leaching and runoff. Available from JRC database.
Drainage Land drainage creates shortcuts between P-enriched soils and water courses. SAPM.
Topography Surface and subsurface runoff depend to a large degree on terrain attributes like slope gradient and upslope contributing area as well as the distance of effective discharge and eroding areas to water courses. Available from JRC database.
Landscape structure The connectivity between P sources and receiving waters depends on field size and shape as well as landscape elements like hedges, roads, ditches and buffer zones. SAPM. Administrative data possibly available as an outcome of the greening of direct payments, if landscape elements are registered in LPIS. Field size available from LPIS as well.


The gross phosphorus balance provides insight into links between agricultural phosphorus use, losses of phosphorus to the environment, and the sustainable use of soil nutrient resources. A persistent surplus indicates potential environmental problems, such as phosphorus leaching resulting in pollution of drinking water and eutrophication of surface waters. A persistent deficit can impair the resource sustainability of agriculture soil through soil degradation, or soil mining, resulting in declining fertility in areas under crop or forage production.

Furthermore, EU-28 depends almost entirely on imports of phosphorus for fertilisation as there are no significant phosphate rock reserves (the main source for the production of phosphorous fertilisers) in the EU[2]. Phosphate rock is a non-renewable natural resource. Estimations on the global availability of phosphate rock and the rate of depletion vary[3]. Though there is uncertainty on the lifetime of remaining high quality phosphate rock reserves, there is a general consensus that the quality of remaining reserves is in decline, that phosphate layers become more difficult to access, that more waste is being generated and that costs are increasing[4]. A sustainable use of phosphorus is needed to ensure food supply in the future and to reduce negative impacts of waste of natural resources on the environment. These include, among others, appropriate fertilisation practices, reduction of imbalances in P inputs and outputs to agricultural soils, recovery of phosphorus from sewage for fertilisation.

Policy relevance and context

International conventions of importance to phosphorus use in agriculture include inter alia UNEP/MAP (United Nations Environment Programme/Mediterranean Action Plan), CBD (Convention on Biological Diversity) and OSPAR (Oslo & Paris Convention to prevent pollution). Such international treaties often give an impetus to harmonise standards amongst all Member States of the European Union. Despite the significant off-site impact that diffuse contamination of phosphorus from agricultural land poses, there is no specific legislation that is directly concerned with the use of phosphorus in agriculture at European level. There is a lack of appropriate institutional arrangements specific to the environmental pollution of phosphorus. Aspects of the phosphorus problem are however integrated in several policy areas and related legal instruments at European level. This section provides an overview of existing regulations and directives dealing with farm-level nutrients (and phosphorus) use and production at international and European level (see also the sections on policy relevance and context in AEI 5 - Mineral fertiliser consumption).

  • The Water Framework Directive (WFD) (Directive 0060/2000) is a legal obligation to protect and restore the quality of waters across Europe. Measures applied under the Water Framework Directive affecting the use of phosphorus in agriculture relate to best environmental practices and include the reduction of nutrient application, the modification of cultivation techniques, the proper handling of pesticides and fertilisers, and the prevention of soil erosion through erosion minimising soil cultivation. Most measures suggested in this context are aimed at reducing the influx of nutrients, such as nitrogen, phosphorus as well as pesticides to the groundwater as well as surface waters. The P balance surplus is a commonly used indicator for identifying areas vulnerable to nutrient pollution in the pressures and impacts analysis.
  • The Nitrates Directive (ND) (Directive 0676/1991), established in 1991 aims to reduce water pollution caused or induced by nitrates from agricultural sources and prevent further such pollution. The Water Framework Directive explicitly refers to the Nitrates Directive for information on diffuse pollution of nitrates from agricultural activities and extends this to phosphates. The measures established within the Action Programmes aim to control diffuse and direct water pollution and also influence the use of phosphorus in farm practice. For instance, by limiting the annual application of nitrogen fertiliser and livestock manure, defining legally binding maximum concentrations of nitrates in drinking water and designating periods when the application is prohibited, the directive clearly aims at establishing and maintaining the natural balance of fertilisers in soils. Through these measures a massive influx of nutrients to ground- and surface water and thus potential eutrophication is prevented, while excess nutrients, oversaturation and a possible ensuing degradation is avoided at the same time. Currently the Action Programmes are no longer expected just to address losses of nitrate-nitrogen but are also judged on their ability to reduce losses of phosphorus.
  • The 7th Environmental Action Programme encourages the full implementation of WFD, in order to achieve levels of water quality that do not give rise to unacceptable impacts on, and risks to, human health and the environment.
  • Rural Development Programme (RDP): Various agri-environment measures have been established throughout the European Union directly or indirectly addressing diffuse contamination by phosphorus. Some of these measures are directed at mitigating soil erosion such as crop rotations, mulch seeding retaining stubble after harvest and ploughing restrictions. Other measures tackle the problem of excess nutrients through reduced fertiliser use. All measures that impact soil erosion and nutrient balances ultimately result in a reduction of diffuse contamination by phosphates from agricultural land.
  • The Directive on Integrated Pollution Prevention and Control (IPPC) (Directive 0001/2008) introduces an integrated cross-media approach, aiming to prevent or minimise emissions to air, water and land, as well as to avoid waste production with a view to achieving a high level of environmental protection as a whole. The IPPC Directive concerns highly polluting industries, among which are intensive pig and poultry farms. A single permit based on the concept of Best Available Techniques (BAT including limit values) must include all arrangements made, including emission limit values for pollutants, for water, air and land, and may, if necessary, contain requirements for the protection of the soil and the groundwater as well as measures or waste management (Art. 9(3)) in order to continuously prevent and reduce pollution. The purpose of the IPPC Directive was to achieve integrated prevention and control of pollution arising from several categories of industrial activities. The indicative list of main polluting substances to be taken into account if they are relevant for fixing emission limit values includes oxides of nitrogen and substances which contribute to eutrophication (phosphates and nitrogen).
  • Habitats Directive (Directive 0043/1992) and Birds Directive (Directive 0147/2009): The main purpose of this directive is to ensure biological diversity through the conservation of natural habitats and wild flora and fauna within the European territory, while taking into account economic, social, cultural and regional requirements. Farmers who have agricultural land in Natura 2000 sites and face restrictions due to the requirements of the Habitat-Directive are eligible to receive payments for the management of these sites by the Rural Development Regulation, which helps promote environmental-friendly farming. Depending on the specific conditions of a certain area, these include measures to reduce the use of pesticides and fertilisers, measures to mitigate the effects of soil compaction, e.g. limitations on the use of machinery or the setting of stocking limits, or measures aiming to regulate the irrigation of agricultural land.

Pollution by phosphorus is also partially covered by the Directive on Bathing Water (Directive 0007/2006), the Directive on Sewage Sludge (Directive 0278/1986) and the Directive on Urban Waste Water Treatment (Directive 0271/1991), which together with the Nitrates Directive and the Directive on Integrated Pollution Prevention Control, have been linked since 2000 in the Water Framework Directive.

Agri-environmental context

The gross phosphorus balance indicates the total potential risk to the environment (water and soil). The input side of the balance counts all P supplied to the soil. The output side of the balance presents the nutrients removed from the field by agricultural production; harvest or grazing. Sustainability could be defined as preserving and/or improving the level of production without degrading the natural resources. The harvest and grazing of crops and fodder means that P is removed from the soil. To sustain soil fertility this removal of P should be compensated. Fertilisers and manure are therefore imperative to supply the crops with the necessary P for growing.

Depending on the soil P capacity, excessive P can be stored in the soil. Not all of P in fertilisers and manure are directly available to the plant, a part is converted from active P to fixed P. The active P pool is the main source of available P for crops, whereas the fixed P pool contains inorganic phosphate compounds that are very insoluble and organic compounds that are resistant to mineralisation by microorganisms in the soil. Phosphate in this pool may remain in soils for years without being made available to plants and may have very little impact on the fertility of a soil. Depleting the active pool through crop uptake may cause some of the fixed P to slowly become active. The storage capacity of the soil, however, depends on soil characteristics like soil texture. An important aspect of the ability of a soil to hold phosphate is that a soil cannot hold increasing amounts of phosphate in the solid phase without also increasing soil solution phosphate. Increased amounts of phosphate in solution will potentially cause more phosphate to be lost to water running over the soil surface or leaching through the soil. Loading soils with very high levels of phosphate will generally not hurt crops but may result in increased phosphate movement to nearby bodies of water[5]. It has been estimated that 25 %[6] or less of P applied annually is actually taken up by the growing crop; the remaining 75 % becomes bond in the soil profile or is lost to water. The crop uptake of P is in sharp contrast to the crop use of N and K fertilisers, where the recovery in the season of application can be as high as 80 %[7]. Yield and therefore the uptake of P by crops is not only determined by inputs but also by uncontrollable factors like type of soil and climate.

The estimated P surplus represents the potential risks to water and soil. The actual P loss from agricultural land to surface waters is a complex function of climate, topography, soil type, soil P status, P fertilisation, and land management. These factors vary greatly in space and over the year, and the hydrological pathways for P losses also vary greatly in space and time. The effects of these individual factors on P loss from agricultural land to surface waters are rather well understood, but in combination the understanding is less well developed. As a consequence, current simulation models do not yet simulate the actual P loss from agricultural land to surface waters very accurately. Further, the ecological effect of P from agricultural land in surface waters depends on the 'bio-availability' of the P. A significant fraction of the P from agricultural land that is lost to surface waters is particulate P and the availability of this P to growing organisms in surface waters is much less than dissolved reactive P, at least in the short term. Hence, ‘the pollution effect’ of P lost from agricultural land to surface waters does not depend on the total P loss but on the fraction 'reactive P'.

The accumulated amount of P represents a larger actual risk for the environment than the amount of P which is applied yearly, or the surplus on today’s balance. However, the future risk is strongly influenced by the P balance of today and the near future, in combination with the capacity of a soil to bind P. This contrasts with nitrogen; the cumulative N balance of the past is much less relevant since N hardly accumulates in soils. This is why the main indicator on its own is not sufficient to indicate areas at risk of phosphorus pollution.

The gross phosphorus balance also gives information on the dependence on mineral fertilisers. Phosphorus, contrary to Nitrogen, is a finite resource – in the sense that phosphate ores are becoming depleted – and after being used in agriculture, the mineral ultimately becomes unavailable for reuse, or can be reused only to a very limited extent. Phosphorus builds up in agricultural soils, it ends up in the sludge from water treatment plants – which is largely incinerated and thereby removed from the agricultural cycle – but mostly it is eroded, thereby ending up in the sediments of lakes, coastal seas and the ocean[8]. Recycling from ocean sediments takes place during a period of millions of years[9].

The natural level of phosphate in the soil is very low and limits plant growth. Before the industrial revolution, phosphate was a limiting factor nearly everywhere on Earth. A breakthrough occurred in the 19th century by means of the external application of phosphate as a fertiliser, where the mineral originated outside the area to be fertilised. As a result of this external source, the local phosphate cycles were broken; at the same time, this created the basis for a potential shortage in the future. This external application of phosphate (and other fertilisers) from external sources was one of the preconditions for the increase in the world population from approximately 1 billion people in 1850 to the current population of 6.8 billion[10]. The EU is almost entirely dependent on imports of phosphate; very little is mined within the EU. Import takes place in two forms: artificial fertiliser and animal feed. In total, 1.6 Mt P per year is imported into the EU, where the phosphate contained in animal feed ultimately ends up in animal manure. The total phosphorus import of the EU concerns nearly 10 % of the entire world production of artificial phosphate fertiliser. It is important to note that the EU, other than is usually assumed, is not at all self-sufficient in food production due to this phosphate dependency[11].

See also

Further Eurostat information



  • Agriculture, see:
Agriculture and environment (aei)
Gross nutrient balance (aei_pr_gnb)

Dedicated section

Methodology / Metadata

Source data for tables, figures and maps (MS Excel)

Other information

Legislation: Commission Staff working document accompanying COM(2006)508 final

External links

  • European Commission - Agriculture and Rural Development
  • European Commission – Environment
  • European Commission – Climate Action


  1. Summer 2003 was exceptionally hot over most of central and western Europe, ranging from Spain to Hungary and from Iceland to Greece (Fink 2004, Schaer 2004). Summer 2007: sharp differences in weather conditions between the eastern (warmer and drier than average) and the western side (cooler and wetter) of Europe; a prolonged dry spell in central Mediterranean, Black Sea and Baltic areas; in contrast, abundant and persistent rain in central and northern EU areas, interfering with field activities and affecting crop yield (Bulletin pdf)
  2. Schröder, J.J., Cordell, D., Smit, A.L; Rosemarin, A. (2010) Sustainable Use of Phosphorus (European Union tender project ENV.B.1/ETU/2009/0025). Report 357, Plant Research International, Wageningen University and Research Centre, Wageningen, The Netherlands, 122 pp.available here
  3. See for example: Schröder, J.J., Cordell, D., Smit, A.L; Rosemarin, A. (2010) Sustainable Use of Phosphorus (European Union tender project ENV.B.1/ETU/2009/0025). Report 357, Plant Research International, Wageningen University and Research Centre, Wageningen, The Netherlands, 122 pp. Available here. Conclusions of the Expert Seminar on the Sustainability of Phosphorus Resource.
  4. Schröder, J.J., Cordell, D., Smit, A.L; Rosemarin, A. (2010) Sustainable Use of Phosphorus (European Union tender project ENV.B.1/ETU/2009/0025). Report 357, Plant Research International, Wageningen University and Research Centre, Wageningen, The Netherlands, 122 pp. available here.
  5. The Nature of Phosphorus in Soils, Lowell Busman, John Lamb, Gyles Randall, George Rehm, and Michael Schmitt, University of Minnesota
  6. Morgan, M.A., 1997. The behaviour of soil and fertiliser phosphorus. In: Tunney, H., Carton, O., Brookes, P. and Johnston, A. (eds.) Phosphorus loss from soil to water. CAB International, Oxon, UK
  7. Bomans E., Fransen K., Gobin A., Mertens J., Michiels P.,Vandendriessche H., Vogels N. Addressing phosphorus related problems in farm practice. Final report to the European Commission. Soil Service of Belgium.
  8. Phosphate – from surplus to shortage. Policy memorandum of the Steering Committee for Technology Assessment of the Ministry of Agriculture, Nature and Food Quality. H.A. Udo de Haes, J.L.A. Jansen, W.J. van der Weijden and A.L. Smit, Utrecht, September 2009
  9. Phosphorus in agriculture: global resources, trends and developments: report to the Steering Committee Technology Assessment of the Ministery of Agriculture, Nature and Food Quality, The Netherlands; Smit, A.L; Bindraban, P.S; Schröder, J.J.; Conijn, J.G.; Meer, H.G.; van der Wageningen: Plant Research International 2009
  10. Phosphate – from surplus to shortage. Policy memorandum of the Steering Committee for Technology Assessment of the Ministry of Agriculture, Nature and Food Quality H.A. Udo de Haes, J.L.A. Jansen, W.J. van der Weijden and A.L. Smit, Utrecht, September 2009
  11. Phosphate – from surplus to shortage. Policy memorandum of the Steering Committee for Technology Assessment of the Ministry of Agriculture, Nature and Food Quality H.A. Udo de Haes, J.L.A. Jansen, W.J. van der Weijden and A.L. Smit, Utrecht, September 2009