D1hu4133i4rt3z.cloudfront.net



Report for Defra Projects NT2601 and NT2602

Evaluation of urea-based nitrogen fertilisers

Edited by

Anne Bhogal, ADAS Gleadthorpe

Peter Dampney, ADAS Boxworth

Keith Goulding, Rothamsted Research

October 2003

Contents

Abbreviations

Glossary of terms

1. Executive summary 6

2. Introduction 12

3. Characteristics of urea 14

3.1 World consumption and production 14

3.2 Chemical and physical properties 14

3.3 Current use on UK farms 14

3.4 Conclusions 15

4. Behaviour and fate of urea-N applied to soil 16

4.1 Fate of urea applied to soil 16

4.2 Conclusions 19

5. Agronomic effectiveness 21

5.1 Arable cropping 21

5.2 Grassland 30

5.3 Horticultural crops 33

5.4 Conclusions 43

6. Environmental impacts 46

6.1 Ammonia emissions 46

6.2 Nitrous oxide emissions 51

6.3 Nitric oxide emissions 56

6.4 Leaching and surface runoff 56

6.5 Conclusions 58

7. Methods to mitigate ammonia emissions 61

7.1 Slow release formulations of urea 61

7.2 Chemical additives 61

7.3 Inorganic salts 62

7.4 Pellet size and soil incorporation 62

7.5 Urease inhibitors 63

7.6 Conclusions 78

8. Effects on soil processes 81

8.1 Factors affecting soil processes 81

8.2 Impacts of urea on biological processes 82

8.3 Impacts of urea on chemical processes 83

8.4 Conclusions 84

9. Modelling ammonia emissions 85

9.1 Modelling process stages 86

9.2 Scenario testing 90

9.3 Choice of models for future use 96

9.4 European approaches to modelling ammonia emissions 97

9.5 Conclusions 99

10. Implications for new research and other studies 101

11. References 103

Abbreviations

|AC |Ammonium carbonate |

|ACl |Ammonium chloride |

|AN |Ammonium nitrate |

|APP |Ammonium polyphosphate |

|ATS |Ammonium thiosulphate |

|AS |Ammonium sulphate |

|ASN |Ammonium sulphate nitrate |

|AnA |Anhydrous ammonia |

|AqA |Aqueous ammonia |

|BSFP |British Survey of Fertiliser Practice |

|CAN |Calcium ammonium nitrate |

|CC |Calcium cyanamide |

|CEC |Cation Exchange Capacity |

|CN |Calcium nitrate |

|Chilean CN |Chilean potassic nitrate |

|CORINAIR |Core Inventory of Air Emissions in Europe |

|CV |Coefficient of Variation |

|CDU |Crotonylidenediurea |

|DAP |Di-ammonium phosphate |

|EF |Emission factor |

|EMEP |Cooperative programme for monitoring and evaluation of the long-range transmision of air pollutants in Europe |

|FMA |Fertiliser Manufacturers Association |

|FSU |Former Soviet Union |

|HSE |Health and Safety Executive |

|IBC |Intermediate Bulk Container |

|IBDU |Isobutylidene urea |

|MgAP |Magnesium ammonium phosphate |

|MgN |Magnesium nitrate |

|MU |Methylene urea |

|MAP |Mono-ammonium phosphate |

|MOP |Muriate of potash |

|N |Nitrogen |

|NARSES |National Ammonia Reduction Strategy Evaluation System |

|OSN |Other straight nitrogen |

|Ox |Oxamide |

|KN |Potassium nitrate |

|PSDA |Product Safety data Sheet |

|SMB |Soil microbial biomass |

|SSP |Single superphosphate |

|NaN |Sodium nitrate (nitrate of soda) |

|SCU |Sulphur coated urea |

|TAN |Total Ammonical Nitrogen |

|TSP |Triple superphosphate |

|U |Urea |

|UAN |Urea ammonium nitrate |

|UAS |Urea ammonium sulphate |

|UCN |Urea calcium nitrate |

|UKAEI |United Kingdom Ammonia Emissions Inventory |

|UNECE |United Nations Economic Commission for Europe |

|UP |Urea phosphate |

Glossary of terms

|Blended fertiliser |Compound fertiliser produced by dry mixing of two or more different particulate or powder materials. |

|Bulk density |Density of a mass of material, often expressed as kg/litre. The mass comprises the particles and the air spaces |

| |between them so bulk density is determined by the shape and size of particles as well as by the true density of the |

| |material from which the particles are formed. Particulate materials show differences in bulk density between loose |

| |and tamped or shaken states, in some materials as great as 15%. The bulk densities shown are intended to describe |

| |those of material in a spreader hopper. A value of 1.00 kg/l means that a 1000 litre hopper should hold 1tonne of |

| |material. |

|Caking |Formation of large hard agglomerations of fertiliser particles due to chemical properties of the materials or to |

| |absorption of water. This phenomenon occurs when fertiliser granules adhere to one another through crystal bridges or|

| |plastic deformation. |

|Complex fertiliser |Compound fertiliser where all particles have the same composition. |

|Compound fertiliser |Product containing more than one of the major nutrients. |

|Deliquesce |Absorption of atmospheric water vapour resulting in the loss of physical structure of particles. |

|Fluid fertiliser |Products supplied in liquid form, either as solution or suspension. |

|Granulation |Methods of forming fertiliser particles, mainly in the range 2 to 4mm diameter. There are two main classes of |

| |granulation: slurry and non-slurry processes. In slurry processes, solid particles of the fertiliser (obtained |

| |through recycling of undersize particles) are coated with a slurry of the fertiliser in successive layers. In |

| |non-slurry processes, a liquid component is added to finely divided particles causing them to agglomerate. Most |

| |granular products are slightly irregular in shape but some, those made by fluidised bed processes for example, are |

| |nearly spherical. |

|Granular fertiliser |Solid fertiliser where particles are all produced by granulation. May be complex or blended though the term is |

| |sometimes erroneously used as an alternative to complex. |

|Hygroscopic |Material absorbs moisture from the air. |

|IBC |Intermediate bulk container or big bag, usually containing 500, 600 or 1000kg of fertiliser. IBC also can refer to |

| |1m3 containers of solution fertiliser. |

|Median size |The particle size at which 50% of the material by weight is smaller and 50% larger. The median size can vary in some |

| |materials and the values shown should be treated as guides. The particle size for most manufactured granular and |

| |prilled fertilisers is in the range 2 to 4mm range. |

|Particle crushing strength |Force that must be applied to cause a particle to shatter or break. Measured in newtons (N). |

|Particle or true density |Density of the solid material from which the particles are formed. Particle density therefore is independent of |

| |particle size and shape. The weight of a particle is determined by it’s size and density and is an important factor |

| |in spreading properties. |

|Prilling |Method of particle formation in which the molten fertiliser is forced through holes in a metal disc or spinning |

| |bucket and allowed to fall as droplets in a tower. The particles solidify as they fall. Prills tend to be more |

| |spherical and slightly smaller than granules |

|Solution fertiliser |Products where the nutrients are present in true solution. |

|Straight fertiliser |Product containing only one of the major nutrients (nitrogen, phosphate or potash) |

|Suspension fertiliser |Products where the nutrients are present partly in solution and partly as finely divided particles in suspension. |

Executive summary

1. This report forms part of the NT2601 and NT2602 projects for Defra. It describes and discusses existing knowledge on the effects of using urea-based nitrogen (N) fertilisers on the performance of arable, grassland and horticultural crops, and likely impacts on the air, water and soil environments. The report discusses possible mitigation options to minimise or avoid adverse effects. The information sources comprised published international literature, as well as information provided by representatives of the UK and international fertiliser industries. Other reports from the NT2601 and 2602 projects cover ‘Nitrogen fertilising materials’ and ‘Production and use of nitrogen fertilisers’.

Characteristics of urea

2. Urea is the predominant source of fertiliser nitrogen used in agriculture throughout the world (50% of total N use). However, in the EU-15, the predominant source (40%) of N is ammonium nitrate (AN) or calcium ammonium nitrate (CAN), and AN is used predominantly on UK farms. There is no production capacity for urea in the UK. All current supplies are imported from within or outside Europe.

3. Urea can be manufactured as prills or granules. However, because urea is very hygroscopic (i.e. absorbs water), its use as a raw material in the production of compound fertilisers is much less flexible and more limited than for AN or CAN.

4. Urea is primarily used as a solid straight N fertiliser (46% N) or in the production of urea ammonium nitrate (UAN) solution (28-30% N w/w). Solid urea represents 9% and UAN 10% of the total UK consumption of N-containing fertilisers. Most (95%) is applied as a topdressing to winter cereals (63%), oilseed rape (16%) or grass (17%), largely in the February to April period; only 12% is applied in the warm and dry months of May to August. Very little urea (2% of total) is used on spring cereals and virtually none on potatoes, sugar beet or horticultural crops.

5. Urea has a lower bulk density than AN which makes accurate spreading by spining disc more difficult. A small particle size, as usually found with urea prills, aggravates this problem. Urea granules are larger (up to 3.5mm diameter), which should improve the spreading accuracy.

Behaviour and fate of urea applied to soil

6. Following application to soil, urea undergoes hydrolysis to ammonium (NH4) which is then subject to the same chemical and biological transformations as AN and other N fertilisers. The hydrolysis process is controlled by the urease enzyme (which is ubiquitous in soil, on vegetation and in surface litter), urea concentration, soil temperature and moisture. Grassland soils have more urease enzyme activity than arable soils. Rates of hydrolysis are generally rapid in most UK soils, but could be slower in arable soils low in organic matter, or in very dry, very wet or very cold weather.

7. Hydrolysis of urea results in a localised very high soil pH which can result in large emissions of ammonia to the atmosphere (see also paras 20-41). This is a well documented major loss process and is the main reason why urea has often been shown to be less effective for crop uptake compared to nitrate based fertilisers (see also paras 10-19). More research is needed to quantify ammonia emissions under different UK soil and agricultural conditions.

8. Unhydrolysed urea is soluble in soils and there is a risk that heavy rain immediately after urea application could wash urea and/or ammonium into surface or groundwaters, but there is little existing data. In soils above neutral pH, nitrite (NO2) could accumulate with risk of plant damage and leaching to waters. Both nitrate (NO3) and nitrite are at risk of loss as nitrous oxide (N2O) gaseous emissions. These transformations and processes make the efficiency of use of urea more difficult to predict and manage compared to AN.

9. There is no evidence that continued use of urea will have any long-term adverse effects on the national soil resource. Compared with the other major factors controlling biological and chemical processes in soils (e.g. pH, organic matter content), impacts arising from nitrogen fertilisers have always been observed to be small. No impact of urea (or the urease inhibitor Agrotain – see paras 29-33) on the soil microbial biomass (SMB) was observed in a 1-year trial. The spatial and temporal variations in microbial, chemical and biochemical properties were found to be much larger than any changes resulting from urea or Agrotain use.

Agronomic effectiveness of urea

Arable crops

10. There have been many studies comparing the agronomic efficiency of solid urea with other N fertilisers, largely from trials between 1960-1980 on winter cereals. The general conclusion was that urea gives more variable results and sometimes only 80-90% of the yield produced from other solid N fertilisers. Yield reductions have largely been attributed to a poorer efficiency of use of urea-N by the crop due to ammonia volatilisation losses post-application. A few studies comparing UAN solution with urea and AN have shown that UAN can give similar yields to urea (and sometimes less), and lower yields than AN.

11. Some trials have reported higher optimum rates of N (Nopt) from use of urea compared to AN, but lower yields at these N rates; however, in most cases the statistical significance of any differences was not reported. Where errors could be estimated there was no significant difference in the mean Nopt. Agronomic studies have shown no clear benefits from splitting urea applications. However, splitting of urea applications might be a strategy to consider to reduce ammonia emissions and increase the effectiveness of urea.

12. Statistically significant decreases in wheat grain N (protein) content have been reported from the use of urea compared to AN, typically ranging between 0.05-0.15% N in wheat (0.3-0.9% protein at 100% DM). Protein content is important for wheat grain marketing. Many reports have shown that foliar applications of straight urea solution are effective for increasing grain protein content, but these have a poor N use efficiency.

13. A few studies have reported that the effectiveness of urea appeared lower on calcareous soils, perhaps because of slower hydrolysis limiting the availability of N at a critical growth stage or due to higher ammonia losses. There appeared to be no effect of soil texture on the effectiveness of urea-N. Some trials have indicated a positive relationship between the effectiveness of urea and rainfall, but most reports gave little or no information on the prevailing weather, soil moisture or wind conditions.

14. A few studies have been done on the effectiveness of urea incorporated into crop seedbeds, but these did not record any adverse effects on germination or establishment of spring cereals where up to 90kg N/ha had been applied. However, higher rates of seedbed N for spring cereals (and oilseed rape) are currently used in the UK. Combine-drilled urea can result in reduced establishment and crop yields, but this practice is not recommended in the UK.

15. For sugar beet, no significant differences in sugar yield between urea and other nitrogen fertilisers were recorded. Reductions in plant population occurred at 3 trials, where 60 or 120kg N/ha as urea was applied in the seedbed, but this was unlikely at the normal seedbed recommendation of 40kg N/ha.

Grassland

16. Most grassland studies comparing urea with other N materials were carried out before 1985. The general conclusion was that urea was as effective as CAN or AN for spring grass production, but can result in 5 to 15% lower yields when used in the summer. Large yield reductions were observed on light textured soils and in dry weather periods. Rainfall in the 3 days after fertiliser application has been shown to increase the effectiveness of urea. Little research has been carried out under grazing conditions.

Potatoes and horticultural crops

17. Many potato experiments comparing urea with other N fertilisers have been carried out, but relatively few on other vegetable crops. Responses tended to vary with soil type and weather conditions but, with a few exceptions, equivalent yields were obtained where urea was used.

18. Some horticultural studies have reported that urea produced lower yields, e.g. on brassicas, lettuce, onion, beans, red beet, tomatoes. These reports far outnumbered those where urea produced larger yields, and suggest that there may be significant risks associated with the use of urea for horticultural crops. Reasons proposed to explain the lower yields were the same as for arable and grass, i.e. ammonia loss and phytotoxicity.

19. The likelihood of adverse reactions to urea is greatest for young plants shortly after N application. At this stage, plant tissues are more sensitive and urea concentrations in the soil are at their highest. Strategies to minimise damage to young plants should be based on avoiding high concentrations of urea in the seedbed – e.g. use of nitrate-based fertilisers, placement or splitting of urea, use of controlled-release fertilisers.

Environmental impacts

Ammonia emissions

20. Atmospheric pollution with ammonia has impacts on the acidification of land and eutrophication of water. The UK has a commitment under the EU National Emission Ceilings Directive and the UNECE Gothenburg Protocol to reduce ammonia (NH3) emissions to 297kt NH3/yr by 2010, compared with emissions of about 348kt NH3/yr in 1999. Because of the much higher risk of ammonia emissions following use of urea compared to nitrate-based fertilisers, a significant change in national fertiliser practice away from AN towards urea would have a serious impact on the UK’s obligations to meet this target.

21. The risk of ammonia emission following use of urea is much greater than following use of nitrate-based fertilisers, but the level of risk varies. The greatest risks are on coarse-textured/low organic matter soils, where crop cover is low, and where conditions are dry, warm and windy following application.

22. Studies measuring ammonia emissions following urea use, and its effect on N use efficiency and crop performance, have given highly variable results. Field measurements of emission factors have ranged from 4 to 47% (arable crops) and 6 to 46% (grassland) of the urea N applied. These can be compared with emissions of less than 4% of N applied as AN or CAN. Within the arable experiments, the greatest emissions have been measured from no-till systems (10 to 47%). Emission factors for urea used on cultivated cereals ranged from 4 to 19%, although there were few studies on tilled land.

23. Ammonia emission inventories have used a range of factors to calculate emissions from urea applications. Early UK experiments suggested an average emission factor of 10% for urea. More recently, a factor of 23% has been used in the UK for urea applied to grassland. In the absence of any direct field measurements, emissions from arable land were considered to be half those from grassland (i.e. 11.5%).

24. Information on ammonia emissions from N fertilisers other than urea is sparse, so ammonia emission inventories have tended to group all N fertilisers together except for urea. Emissions from AN and CAN tend to be small and current inventories use an emission factor of 1.6% and 0.8% of the N applied to grass and arable crops, respectively. Ammonium sulphate (AS) is also often grouped with AN and CAN, but separate factors have been proposed depending on the soil pH (2% on soils with pH7).

25. The UK Ammonia Emission Inventory (UKAEI) emission factors proposed for grassland and arable land are:-

urea 23 % (grass) 11.5% (arable)

all other N forms 1.6% (grass) 0.8% (arable)

Ammonia emissions - mitigation options

26. Ammonia emissions from urea could potentially be reduced by slow release formulations, chemical additives, larger particle size, soil incorporation or use of urease inhibitors. The use of a urease inhibitor offers the best prospect; the other options have constraints of very high cost (e.g. slow release fertilisers), limited effect (e.g. large particle size), or lack of practical opportunities (e.g. soil incorporation).

27. Incorporation of urea below the soil surface will minimise ammonia volatilisation but this would be impractical in most agricultural systems as 95% of fertiliser N is topdressed to growing crops. Soil incorporation may be possible for fluid fertiliser applications if some means of injection into the soil was adopted, but capital and time-related costs would be major limiting factors. Band-spread urea has some potential as less urea is in contact with the soil compared with broadcast urea. But the high concentrations of urea might increase ammonia emissions and root growth in the band might be restricted.

28. Inhibiting urease activity slows the conversion of urea to ammonium-N and hence the potential for ammonia volatilisation and seedling damage. Slowing the hydrolysis allows more time for the urea to diffuse into the soil, or for rain or irrigation to occur. Thousands of chemicals have been tested for their potential as inhibitors of soil urease activity, but few have proved both effective and commercially attractive.

29. The only current practicable option is N–(n-butyl) thiophosphoric triamide (nBTPT) which is commercially available (trade name Agrotain). Numerous field studies with nBTPT-coated urea have been conducted in the USA with arable crops and grass, where its use has increased yield and N uptake compared with untreated urea.

30. Apart from work in Northern Ireland and Defra funded studies in 2003, there is no British data on the use of nBTPT treated urea. A few studies in Europe have shown that nBTPT-coated urea can reduce ammonia losses from surface applications of urea. Increasing the concentration of nBTPT has been shown to reduce ammonia losses according to the law of diminishing returns; the inhibitor was very effective at low concentrations, resulting in approximately 50% inhibition at a concentration of 0.01%. There was little benefit in using concentrations above 0.1%. Use of nBTPT has been shown to reduce seedling damage from seed-placed urea and to improve the emergence of cereal seedlings with urea under simulated combine drilling conditions in a greenhouse.

31. In Northern Ireland studies on grassland, there was no evidence of any long-term adverse effects on grass production with repeated applications of nBTPT-coated urea over a 3-year period, and no indication that its efficacy to reduce ammonia emissions declined when used repeatedly. nBTPT has been shown to have no effect on N mineralisation or on the size and activity of the soil microbial biomass; it does not inhibit nitrification or denitrification.

32. Agrotain has successfully passed US-EPA toxicological and environmental tests and degrades into fertiliser elements N, P and S after approximately 10-14 days. Some information indicates that the shelf life of Agrotain treated solid urea is dependent on the nBTPT concentration and the way that it is applied, but other information indicates that nBTPT in treated urea does degrade within months. A new stabilisation technique has suggested that Agrotain added to the urea melt (thus incorporated within the urea granule) is stable for several years. nBTPT can be added to UAN solutions, but advice is that these should be applied soon after mixing, as nBTPT gradually decomposes in the presence of water.

33. Agrotain treated urea offers the best current prospect for a modified urea fertiliser that might provide an effective alternative to AN for crop production with minimal impact on the environment. However, more UK-based research is needed.

Ammonia emissions - modelling

34. Ammonia emission models are required to predict future emissions in the event of increased urea usage and to assess the potential impact of different mitigation methods. Unfortunately, specific urea-based models are relatively few. If mechanistic modelling was required, then the model of Rachhpal-Singh and Nye would be the best one to develop. In the short term, the UK NARSES (National Ammonia Reduction Strategy Evaluation System) model provides the best platform to build on for predicting ammonia emissions, following different scenarios of urea use and taking account of a range of factors to predict losses.

35. Generally the approach in mainland Europe is the same as in the UK, utilising inventories and some modelling. The MARACCAS model (Model for the Assessment of Regional Ammonia Cost Curves for Abatement Strategies) has been used to compare emissions from agricultural activities in different European countries and to assess the applicability and efficacy of potential ammonia abatement measures. In 1998, the UNECE Ammonia Expert Group adopted the emission factors used in MARACCAS to revise the guidelines for calculating ammonia emissions. The MARACCAS model is being updated and adopted for use with disaggregated ‘activity’ data in the NARSES model.

36. The impact of a complete switch from use of AN to use of urea on ammonia emissions, and the impact of possible mitigation options, was tested using current emission inventory data. Three approaches were compared - the UK Ammonia Emissions Inventory (UKAEI), the ‘prototype’ NARSES model and the EMEP/CORINAIR Emission Inventory Guidebook. All gave similar results.

37. Total ammonia emissions from current manufactured fertiliser N forms applied to UK agricultural land in 2001 were estimated as 34.7, 37.9 and 49.8kt using the UKAIE, NARSES and EMEP/CORINAIR emission factors, respectively. The models predicted that, if all this fertiliser N was applied as urea, the total emissions from manufactured N fertilisers would increase by around 220kt NH3 to 260kt NH3. This would represent an increase of 75–85% in the total of all ammonia emissions from UK agriculture, including those from livestock manures. For total emissions from fertiliser N to remain the same as 2001, the emission factor for urea would have to be reduced to 2.25% of applied N if urea was used to replace all other N fertilisers, or to 3.1% if urea was used to replace AN fertiliser only.

38. Three potential abatement scenarios were tested: (1) the use of urease inhibitor nBTPT (Agrotain), (2) the application of urea in liquid form and, (3) an increased proportion of urea applied to arable land and incorporated into the soil. The results suggest that, if all urea was treated with the urease inhibitor nBTPT, and assuming an 80-90% reduction in ammonia emissions compared to untreated urea, then the impact on total ammonia emissions from UK agriculture would only be a 5% increase. Liquid application was estimated to half emissions; soil incorporation was assessed to have little effect because of the potential difficulty of incorporation for most tillage crops.

Nitrous oxide emissions

39. Agricultural soils are a major source of nitrous oxide (N2O) emissions, contributing c.50% of total UK emissions of N2O. Modelling has predicted that c.77% of the nitrous oxide from soils is derived from N fertiliser.

40. There have been many studies on the effect of N fertilisers on nitrous oxide emissions, but only a few have studied the form of N used. Current IPCC guidelines for greenhouse gas inventories suggest the use of a single nitrous oxide emission factor of 1.25% (( 1%) for fertiliser applications, with no allowance made for different fertiliser types.

41. Assuming denitrification is the dominant source process for nitrous oxide, emissions will be greater from nitrate-based fertilisers than ammonium-based fertilisers (e.g. urea); the difference will increase as soil moisture content increases. Use of urea in wet springs (when there is a high potential for denitrification) is therefore likely to result in a significant reduction in total annual emissions compared to use of AN. Results from Scotland and reviews of European research support this, but differences are generally small in numerical terms. In Scotland, differences between fertiliser forms have also been more apparent on grassland than tilled land, but the magnitude of these differences was very dependent on the season.

Nitric oxide emissions

42. There is very little published information on nitric oxide (NO) emissions from different forms of N fertiliser. Most nitric oxide emissions are associated with nitrification, so urea would be expected to result in larger emissions that AN. Measured losses of nitric oxide have ranged from 0.003 to 11% of applied fertiliser N, with a mean of 0.3%. The CORINAIR Emissions Inventory Guidebook uses an emission factor of 1%, with no division for fertiliser form or cropping system.

Leaching and surface runoff

43. The risk of direct leaching of any N fertiliser following application is generally regarded as small, unless rainfall follows applications to heavy soils and results in drainflow or surface runoff, or N is applied to young crops with limited rooting systems (e.g. potatoes and spring cereals) in wet springs. There is little evidence of direct leaching of residual, unused fertiliser N at the end of the growing season if the correct amount of N is applied. However, numerous studies have shown an increase in soil mineral N and associated leaching losses from applications in excess of the economic optimum, which could occur if farmers overfertilise with urea N to provide ‘insurance’ against potential NH3-N losses.

44. Urea is non-ionic and therefore susceptible to leaching and runoff. Although there has been very little research on N leaching or runoff from urea or its decomposition products, the potential for leaching has been demonstrated in leaching columns under laboratory conditions; urea was considered to be more susceptible to leaching than ammonium-N, but less than nitrate-N. In one field study, 24% of the applied unhydrolysed urea was lost in runoff following 10mm of rainfall shortly after application to an impermeable grassland soil. Use of urea treated with a urease inhibitor (see paras 28-33) may exacerbate the problem, as this urea will remain unhydrolysed for longer.

45. Whether leached urea would persist until it reaches a watercourse is uncertain. However, the hydrolysis of urea within watercourses is likely to impact on concentrations of ammonium-N, nitrite-N and nitrate-N. This could increase ammonium-N concentrations above the European guidelines for Salmonid and Cyprinid waters. The EU Freshwater Fish Directive (FFD) has set mandatory threshold concentrations for total ammonium-N of 0.78mg/l, and guide levels of 0.03mg/l for Salmonid and 0.16mg/l for Cyprinid fish. Guide levels of nitrite-N have been set; 0.003mg/l for Salmonid and 0.009mg/l for Cyprinid fish.

Introduction

This report summarises existing international knowledge on the use of urea-based fertilisers as a source of nitrogen for use in agriculture. It is part of the suite of reports produced as part of the Defra projects NT2601 and NT2602. Each section or sub-section has been written by a lead author as indicated, with overall editing by Anne Bhogal and Peter Dampney (ADAS), and Keith Goulding (Rothamsted Research).

Section 3 summarises the supply, characteristics and current use of urea. Much of this information has been presented and discussed in detail in the NT2601/2602 reports ‘Nitrogen fertilising materials’ and ‘Production and use of nitrogen fertilisers’. These reports provide detailed information on other N fertiliser materials as well as urea.

Section 4 describes the behaviour of urea-N in soil and the wider environment.

Section 5 reviews existing knowledge on the agronomic effectiveness of urea-based fertilisers compared to other N fertiliser materials in arable, grassland and horticultural cropping systems.

Section 6 reviews existing knowledge on the effects of using urea on the air, water and soil environments. The potential for urea to emit ammonia to the atmosphere is a major concern.

Section 7 considers methods that are or might be used for mitigating the adverse effects of urea on agronomic production and the environment.

Section 8 discusses the potential effects on the biological and chemical sustainability of the soil resource.

Section 9 discusses the potential use of models, including estimates of the effect of using urea on UK ammonia emissions.

Section 10 highlights the implications of the review for new research and other studies.

Acknowledgements

The conclusions and recommendations contained in this report have been considered by the contractor organisations collaborating in the Defra NT2601 and NT2602 research projects (as shown below), and represent a concensus agreement of these organisations. The willing help provided the FMA, and representatives of Hydro Agri (UK) Ltd., Kemira Growhow (UK) Ltd. and Terra Nitrogen (UK) Ltd., is gratefully acknowledged.

Contractor organisations:-

• ADAS

• Edinburgh University (School of Geosciences)

• Horticulture Research International (HRI)

• Institute of Grassland and Environmental Research (IGER)

• Queen’s University, Belfast (Dept. of Agricultural and Environmental Sciences - QuB)

• Rothamsted Research

• SAC

• Silsoe Research Institute (SRI)

Characteristics of urea

(Lead author: Peter Dampney, ADAS)

This section summarises the supply, characteristics and current use of urea-based fertilisers. More details are given in the NT2601 reports ‘Nitrogen fertilising materials’ and ‘Production and use of nitrogen fertilisers’.

1 World consumption and production

Urea is the predominant source of fertiliser nitrogen used in agriculture throughout the world (50% of total N use). In the last 30 years, there has been an approximate 2-fold increase in the global productive capacity of urea compared to little change for other N-containing fertilisers. There is no production capacity for urea in the UK and all current supplies are imported from within or outside Europe.

In the EU-15, the predominant source of N is ammonium nitrate (AN) or calcium ammonium nitrate (CAN); together they represent over 40% of the consumption of N-containing fertilisers. Solid urea represents only 13%, and UAN 11%, of the total consumption of N-containing fertilisers in EU-15; France, Germany, Italy, Spain and the UK are significant users of urea-based fertilisers.

2 Chemical and physical properties

Urea (CO (NH2)2) contains 46% N and is produced by reacting ammonia with carbon dioxide. The molten urea is solidified by granulation or prilling, and the final product may be coated with an anti-caking agent. Urea is primarily used as a solid straight N fertiliser (46% N) or in the production of urea ammonium nitrate (UAN) solution fertiliser (28-30% N w/w). Because urea is very hygroscopic (absorbs water), its use as a raw material in the production of compound fertilisers is more restricted than for AN or CAN.

Urea can be manufactured as prills or granules. Around 30% of world urea production is as granules. Urea has a lower bulk density than AN (prills 0.73kg/l; granules 0.77kg/l; AN 1.00kg/l) ) which makes accurate spreading by spining disc more difficult. A small particle size, as usually found with urea prills (e.g. 2mm diameter), will aggravate this problem so that urea prills cannot usually be spread satisfactorily by spinning disc spreaders used in 24m wide tramline systems. Urea granules have a larger particle size, some with 3.5mm diameter, which should improve the spreading accuracy.

3 Current use on UK farms

AN is the main source of N-containing fertilisers used on UK farms, either as straight AN or AN used in the production of compound fertilisers. Overall, the use of solid urea represents 9% (c.100,000t N) of the total use of fertiliser N (1.1 million tonnes). Most (95%) of the solid urea used is applied as topdressings to winter cereals (63%), oilseed rape (16%) or grass (17%). Very little urea (2% of total) is used on spring cereals and virtually none on potatoes, sugar beet or horticultural crops - this may reflect concerns about possible crop phytotoxicities from use of urea to these crops. Nitrogen applied as UAN solution (50% AN:50% urea) represents a further c.10% of the total N use. A high proportion of the potato area (30%) receives N in fluid form as straight UAN or compound fluid N materials.

Nearly 80% of solid urea and UAN solution is applied in February, March and April when soil and weather conditions are more likely to be cool and moist. Only 12% of urea-N is applied in the warm and dry months of May, June, July and August suggesting that farmers may perceive a poor efficiency from urea applied at this time.

4 Conclusions

1. Urea is the predominant source of fertiliser nitrogen used in agriculture throughout the world (50% of total N use). However, in the EU-15, the predominant source (40%) of N is ammonium nitrate (AN) or calcium ammonium nitrate (CAN), with AN used predominantly on UK farms. Solid urea represents only 9% and urea ammonium nitrate (UAN) solution 10% of the total UK consumption of N in manufactured fertilisers. There is no production capacity for urea in the UK. All current supplies are imported from within or outside Europe. A switch from AN to urea would require a major restructuring of UK agriculture.

2. Urea is primarily used as a solid straight N fertiliser (46% N) or in the production of UAN solution (28-30% N w/w). Urea can be manufactured as prills or granules. However, because urea is very hygroscopic (absorbs water), its use as a raw material in the production of compound fertilisers is much less flexible and more limited than for AN or CAN.

3. Urea has a lower bulk density (0.7-0.8kg/l) than UK manufactured AN (0.99kg/l) which makes accurate spreading by spining disc more difficult. A small particle size, as usually found with urea prills, aggravates this problem. Urea granules have a larger particle size (up to 3.5mm diameter), which should improve the spreading accuracy.

5. Most (95%) of the solid urea used in the UK (100,000t) is applied as a topdressing to winter cereals (63%), oilseed rape (16%) or grass (17%), largely in the February to April period. Very little urea (2% of total) is used on spring cereals and virtually none on potatoes, sugar beet or horticultural crops. Only 12% of urea-N is applied in the warm and dry months of May to August when the risk of ammonia emissions will be highest.

Behaviour and fate of urea-N applied to soil

(Lead author: Keith Goulding, Rothamsted Research)

When urea is applied to soil, it enters the soil nitrogen cycle and, like other N fertilisers, becomes subject to various biological and physicochemical processes. The following biological processes affect urea (Myers, 1974; Kissel et al., 1985):

1. Urea hydrolysis. CO(NH2)2 → NH4+

2. Nitrification NH4+ → NO2- → NO3-

3. Denitrification NO3- → NO2- → NO → N2O → N2

4. Ammonification-immobilisation NO3- → NH4+ → organic N

All biological processes are mediated by the micro-organisms in the soil. Thus the amount, diversity and activity of the soil microbial biomass (SMB) is of great importance in determining the reaction rates of these processes. There is as yet, however, no clear understanding of the link between microbial function and process in soils. Most UK soils, other than those contaminated by industrial works or mining, appear to sustain a SMB that can carry out all necessary functions and processes. Thus we would not expect any of the above processes to be inhibited by anything other than the normal controls such as temperature and moisture.

In addition to the biological processes, N fertilisers are affected by the following physicochemical processes (Myers, 1974; Kissel et al., 1985):

5. Diffusion

6. Convection or mass flow

7. Sorption-desorption

Diffusion and convection transport the urea or its by-products through the soil to the plant. Sorption-desorption are especially important for NH4+, which can be fixed by soil clays particles thus preventing rapid nitrification and N loss to surface waters. Following hydrolysis, urea-N can be lost to the atmosphere through ammonia volatilisation, or urea and its by-products may be washed out of the soil, should there be sufficient rainfall to generate runoff or drainflow. The impacts of these processes on the efficiency of use of urea fertiliser, and their interaction with growing crop plants, are explained below.

1 Fate of urea applied to soil

When urea is applied to soils the processes above begin to act. The first step is hydrolysis to NH4+ (process 1 above) which is controlled by the activity of the urease enzyme and urea concentration (Figure 4.1), as well as temperature (Figure 4.2) and soil moisture content (Myers, 1974). Ureolytic micro-organisms that produce urease are ubiquitous in soil, on vegetation and in surface litter. Urease activity is related to soil organic matter content and pH, with an optimum rate at pH 7-8.5 (Kissel et al., 1985).

On soils with a pH of c.6.3 or greater, urea is hydrolysed mainly to ammonium bicarbonate:

CO(NH2)2 + H+ + 2 H2O ( 2NH4+ + HCO3-

whereas on more acid soils, the hydrolysis occurs as follows:

CO(NH2)2 + 2H+ + 2 H2O ( 2NH4+ + H2O + CO2

Thomlinson (1970) quotes work showing that, in a silt loam at 24% moisture content, rates of hydrolysis for 224 kg N/ha applied urea were c. 20 kg N/ha/day at 4OC rising to 105 kg N/ha/day at 20OC, and O’Toole & Morgan (1988) calculated mean rates of 510, 340 and 160 kg urea-N/ha/day at 24, 16 and 8OC (N applied = 500 mg/kg) from a laboratory experiment. Thus urea hydrolysis is generally rapid and unlikely to be a major factor in the efficiency of use of urea in most UK arable soils in which the pH is maintained at near neutral values. It could be slowed in grassland for which the optimum recommended pH is 6.0 (MAFF, 2000) or greatly reduced in grassland that has been allowed to become acid. Skinner and Todd (1998) found from Representative Soil Sampling Scheme data that the only soils with declining pH were under permanent grassland. These had declined from an average pH of 5.7 in 1970 to 5.4 in 1992, sufficient to affect urea hydrolysis.

The rate of urea hydrolysis could also be slowed in arable soils depleted of organic matter, or slowed or stopped in very dry, very wet or very cold weather. It is unlikely to be too hot for urea hydrolysis. However, Powlson et al. (1988) found that denitrifiers were adapted to local environments, with maximum rates at c.10oC in the UK and c.20oC in sub-tropical areas of Australia; similar adaptation could occur to other soil micro-organisms controlling N cycling.

[pic]

Figure 4.1. Rate of urea hydrolysis vs urea concentration.

[pic]

Figure 4.2. Rate of urea hydrolysis vs soil temperature (Note log scale for hydrolysis).

Since unhydrolysed urea is soluble in water, any urea that is not hydrolysed is at risk of leaching into surface or ground waters. There is little research information on direct leaching of urea but, since hydrolysis of urea is usually very fast, leaching directly into waters is generally considered unlikely except in high risk situations such as intense rainfall on very sandy or cracking clay soils immediately after urea application. Further research is needed on this issue (also see section 6.4).

The effect of pH on the ammonia/ammonium equilibrium is shown in Figure 4.3 (Thomlinson, 1970). The ammonium produced is in equilibrium with free ammonia in the soil solution, which is in turn in equilibrium with atmospheric ammonia. In solution, the proportion of ammoniacal-N present as ammonia increases with increasing pH; thus there is greater likelihood of ammonia emission at a higher pH. This has given rise to the idea that ammonia loss is more likely from high pH soils such as calcareous soils. However, this is not necessarily true because urea hydrolysis by itself will result in a temporary rise in pH up to about pH 9, which tends to override the bulk soil pH. The small volumes of high pH around individual urea particles has been termed ‘alkaline microsites’ (Tomlinson, 1970).

[pic]

Figure 4.3. The percentage of total N present as ammonia (■) or ammonium ions (NH4+; ▲)

at various soil pH values.

Once urea has hydrolysed, its reaction products are subject to the many competing and interacting processes listed above. The first breakdown product, NH4+, can be taken up by plants, fixed on clays, immobilised by the SMB, nitrified or volatilised as NH3 (Myers, 1974; Kissel et al., 1985). Ammonia emission is considered in detail in section 6.1. Although ammonia emissions from surface-applied urea can be considerable, much of this ammonia could be re-absorbed by any crop plants present (Hutchinson et al., 1972), or washed out of the air and returned in rainfall (Viets, 1971). Regarding re-absorption, Sommer et al. (1993) found that only a small amount (c. 2-3%) of volatilised ammonia was absorbed by cereal crop canopies. However Ping et al. (2000) found that, following a 100 kg N/ha topdressed application of urea to spring wheat, 13% was volatilised over 7 days and, of this, up to 15% was absorbed by the crop canopy. It might be expected that the amount of N absorbed by crop foliage will vary depending on various factors such as canopy cover and, probably more important, air flow. Absorption would be more likely in still rather than windy conditions. Regarding return in rainfall, Yaalon (1964) calculated that an equivalent amount of ammonia to that volatilised from soils in Israel each year was returned in rainfall.

Nitrification (process 2, above) is sensitive to soil pH, temperature and moisture. Nitrification can occur between soil pH 5-10, but the optimum pH is in the range 6.0-8.0 (Paul & Clark, 1989); little nitrification occurs below pH 5 or above pH 8, with the exact value depending on soil texture and the character of the SMB (Boswell et al., 1985). Since nitrification produces protons and thus local acidification, the rate of nitrification can decrease with time, especially in poorly buffered, unlimed sandy soils.

Rates of nitrification and ammonification-immobilisation increase with soil moisture up to a maximum water potential of about –0.1 bar. At water contents above this, oxygen limitation begins to become important and the rates of these aerobic processes decline rapidly, e.g. by 50% at 2% oxygen content (Boswell et al., 1985), while that of denitrification increases.

The first product of nitrification of NH4+ is nitrite (NO2-). This is generally short-lived and rapidly oxidised to nitrate (NO3-). However, ammonium oxidisers Nitrosomonas have an optimum pH of 7-9 but nitrite oxidisers Nitrobacter have an optimum pH of 6.5-7.5. Thus, above pH 7 the conversion of NO2- to NO3- is inhibited and NO2- can accumulate. Thus the local increase in pH around hydrolysing urea can result in the presence of significant concentrations of NO2- that can be toxic. The NO2- could also be leached, causing pollution of waters, denitrified as described in process 3, or nitrified as in process 2.

Both nitrite and nitrate are at risk of denitrification (process 3, above). This is the anaerobic, strictly anoxic, reduction of oxidised forms on N through to the gases nitrous oxide (N2O) and nitrogen (N2). The process causes an economic loss to the farmer by reducing the efficiency of N use by the crop and, if denitrification stops at N2O, causes environmental pollution because N2O is a potent greenhouse gas. Denitrification and N2O production are discussed in detail in section 6.2.

Both NH4+ and NO3- can be immobilised, i.e. taken up into the bodies of the SMB (process 4, above). Recent research has shown this to be a very rapid process (Murphy et al., 2003); the SMB competes very effectively with plant roots for inorganic N in the soil solution. The NH4+ form is preferred to the NO3- form by the SMB, which could be a cause of the reduced effectiveness of urea compared to nitrate forms of N (see section 5)

However, immobilised N is not lost but is likely to be made available again when the SMB dies and is mineralised and nitrified, i.e. converted to NH4+ and NO3-. Subsequent release is probably slow, so immoblised N from urea may be effectively lost to the current year’s crop. The understanding, modelling and control of this mineralisation-immobilisation turnover (MIT) have been at the centre of much of Defra's research in the last 10 years.

2 Conclusions

1. When urea is applied to soil, it enters the soil nitrogen cycle and becomes subject to hydrolysis, nitrification, denitrification and ammonification-immobilisation processes. Following hydrolysis, urea-N can be lost to the atmosphere through ammonia volatilisation or urea or its decomposition products may be washed out of the soil should there be sufficient rainfall to generate runoff or drainflow. Diffusion and convection transport urea or its by-products through the soil to the plant. Sorption-desorption processes are especially important for ammonium N (NH4+-N), which can be fixed by soil clay particles and can potentially reduce ammonia volatilisation losses, delay nitrification and subsequent N loss to water systems.

2. Hydrolysis of urea to NH4+-N is controlled by the activity of urease enzymes, urea concentrations, soil temperature and soil moisture contents. Ureolytic micro-organisms that produce urease are ubiquitous in soil, on vegetation and in surface litter. Rates of hydrolysis are generally rapid in most UK soils and are unlikely to affect the efficiency of urea use by crops. However, hydrolysis could be reduced in arable soils depleted of organic matter, or in very dry, very wet or very cold weather. Hydrolysis could also be slowed in grassland soils that have been allowed to become acid (optimum range for hydrolysis: pH 7-8.5).

3. Urea is soluble in water and therefore at risk of leaching into surface or ground waters. There is little information on direct leaching of urea, but hydrolysis is usually considered to be so fast as to make leaching unlikely, except in high risk situations such as intense rainfall on ‘wet’ soils soon after application, that results in surface runoff or drainflow.

4. The NH4+ produced when urea hydrolyses is in equilibrium with free ammonia (NH3) in the soil solution, which is in turn in equilibrium with atmospheric NH3. The proportion of NH4+ present as NH3 increases with increasing pH, thus the risk of a NH3 loss increases with pH. Urea hydrolysis causes a temporary rise in pH (up to pH 9) in the environment surrounding the applied urea-N which exacerbates the problem. Ammonia volatilisation is the major N loss process responsible for the lower agronomic efficiency of urea compared to AN.

5. The first product of nitrification of NH4+ is nitrite (NO2-). This is generally short-lived and rapidly oxidised to nitrate (NO3-). However, above pH 7 the conversion of NO2- to NO3- is inhibited and NO2- can accumulate. Thus the local increase in pH around hydrolysing urea can result in the presence of significant concentrations of NO2- that can be toxic. The NO2- could also be leached (causing pollution of waters) denitrified or nitrified. Both NO2- and NO3- are at risk of denitrification.

6. Both NH4+ and NO3- can be immobilised. The NH4+ form is preferred to the NO3- form by the soil microbial biomass (SMB), which could be a cause of the reduced effectiveness of urea compared to nitrate based fertilisers. Subsequent release is probably slow, so immobilised N from urea may be effectively lost to the current year’s crop.

Agronomic effectiveness

This section reviews information on the agronomic effectiveness of solid urea and UAN solution fertilisers, without the use of urease inhibitors, compared with ammonium nitrate (AN), calcium ammonium nitrate (CAN), calcium nitrate (CN) and ammonium sulphate (AS). Effects on the main arable, grassland and horticultural crops grown in the UK are considered. Information has been sourced from both published literature and unpublished information.

The use of urease inhibitors is reviewed in section 7.2.

1 Arable cropping

(Lead author:- Tony Lloyd, ADAS)

1 Seedbed N and combine drilling

Nitrogen fertiliser is recommended for seedbed application for certain spring sown or planted crops. Currently, very little urea is used for seedbed applications (less than 2% of total urea-N use, see NT2601 report ‘Production and use of nitrogen fertilisers’), but it is important to know if seedbed applications of urea might adversely affect crop germination and early growth. The maximum recommended amounts of N for seedbed application are given in Defra (2000) as shown below:

• For later drilled spring wheat crops, up to 180 kg N/ha or for light sandy soils up to 70 kg N/ha.

• For spring rape, up to 120 kg N/ha or for light sandy soils 80 kg/ha.

• For potatoes up to 270 kg N/ha or, for light sands and shallow soils, only two thirds of this amount. DAP is commonly used as a seedbed dressing for potatoes.

• For sugar beet, a maximum of 40 kg N/ha.

The risk of adverse effects is greatest when the fertiliser is combine drilled with the seed. This is largely thought to be due to ammonia toxicity in the vicinity of the germinating seed. Buiret toxicity used to be of concern for seedbed applications, but under the Fertiliser Regulations (1990) urea fertiliser must now contain no more than 1.2% buiret (formed as a by-product during urea manufacture) which, at this concentration, is considered to have no adverse effect on crop growth.

At 4 trials, Widdowson & Penny (1960) compared the effect of 30-95 kg N/ha of combine drilled urea and AS on the yield of spring barley. Urea supplying 60 kg N/ha checked early growth while at 95 kg N/ha growth was severely checked and some plants died.

Widdowson et al. (1964) also compared 45 or 90 kgN/ha of urea and AS applied to spring barley either combine drilled in contact with the seed, or placed one inch to the side of the seed. Combine drilling urea (even at the lower N rate) killed some plants and reduced yield whereas the same effect did not occur with AS. Placing urea one inch from the seed reduced the adverse growth effects noted with combine drilled urea.

Devine & Holmes (1963a) investigated the effect of combine drilling urea and other fertilisers on spring barley at 21 trials. Combine drilling urea at 50 kg N/ha had no effect on early growth, but at 80 or 100 kg N/ha there was a serious delay to brairding and reduced plant population, with resultant lower yields.

Combine drilling is now rarely practised in the UK and is not recommended, so a switch to urea-based fertilisers should not cause concern, unless this practice increased. Comparisons of fertilisers for non-combine drilled seedbed applications are given in the relevant crop sections.

2 Topdressed urea

Many early trials in the 1960s and 1970s have been discussed in reviews (Gasser, 1964; Tomlinson, 1970). Tomlinson (1970) concluded that urea had variable effectiveness compared to other nitrogen fertilisers but that significant differences only occurred in a minority of cases; where they did occur, urea was usually the least effective fertiliser. He considered that, although environmental conditions had not been reported in sufficient detail to help explain any differences, the main factor affecting the efficiency of urea was ammonia loss (i.e. reduced N availability). He added that urea tended to be more effective when cultivated into the soil than when broadcast on the soil surface, which also suggests ammonia volatilisation to be responsible. Some authors also suggest direct toxicity as a mechanism for reduced effectiveness of urea (Court et al. 1964) due to a reduction in rhizosphere pH, induction of cation deficiencies or plant water stress in carbohydrate metabolism associated with the detoxification of ammonium-N within the plant (Haynes, 1986). However, there is no evidence to suggest that this occurs for UK arable crops (Tomlinson, 1970).

The greatest potential for loss of ammonia is when urea is topdressed (broadcast) on the soil surface. However this will be strongly dependent on soil moisture conditions and rainfall at and following the time of application.

Provided the soil is dry, very little hydrolysis of the urea is likely to occur. Terman (1979) reported that urea applied to air-dry soil does not hydrolyse and suggested that, even at high humidity, urea on a dry soil will not take up enough moisture to support quantitative hydrolysis. The inference from this is that urease enzyme activity is not observed in the highly concentrated solutions formed by deliquescence of the solid. On the other hand, if a large amount of rain falls after application, then the urea will be washed into the soil and ammonia emissions will be considerably reduced; this can have the same effect as physical incorporation of the fertiliser.

Between these two extremes, there will be situations where urea will remain on the surface of a moist soil and be at risk for ammonia loss, for example:

1. Where the soil is moist before application (either from previous rain or dew) and there is little rain for a few days following application.

2. Where the soil is dry before application and then there is sufficient rain to moisten the surface but not to wash the urea into the soil. In this circumstance, hydrolysis will occur and, as there is no soil cover to adsorb the ammonia, considerable loss to the atmosphere will occur.

In both cases, the amount of loss will be affected by wind speed across the soil surface.

Several workers have commented on the effect of weather on the effectiveness of topdressed urea. Lloyd et al. (1997), studying the effect of a single application of urea to winter cereals at growth stage (GS) 31, found that grain N offtake increased with increasing rainfall on the day of application but not on subsequent rainfall; on chalk soils grain N offtake increased with increasing rainfall up to the fifth day after application. Gately (1994) found that the drier the weather around the time of N application, the poorer was the performance of urea relative to CAN for topdressing winter wheat. Sanderson & White (1987) found that, for potatoes, 80% of yield variation between the use of urea and AN could be explained by the temperature and accumulated rainfall in the week prior to and following application, as well as the temperature two weeks after planting. Fox et al. (1996) measured ammonia losses from urea topdressed to maize crops; c. 30% of applied N was lost as ammonia and this relatively high amount was attributed to the relatively rain-free period for at least 6 days after application in each year. They quoted five other reports suggesting that a rainfall of 5-10 mm within 6 days of application was sufficient to significantly reduce ammonia volatilisation.

The following sections reviews trial results carried out on arable crops. The results are summarised in Appendix 1.

1 Winter cereals

Devine & Holmes (1963b) compared spring topdressing of AN and urea for winter wheat at 17 trials during 1958-61. Two low rates of N (28 and 50 kg/ha) were tested with nil N controls and the response (yield increase above nil N) was adjusted to 39 kg/ha (assuming an exponential response curve). Over all the trials, the response to urea was 96% of that to AN, but the difference was not statistically significant; there was no effect of soil pH or texture. However, three of the trials showed a significantly (P 250-260 milliequivalent/kg irrespective of the season of application, whereas there was a very high risk of NH3 volatilisation on soils with a CEC < 160 me/kg. The greatest risk of NH3 emissions from urea is therefore likely to occur on coarse-textured soils, with a low organic matter content and where there are low amounts of crop cover. High emission rates are also generally associated with drying soils, when urea is applied to wet (near field capacity) surface soils and followed by several days of little (0.1); however, a more detailed examination of the data revealed that the differences were principally on the wetter grassland site in cool spring conditions, and to some extent in early summer. Again, the most likely explanation was that the denitrification pathway to N2O (from nitrate) was favoured over the nitrification pathway under these conditions.

Table 6.4. Total seasonal emissions of N2O from grassland fertilised with either AN or urea

(data from Dobbie and Smith, 2003; including data from Clayton et al., 1997).

| | |Seasonal N2O flux (kg N2O-N ha-1) |

| | |Mean |Standard error |

|Season |Region |AN |Urea |AN |Urea |

|92-93 |Central Scotland |1.5 |3.0 |0.2 |0.2 |

|93-94 |Central Scotland |4.2 |5.2 |0.3 |0.6 |

|94-95 |Central Scotland |0.8 | | | |

|96-97 |Central Scotland |1.9 |1.4 |0.3 |0.7 |

|97-98 |Central Scotland |13.9 |7.0 |2.8 |0.8 |

|98-99 |Central Scotland |22.6 |20.2 |2.6 |2.7 |

|99-00 |Central Scotland |11.6 |3.9 |3.4 |2.3 |

|00-01 |Central Scotland |16.0 |9.1 |4.1 |3.9 |

| | | | | | |

|96 |South-East Scotland |3 |2 |0.2 |0.3 |

|97 |South-East Scotland |7.9 |7.3 |1.4 |1 |

|98 |South-East Scotland |17.2 |19.3 |1.5 |1.9 |

| | | | | | |

|96 |South-West Scotland |4.5 |6.5 |0.5 |1.2 |

|97 |South-West Scotland |4 |2.6 |0.4 |0.1 |

|98 |South-West Scotland |10.6 |7.9 |1.3 |0.6 |

In The Netherlands, evidence has been obtained that supports these findings. Velthof et al. (1997) described experiments comparing emissions from AS, CN, urea and CAN, on grass growing on clay and sandy soils of differing drainage classes. Rising water tables after heavy rain, following fertiliser application, gave much higher fluxes from CN and CAN than from AS or urea, at soil temperatures below 10(C.

In contrast with the results for grassland in Scotland, those comparing AN and urea on arable crops in Scotland (Dobbie and Smith, 2003a) showed no difference in total seasonal N2O emissions between these two fertiliser types (Table 6.5).

Table 6.5. Total seasonal emissions of N2O from arable crops fertilised with ammonium nitrate or urea (data from Dobbie and Smith, 2003).

| | |Seasonal N2O flux (kg N2O-N ha-1) |

| | |Mean |Standard error |

|Season |Crop |AN |Urea |AN |Urea |

|96-97 |Winter wheat |0.6 |0.7 |0.2 |0.2 |

|97-98 |Winter wheat |0.9 |1.0 |0.3 |0.3 |

|96-97 |Potatoes |3.5 |3.7 |1.1 |1.3 |

|97-98 |Potatoes |5.0 |4.8 |1.0 |1.0 |

|98-99 |Oilseed rape |1.6 |1.4 |0.4 |0.3 |

Granli and Bøckman (1994) compiled data on N2O losses from different fertilisers, based on reviews by five authors. These are shown in Table 6.6. Granli and Bøckman concluded that the N2O loss (i.e. the Emission Factor, in current parlance) was usually in the range of about 0.1-2%, with no single mineral fertiliser (with the possible exception of anhydrous ammonia) giving more emission than the others. They also concluded that some situations could be associated with high emissions:

1) application of urea/ammonium compounds under conditions favouring N2O production by both nitrification and denitrification, e.g. in moist but well-aerated soil;

2) use of nitrate fertilisers where denitrification is favoured, e.g. on clay soils in wet climates;

3) injection of anhydrous (but not aqueous) ammonia.

Table 6.6. Median N2O yields (%) for different fertiliser types (with ranges) from reviews by 5 authors.

|N form |A |B |C |D |E |

|Nitrate |0-07 |0.04 |0.07/0.04 |0.05 |0.04 |

| |(0.01-1.8) |(0.001-1.3) |(0.001-0.5) | | |

|Ammonium |0.12 |0.15 |0.15 |0.11 |0.173 |

| |(0.03-0.9) |(0.03-1.5) |(0.05-1.8) | | |

|Urea |0.11 |0.1 |0.2/0.61 |0.52 | |

| |(0.07-0.2) |(0.01-0.6) |(0.1-2.1) | | |

|AN |0.1/0.4 |0.7 | | | |

| |(0.04-1.7) |(0.3-1.6) | | | |

|Ammonia |1.2/1.4 |0.1 | | | |

| |(0.9-6.8) |(0.01-2.05) | | | |

Authors: A: Eichner (1990); B: Bouwman (1990); C: Keller et al. (1988);; Galbally (1985); Bolle et al. (1986). 1NH4NO3 included; 2NH3 included; 3urea included.

In their review, Harrison and Webb (2001) proposed a scheme for assessing the relative emissions of N2O from different fertilisers (Table 6.7). In most cases, N2O emissions will be greater from NO3-based fertilisers compared with NH4-based fertilisers, this difference increasing with moisture content (assuming denitrification to be the dominant source of N2O). For example, Leick & Engels (2001) measured higher emissions from CN and CAN compared with AS, due to an increase in the soil NO3-N content.

Table 6.7. Proposed scheme for assessing the relative emissions of N2O from different fertilisers (from Harrison & Webb, 2001).

|Moisture |Relative emission for N forms |Relative emission from urea |

|dry |low |nitrate ( ammonium |urea ( ammonium |Rate of urea hydrolysis limited |

|wet |high |nitrate > ammonium |urea >> ammonium |Rate of urea hydrolysis increases |

| | | | |with temperature |

|v. wet |high |nitrate >> ammonium |urea ( ammonium |High pH associated with hydrolysis|

| | | | |dispersed by moisture |

Urea behaves differently from ammonium forms of N. During hydrolysis, the rise in soil pH in the vicinity of the fertiliser granule can inhibit the oxidation of nitrite (NO2) to nitrate (NO3) by nitrobacter (Gould et al., 1986). The accumulation of nitrite in soils can then lead to increased N2O (and nitric oxide, NO) losses (see below). Therefore, conditions which lead to rapid urea hydrolysis (i.e. warm and wet soils) can lead to N2O emissions which are much greater than those from NH4-based fertilisers and may even exceed those from NO3-based fertilisers.

In this context, the grain size of the urea granules could have an important effect on N2O emissions. In a laboratory study, Tenuta and Beauchamp (2000) found that the production of N2O (and the concentration of NO2-) in soil increased steadily as the granule size increased from a powder to prills and to larger granules. A high concentration of prills produced a similar but greater effect than large granules. The proportion of urea transformed into N2O increased with granule size but did not exceed 1.24%, but a high concentration of urea prills resulted in 2.80% being transformed into N2O.

In comparison to urea, AS applications tend to decrease soil pH and therefore inhibit nitrification, so that N2O losses tend to be least from this fertiliser form.

In Guelph, Canada, continuous measurements of N2O (and NO and NO2 fluxes, see below) were made for three growing seasons by micrometeorological techniques over turf grass fertilised with AN, urea and slow release urea (Maggiotto et al., 2000). N2O fluxes were found to be dependent on weather conditions and soil moisture at the time of fertiliser application. Largest fluxes of N2O were observed from AN-fertilised grass in two out of three seasons. In the third season, N2O fluxes from AN were smaller than from the two forms of urea. However, averaged over the three seasons, N2O emissions from AN were three times larger than from the two urea treatments. This difference was almost entirely the result of an extreme peak in N2O flux of two to three days duration following AN application when the soil was particularly wet. While emissions from slow release urea were initially smaller, they increased above those from conventional urea in the third season. This might be related to a increase in mineral N concentrations resulting from longer term application of slow release urea, a trend also observed by Smith and Dobbie (2002).

1 Chemically amended fertiliser materials

The use of nitrification inhibitors has been shown to reduce N2O emissions from NH4-based fertilisers. Earlier work with Nitrapyrin (2-chloro-6-(trichloromethyl)-pyridine, also called N-serve) has been reviewed by Granli and Bøckman (1994), who quoted several authors as showing a reduction in N2O emission with this compound. Work since the mid-1990s, mainly with dicyandiamide (DCD), has also shown reduced emissions (McTaggart et al., 1997; Velthof et al., 1997; Leick and Engels, 2001). Under cool Scottish conditions on a relatively wet grassland soil, the DCD (as 'Didin Fluid'), reduced N2O emissions from urea by 72% for the period from June 1999-June 2000 and by 46% from June 2000 to June 2001 (Dobbie and Smith, 2001). This confirms earlier work in Scotland by McTaggart et al. (1994, 1997) and in The Netherlands by Velthof et al. (1997). McTaggart et al. (1994) also showed a 50% reduction in emission from AS amended with DCD. In contrast, neither the urease inhibitor ('Agrotain', N-(n-butyl) thiophosphoric triamide (nBTPT)), as a sole additive to urea, or in combination with the nitrification inhibitor DCD, nor controlled release urea, reduced N2O emissions significantly below those from conventional urea under these conditions (Dobbie and Smith, 2003b).

2 Emission factors

In a review of over 100 experiments, Eichner (1990) suggested the following emission factors for the loss of N2O from N fertilisers: anhydrous NH3-N 2.7% (range 0.86-6.84); AN 0.44% (range 0.04-1.71); ammonium-N 0.25% (range 0.02-0.90), urea 0.11% (range 0.07-0.18); nitrate-N 0.07% (range 0.001-0.5). However, these emission factors were later considered to be unrepresentative, due to the high proportion of studies conducted on fallow land in the absence of a competing crop.

The IPCC guidelines for greenhouse gas inventories suggest the use of a single N2O emission factor for fertiliser applications, with no allowance made for different fertiliser types because (a) they were based on very limited data, and (b) it was thought that fertiliser type was likely to have little impact on the total emission (IPCC, 1997). A factor of 1.25% ( 1% was adopted, based on the work of Bouwman (1994, 1996). This factor was used in the UK inventory of N2O emissions from farmed livestock (Chadwick et al., 1999) and has been recommended in the EU Emissions Inventory Guidebook (Anon., 2003).

More recently, it has been recognised that measured emission factors for N2O are log-normally distributed, with an uncertainty range from one-fifth of the mean to 5 times the mean. This translates into a range from 0.25% to c. 6% of the N applied, assuming no change in the IPCC default average value, rather than the 0.25-2.25% implied by the formula of 1.25( 1% (Mosier and Kroeze, 1999; Smith et al., 2002).

This is broadly borne out by the research reviewed by Eichner (1990) and Harrison & Webb (2001): N2O emission factors ranged from ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download