NT2605 Final Report WP4 Nitrogen losses to surface and ground waters

______

Component report for Defra Project NT2605 (CSA 6579)

WP4 Nitrogen losses to surface and ground waters

Lead Authors

Andrew Macdonald and Keith Goulding (Rothamsted),

Anne Bhogal, Fiona Nicholson, Brian Chambers

and Lizzie Sagoo (ADAS)

and

Liz Dixon and David Hatch (IGER)

February 2006


Contents

1.  EXECUTIVE SUMMARY………………………………………………………………………………….4

2.  INTRODUCTION…………………………………………………………………………………………...5

3.  EXPERIMENTAL DESIGN, TREATMENTS AND METHODS………………………………...……..9

4.  RESULTS AND DISCUSSION………………………………………………………………………….15

5.  KEY CONCLUSIONS…………………………………………………………………………………….29

6.  REFERENCES……………………………………………………………………………………………32

7.  APPENDICES…………………………………………………………………………………………….34


Abbreviations

Ag / Agrotain (trade name) urease inhibitor (active ingredient is nBTPT)
AnA / Anhydrous ammonia
AN / Ammonium nitrate
AS / Ammonium sulphate
AU-N / Ammonium-N + urea-N
nBTPT / N-(n-butyl)-thiophosphoric triamide urease inhibitor
CAN / Calcium ammonium nitrate
CEC / Cation exchange capacity
DCD / Dicyandiamide nitrification inhibitor
DNDC / DeNitrification - DeComposition
EF / Emission factor
EU / Edinburgh University
FFD / Freshwater Fish Directive
GC / Gas chromatography
HRI / Horticulture Research International
IGER / Institute of Grassland and Environmental Research
IPCC / Intergovernmental Panel on Climate Change
N / Nitrogen
NH3 / Ammonia
NH4 / Ammonium
N2O / Nitrous oxide
NO2 / Nitrite
NO3 / Nitrate
NVZ / Nitrate Vulnerable Zone
QuB / Queens University, Belfast
RR / Rothamsted Research
U / Urea
U+Ag1000c / Urea granules with 1000ppm of nBTPT urease inhibitor (coated onto granule)
U+Ag500m / Urea granules with 500ppm of nBTPT urease inhibitor (in the melt)
UAN / Urea ammonium nitrate solution
UAN+Ag500 / Urea ammonium nitrate solution with nBTPT urease inhibitor (500ppm of urea-N)
UAS / Urea ammonium sulphate
UKAEI / UK Ammonia Emissions Inventory
WFPS / (Soil) water-filled pore space
ppmv / gas concentration in parts per million by volume (equivalent to mmol mol-1)


1. Executive Summary

·  Nitrogen losses following the application of N fertilisers to ‘wet’ soils in spring were dependent upon the timing and amount of rainfall following application, and water movement pathways through the free-draining sandy soils (to ground waters) and drained/undrained clay soils (to surface waters).

·  Total N losses ranged between 0.7-23% of the total N applied and represented a ‘worst-case’ situation in terms of likely N losses, as irrigation water was applied at a number of the sites to generate water movement soon after fertiliser spreading. There were no consistent differences in total N losses between AN, urea and urea+Agrotain (after taking into account ammonia volatilisation losses to air from urea) at all eight study sites. Hence, a switch from AN to urea or urea+Agrotain is not likely to substantially change total N losses to surface and ground water systems.

·  Agriculture is estimated to be responsible for 5-10% of Freshwater Fish Directive (FFD) non-compliance cases, with c.50% of these cases estimated to be from point-sources and c.50% from diffuse sources (mainly from poor manure management). Ammonium-N concentrations in waterflows from the six surface water sites commonly exceeded the FFD mandatory level of 0.78 mg/l NH4-N. Peak NH4-N concentrations in drainflow from the drained study sites were up to 42 mg/l, and in surface runoff/subsurface flow from the undrained study sites were >80 mg/l. There were no consistent differences in NH4-N concentrations and loads between the N fertiliser materials at the four drained sites and two of the undrained sites. However, NH4-N loads in surface runoff/subsurface flow from the two Rowden undrained lysimeter studies were c.6-fold greater from AN than urea. In summary, a switch from AN to urea or urea+Agrotain would not significantly change the contribution of fertiliser N to FFD non-compliance.

·  Peak urea-N concentrations in drainflow from the drained study sites were > 60 mg/l, and in surface runoff/subsurface flow from the undrained study sites were > 100 mg/l.

·  Nitrite concentrations commonly exceeded the FFD guideline level of 9 mg/l NO2-N in waterflows, with no consistent differences in concentrations or loads between the 3 fertiliser types.

·  A switch from AN to urea would change the balance of N loss forms to surface waters. Typically, ammonium + urea-N (AU-N) loads were 2 to 3-fold higher from urea than AN, and NO3-N loads 3 to 6-fold higher from AN than urea. The balance between N loss forms depended upon the degree of urea hydrolysis to NH4-N, and nitrification of NH4-N to NO3-N, in relation to the timing of waterflows. Nitrogen losses from the free draining sandy soils were dominated by the NO3-N form.


2. Introduction

2.1. The NT26 Research Programme

The NT26 research programme was set up by Defra to investigate the nitrogen (N) loss pathways, the environmental and economic impacts, and the response of agricultural and horticultural crops to different forms of fertiliser-N. The NT2605 project was part of a suite of projects in this programme as shown below (Final report submission dates shown in brackets).

NT2601 / Desk study reports on:
·  Nitrogen fertilising materials (June 2003)
·  Production and use of nitrogen fertilisers (August 2003)
NT2602 / Desk study report on:
·  Evaluation of urea-based nitrogen fertilisers (October 2003)
NT2603 / Report of field studies (2002/03 cropping season):
·  The behaviour of some different fertiliser-N materials (March 2004)
NT2604/06 / Facilities construction:
·  Ammonia emissions from nitrogen fertilisers – wind tunnel construction (March 2004)
NT2605 / This project
NT2610 / Report of field studies (led by Silsoe Research Institute):
·  Spreading accuracy of solid urea fertilisers (August 2005)

The following leading UK agri-environment research organisations participated in all the NT26 projects (except NT2610), including the NT2605 project reported here.

·  ADAS UK Ltd

·  Edinburgh University (EU)

·  Warwick HRI (HRI)

·  Institute of Grassland and Environmental Research (IGER), North Wyke

·  Queens University, Belfast (QuB)

·  Rothamsted Research (RR)

·  SAC Commercial Ltd (SAC)

The project was led by Peter Dampney, Principal Research Scientist, ADAS Boxworth Research Centre, Cambridge who was the main point of contact with the Defra NT26 Steering Group.

2.2 The NT26 Project

The NT2601, NT2602 and NT2603 projects provided the basis for the field experimental and other work carried out in NT2605, in cropping seasons 2003/04 and 2004/05. The overall aim of the project was to develop working decision support systems (DSS) to evaluate the agronomic, environmental and economic impacts that would result from changes in the use of different fertiliser-N materials in UK agriculture. More specifically, project work packages (WP) covered the following topic areas:-

WP1a / To investigate crop responses to different fertiliser N forms.
WP1b / To generate robust ammonia emission algorithms and emission factors for predicting the loss of ammonia following application of different fertiliser N forms under a range of crop, soil and environmental conditions. To evaluate the relationship between ammonia loss and crop N use efficiency as a potential basis for revising current national standard nitrogen fertiliser recommendations (Defra, 2000).
WP2 / To generate robust nitrous oxide emission factors for predicting losses following application of different fertiliser N forms under contrasting crop, soil and environmental conditions.
WP3 / To determine the optimum formulation method, addition rate and method of use of urea treated with the urease inhibitor nBTPT (Agrotain), to maximise its ammonia abatement potential and efficiency of N use by crops, whilst minimising any adverse phytotoxic effects.
WP4 / To assess the risk of ammonium-N, nitrite-N or urea-N losses to surface waters and groundwaters following the application of urea-based N fertilisers.
WP5 / To assess the potential for urea or urea+Agrotain to cause phytotoxic effects during establishment, in growing crops, or in marketable produce.
WP6 / To construct a decision support system that will assess the economic impacts of changes in the availability of different forms of N fertiliser on different farm types and UK agriculture.
WP7 / To estimate and evaluate the agronomic, environmental and economic impacts at both farm and national levels that would result following different hypothetical scenarios concerning the availability of N-containing fertilisers to UK farmers.

Reporting of the NT2605 has been structured into a suite of 8 component reports, one for each work package plus an over-arching Executive Summary for the whole project. Each report is self contained with its own Executive Summary, but interacts with data and conclusions from other WPs where appropriate.

2.3. Work Package 4

Very little work had been conducted in the UK linking differences in fertiliser type to nitrogen (N) leaching losses and forms. Indeed, once urea or other ammonium-based fertilisers are transformed to nitrate and start to move into the soil profile, their fate is likely to be very similar to that of nitrate derived from ammonium nitrate (AN) fertiliser. However, if there is heavy rainfall creating surface runoff or drainflow immediately after fertiliser application, or if the transformation of urea or ammonium-based fertiliser to nitrate-N (NO3-N) is delayed, either artificially (using inhibitors) or because of environmental conditions (soil temperature, moisture and pH), then differences between fertiliser products may arise.

Urea is non-ionic and can be susceptible to leaching and runoff (Gould et al., 1986). The potential for urea leaching has been demonstrated in leaching columns under laboratory conditions (Broadbent et al., 1958), where urea was considered to be more susceptible to leaching than ammonium-N (NH4-N), but less than NO3-N. Sherwood and Fanning (1985) measured substantial losses of unhydrolysed urea (c.24% of that applied) in surface runoff following 10 mm of rainfall shortly after application to an impermeable grassland soil. However, apart from these early studies, there was very little information on urea leaching or runoff losses under experimental conditions. Indeed, urea has been measured in tributaries of the Chesapeake Bay, USA (with peak concentrations coinciding with fertiliser use) and in surface waters in South Africa (Glibert, 2004). The use of urease inhibitors (to minimise ammonia emissions) may exacerbate the problem, particularly where there is runoff or drainflow as a result of heavy rainfall following application. Hydrolysis of urea within watercourses is likely to impact on both NH4-N and NO3-N concentrations, and could increase NH4-N concentrations above the guidelines for Salmonid and Cyprinid fish in freshwaters (EC, 1978). Previous work in Defra project NT2603 (Dampney et al., 2004) showed that there was a risk of movement through soils of elevated concentrations of NH4-N, NO3-N, nitrite-N (NO2-N) and urea-N to surface and ground waters in the first few days after fertiliser N applications.

The EU Freshwater Fish Directive (FFD) has set mandatory threshold concentrations for total NH4-N of 0.78 mg/l and guide levels of 0.03 mg/l and 0.16 mg/l for Salmonid and Cyprinid fish, respectively (EC, 1978). For NO2-N, the guide levels are 0.003 mg/l and 0.009 mg/l for Salmonid and Cyprinid fish, respectively. Ammonium-N is usually relatively immobile (until it has nitrified), due to adsorption onto the soil exchange complex. However, concentrations of up to 5 mg/l NH4-N have been measured in drainage waters following the application of AN in both March and May 2000 at ADAS Boxworth, and of up to 4.5 mg/l in surface runoff from a clay soil in Devon (Hatch et al., 2004). There was no UK information quantifying the leaching of NO2-N from different fertiliser-N materials.

Agriculture is estimated to be responsible for 5-10% of FFD non-compliance cases, with c.50% of these cases from point sources and c.50% from diffuse sources (Defra, 2003). Data were sourced from the Environment Agency’s Harmonised Monitoring Scheme, which showed that the mean NH4-N concentration in surface waters in Great Britain (1980-2003) was 0.16 mg/l (90 percentile value = 0.36 mg/l; 10 percentile value = 0.02 mg/l). Further data were kindly supplied by the Environment Agency (Melanie Newson) on 492 non-compliance cases in five English EA regions (1993-2003). Of these non-compliance cases, agriculture was responsible for 19%, with some regional variation (26% in north-west and south west regions; 11% in south, north-east and Thames regions). Of the 92 non-compliance cases from agriculture, the majority were linked with poor manure management, e.g. contaminated farm/agricultural drainage waters, surface runoff following manure spreading and other diffuse sources. From these data, it can be concluded that the majority (80-90%) of agricultural non-compliance appears to be from poor manure management. However, a significant increase in the use of urea was considered to have the potential to increase the contribution of fertiliser N to FFD non-compliance cases.

The main objective of this work was to assess the likely impact of an increase in the use of urea fertilisers in UK agriculture on N losses as NO3-N, NH4-N, NO2-N and urea-N to ground waters (2 experiments) and surface waters (6 experiments). In addition, the effect of adding a urease inhibitor (Agrotain) to urea fertiliser on N losses in drainage and surface runoff waters was examined.


3. Experimental design, treatments and methods

3.1 Experimental sites

Eight field experiments were carried out at four sites in spring 2004 and spring 2005 to assess the potential for losses of fertiliser N to ground and surface waters (Table 1). Two experiments were carried out on the free draining sandy soil at Woburn, and six experiments on poorly drained soils (one each at Rosemaund and Boxworth, and four at Rowden). At each site, N losses (NO3-N, NH4-N, NO2-N and urea-N) in surface runoff and drainage waters were measured following contrasting spring fertiliser N applications (AN, urea and U+Ag).

Table 1. Experimental details (2004 & 2005)

Site / Year / Soil texture / Exp No / Crop / N Application
Type / Sowing date / Date (day/month) / N rate
(kg/ha)
Free draining soils
Woburn / 2004 / Sandy loam / 1 / W. Wheat / 4/10/03 / 5/4 / 100
2005 / Sandy loam / 2 / W. Wheat / 23/09/04 / 4/4 / 100
Poorly drained soils
Rosemaund / 2004 / Silty clay loam / 3 / Grass / - / 23/3 / 100
Boxworth / 2005 / Clay loam / 4 / Stubble / - / 28/2 / 100
Rowden / 2004 / Silty clay loama / 5 / Grass / - / 5/2, 18/3 / 80, 40
2005 / Silty clay loama / 6 / Grass / - / 9/2, 14/3 / 70, 40
2004 / Silty clay loamb / 7 / Grass / - / 19/2, 18/3 / 60, 90
2005 / Silty clay loamb / 8 / Grass / - / 14/2, 14/3 / 60, 90
a Rowden drainage plots, b Rowden lysimeters.

At Woburn on the sandy textured soil (Experiments 1 and 2), N losses to a depth of 90 cm were measured using teflon suction cup samplers following the spring application of fertiliser N to winter wheat.

On the poorly drained soils, losses of fertiliser N in surface runoff and drainage waterflows were measured at Rosemaund (Experiment 3) and in drainage waterflows at Boxworth (Experiment 4). At Rosemaund, N concentrations in surface runoff and drainage waterflows (collected at a depth of 100 cm) were measured separately. At Boxworth, drainage waterflows were collected from pipes at a depth of 100 cm, which were supplemented by mole drains at a depth of 55 cm. In the Rowden drainage experiments (Experiments 5 and 6), waterflows were collected from the 0-30 cm layer on the undrained plots, and from mole (50 cm depth) and pipe (85 cm depth) drains on the drained plots. Waterflows from the undrained Rowden lysimeter experiments (Experiments 7 and 8) were collected from the 0-30 cm layer only. Cumulative N losses were calculated by multiplying N concentrations by the measured (or calculated) waterflow volumes at each site in each year. Additional site details, including soil physical and chemical properties, are shown in Table 1 of Appendix II