transport of nitrogen in soil water following the ...€¦ · transport of nitrogen in soil water...

Hydrologicai Sciences-Journal-des Sciences Hydrologiques, 45(1) February 2000 61

Transport of nitrogen in soil water following the application of animal manures to sloping grassland

R. J. PARKINSON Department of Agriculture and Food Studies, Seale-Hayne Faculty, University of Plymouth, Newton Abbot TQI2 6NQ, UK e-mail: [email protected]

P. GRIFFITHS & A. L. HEATHWAITE Department of Geography, University of Sheffield, Sheffield S10 2TN, UK

Abstract A field experiment was conducted on a sloping grassland soil in southwest England to investigate the downslope transport of nitrogen in soil water following the application of cattle manure, slurry and inorganic fertilizer. Transport of nitrogen (N) species was monitored on hydrologically isolated plots. Manure (50 t ha"1), slurry (50 m3 ha"1) and fertilizer (250 kg N ha") were applied in February/March 1992. Subsurface water movement, by both matrix and preferential flow, was the dominant flow route during the experiment. Subsurface and surface nutrient flow pathways were monitored by analysing soil water and surface runoff for NO3-N, NH4-N and total N. Subsurface flow chemistry was dominated by NO3-N, with concentrations usually between 2 and 5 mgN0 3 -N dm"3. Differences between fertilizer and manure treatments and the untreated control were not significant. Significantly elevated NO3-N concentrations were observed in soil water in the buffer zone, indicating the importance of a buffer zone at least 10 m wide between manure spreading zones and an adjacent water course.

Transport d'azote après l'épandage de fumier animal sur des prairies en pente Résumé Une expérience a été réalisée sur le sol de prairies en pente du sud-ouest de l'Angleterre pour étudier le transport d'azote après l'épandage de fumier de vache, de purin et d'engrais inorganiques. Le transport d'azote (N) a été mesuré sur des parcelles isolées sur le plan hydrologique. Le fumier (50 t ha"1), le purin (50 m3 a"1) et les engrais (250 kg ha"') ont été épandus en février/mars 1992. L'écoulement souterrain de l'eau, que ce soit dans la matrice ou selon des écoulements préférentiels, représente la fraction dominante de l'écoulement au cours de l'expérience. Les flux des éléments nutritifs, sous et sur la surface ont été évalués à partir de l'analyse de NO3-N, NH4-N et N-total de l'eau du sol et celle du ruissellement. Le NO3-N domine la composition chimique de l'écoulement de subsurface avec des concentrations allant de 2 à 5 mg de NO3-N dm"". Les différences entre les zones soumises à l'épandage d'engrais ou de fumier et une zone témoin non-traitée ne sont pas considérables. Des concentrations de NO3-N importantes ont été observées dans l'eau d'écoulement de la zone tampon, soulignant l'importance d'une zone tampon d'au moins 10 m de large entre les zones d'épandage et le cours d'eau.

INTRODUCTION

Recent annual surveys of water pollution incidents in England and Wales by the Environment Agency (1998, 1999) indicate that inadequate procedures for handling and disposal of animal manure and slurry continue to be the most common cause of serious water contamination from agricultural sources. In many livestock/grassland systems, animal waste storage capacity is considerably less than the total production during the period of winter housing, resulting in the need to land spread liquid wastes,

Open for discussion until 1 August 2000

mailto:[email protected]

62 R. J. Parkinson et al.

such as slurry and dirty water, onto soils at field capacity (Parkinson, 1993). Such activity has been shown to carry considerable risks of nutrient export, notably ammonium-nitrogen (NH4-N) and phosphorus, to water courses (Parkes et ah, 1997). On sloping soils, saturation-excess overland flow may well occur following low rates of liquid waste application during the winter months. Brookman et al. (1994) noted that the application of 5 mm of dirty water (<2% dry matter) to a clay loam soil in winter resulted in 35% runoff.

Traditionally, research into agricultural sources of nutrients entering water courses has focused on the losses of inorganic forms of nitrogen (N) from either arable or grassland soils, particularly following application of inorganic fertilizers. More recently, attention has been directed at organic sources of these nutrients, particularly animal manures (Sharpley et al., 1994; Heathwaite et al, 1997). Experiments conducted in The Netherlands on N (Hack-ten Broek et al, 1996) have shown that management of livestock and manure strongly controls losses from grassland soils. Such losses need to be characterized in terms of both organic and inorganic forms of N and P in order to indicate the potential contribution of fertilizer and manure sources to runoff.

Export of nutrients contained in livestock wastes in soil water or runoff depends upon a number of factors, such as manure characteristics, time of application, environmental conditions post-spreading, topography and crop management. For example, Beckwith et al. (1998) observed that application of slurry to land during the winter months resulted in a loss of 27% of slurry-N compared to only 6% of manure-N. Intensive livestock farming on steeply sloping grassland may lead to reduced nutrient retention and increased delivery of nutrients to aquatic environments (Heathwaite et al., 1990). The application of large volumes of animal manures in either solid or liquid form can modify both the magnitudes of nutrient export and the surface and sub-surface pathways that nutrients follow from soil to receiving waters. Muscutt et al. (1993) have shown that the development of dynamic models of loss incorporating hydrological, topographical and land management controls, for example buffer strips, can aid prediction of these nutrient losses and hence assessment of the impacts on aquatic systems.

The results reported here form part of a wider study of the pathways and processes involved in the transport of nitrogen (N) and phosphorus (P) following applications of cattle manure to sloping grassland. The overall aim of the research was to increase understanding of magnitude, timing, route and fractionation of N and P in order to assess relative contribution to water course pollution, and was described in full by Griffiths (1994). The present paper describes the patterns and pathways of N export in soil water from sloping manure-amended grassland under natural rainfall. A previous paper (Heathwaite et al., 1998) reported the results of experiments conducted under artificial rainfall.

EXPERIMENTAL METHODS

Site and soil

A plot-scale experiment was established within a 2 ha permanent pasture field on Seale-Hayne Farm, University of Plymouth, UK. The site lies approximately 50 km east of Plymouth, in southwest England (National Grid Reference SX828721).

Transport of nitrogen in soil water following application of animal manures to grassland 63

Fertilizer, FYM or slurry application

Control (no treatment)

Buffer (no treatment)

minimi Surface runoff collection trough Fig. 1 Experimental layout for the central block showing upper (Zone 1) and lower (Zone 2) treatment zones and the buffer zone for each of the four plots within the block.

Average annual rainfall for the site is 967 mm, of which approximately 60% falls between the months of October and March inclusive. Groundslope across the plots ranges from 10-15% over the lower 10 m section, adjacent to a water course, to approximately 20-25% on the upper 20 m. Slope angles of this magnitude are typical of many fields in this area of southwest England, and previous research (Heathwaite & Burt, 1991; Heathwaite et al. 1990) has demonstrated that these slope sections adjacent to water courses exert a strong influence on patterns of nutrient loss to water courses. Figure 1 shows the experimental design for the central block of three replicate blocks at the research site.

The soil of the experimental site is a weakly structured brown earth of the Denbigh Association (Findlay et al, 1984). The silty clay loam topsoil merges at 0.20-0.25 m into a subsoil horizon of similar texture over weathered Devonian slate. Depth to slate varies from 0.3 m at the top of the experimental site to >1.0 m adjacent to the water course. Twomlow et al. (1990) carried out detailed investigations into the drainage


characteristics of these soils, and noted that average winter topsoil water contents range from 0.40 to 0.46 m3 m"3. Topsoil pH (1 organic matter (by loss on ignition) was 0.127 g g" range from 0.40 to 0.46 m3 m"3. Topsoil pH (1:2.5 watensoil ratio) was 5.5, and

Experimental design

The objective of the experimental programme was to evaluate downslope N and P transport following the application of liquid and solid animal wastes to grassland. Current recommendations in England and Wales (MAFF, 1998), which have remained unchanged since the 1991 edition of the Code of Good Agricultural Practice for the Protection of Water, state that maximum individual application rates should not exceed 50 m ha" (liquids) or 50 t ha"1 (solids), and that an unspread strip at least 10 m wide should be left adjacent to water courses. In addition, applications should not be repeated within three weeks. Using these guidelines, three treatments were selected: solid farmyard manure (FYM) at a rate of 501 ha"1, slurry at 50 m3 ha"1 and inorganic fertilizer at 250 kg N ha"1. Each replicate block contained an untreated control plot (Fig. 1). The plots were replicated three times across the slope, orthogonal to the contours. The position of each plot within the block was allocated on a random basis. Plots were 5 m wide and 30 m long, the lower 10 m of which received no applications of manure, slurry or fertilizer. Hydrological isolation of the experimental area at the upper boundary was ensured by the installation of an impermeable membrane inserted into the slate at the base of the soil profile. Plots were not isolated hydrologically from one another but were separated by a 1 m untreated zone. As the ground slope throughout the experimental area was uniform, and the contours were parallel to the lower plot boundary, it was assumed that the dominant direction of water movement would be orthogonal to the contours and the lower plot boundary.

The timing of manure and fertilizer application for the main phase of experimentation, which took place during winter and spring 1992, is given in Table 1. All treatments were applied to the upper 20 m treatment section of each plot, allowing the lower 10m section to act as an untreated buffer zone. Analysis of the nutrient content of both the slurry and manure after application revealed that total N contents of each were variable. Manure treated plots received one application, containing 360 kg N ha"1, while slurry treated plots received two applications, containing 140 kg N ha"1 on each occasion. Dates of application are given in Table 1. Inorganic

Table 1 Experimental treatments, manure and fertilizer application rates and quantities of nutrients applied.

Date Treatment Rate Nitrogen applied (kg N ha"1)

12 February 1992

13 March 1992

Manure Slurry Fertilizer Manure Slurry Fertilizer

50 t ha-1

50 m3 ha"1

none nonef

50 m3 ha"1

360 140 _ -140 250

* Too early in the year for spring fertilizer application. * Manure residues from previous application remained on the plot.


fertilizer was applied as a mixture of ammonium nitrate, ammonium phosphate and potassium chloride.

Field instrumentation

The experimental plots were subdivided into three 10 m zones. The zone adjacent to the water course received no treatment, and hence acted as a buffer strip (Fig. 1). Measurements of matric potential and water content were confined to the control plot in each replicate block, in order not to influence soil water chemistry on treated plots. Duplicate mercury manometer tensiometers were installed centrally at 0.10, 0.35 and 0.70 m depth within each of Zone 1, Zone 2 and the buffer zone, and were monitored daily during the experimental period. Soil water content was measured weekly by time domain reflectometry (TDR; Topp et al, 1980), using probes inserted horizontally into the soil at the same locations and depths as the tensiometers. Piezometers were installed at the same location as the tensiometers, at depths of 0.35, 0.50 and 0.70 m. Dipwells were located in the buffer zone of each control plot, at a depth of 2.00 m, and at the base of Zone 2 at a depth of 1.00 m.

Soil water for chemical analysis was collected using ceramic suction cup samplers installed at two depths, 0.10 and 0.35 m, in all three zones of each treatment in two of the three replicate blocks. Suction cup samplers have been used successfully to assess solute flow in shallow, stony soils (Johnson & Smith, 1996; Williams & Lord, 1997), but due to the uncertainties in the sphere of influence of suction samplers noted by Grossman & Udluft (1991), these samples were only used to indicate broad changes in the concentrations of N species under the different treatments. Additional measurements of solute movement were taken using surface runoff collection troughs 0.80 m wide, which were located in all three zones of each of the four plots in the central block, arranged offset from one another to allow the free flow of water down the length of the slope. Samples were collected daily during periods of significant rainfall which generated surface runoff. Throughflow pits 0.80 m wide were sited at the bottom of each of the four plots in the central block of the site. Pits were installed to a depth of 0.80 m, and lined with polythene to retain water draining from the upslope profile. Discharge volumes were recorded on a daily basis. Samples were collected for analysis at weekly intervals, or daily during periods of significant rainfall.

Analytical methods

All water samples were filtered through Whatman GFC papers, refrigerated and then analysed within 48 h of collection for NO3-N and NH4-N on a Technicon 2 autoanalyser. NO3-N was measured as total oxidizable nitrogen (NO3-N + NO2-N) using sodium hydroxide, hydrazine-copper and sulphanilamide reagents, while NH4-N was measured using salicylate and dichloroisocyanuric acid-sodium salt reagents and a citrate buffer. Unfiltered portions of each sample were analysed for total N using a simultaneous persulphate digestion technique (Hosomi & Sudo, 1986). Organic plus particulate N within each sample was then calculated by difference. The


total nutrient content of manure and slurry samples was determined by standard methods (MAFF, 1986).

RESULTS AND DISCUSSION

Rainfall and infiltration

Both immediately prior to, and during the experimental period, rainfall was less than the long-term mean. Monthly rainfall for the experimental period as a percentage of the long-term mean (1961-1991) was as follows: January 27%; February 57%; March 61%; April 122%. The maximum daily total recorded during the experimental period was 11 mm (Fig. 2), and the maximum hourly intensity recorded was S m m n 1 . Rainfall was never of sufficient intensity or duration to generate significant quantities of either infiltration-excess or saturation-excess surface runoff. Nitrogen concentrations in rain water were found to be negligible throughout this period. Ammonium-N was below the limits of detection by autoanalysis, while NO3-N was always <0.5 mg l"1.

M Rainfall

14-Feb 24-Feb 19-Mar 29-Mar

Fig. 2 Daily rainfall and resultant flow into throughflow collection pits during the experimental programme, compared with estimates from other field data. (Observed flow values are mean flow rates into all the throughflow collection pits.)

Soil water content

Soil water contents for the upper and lower treatment zones and the buffer zone (Fig. 1) for two example occasions, 8 March and 8 April 1992 are shown in Table 2. Values ranged from 0.38 to 0.45 m3 m"3 at 0.10 m depth throughout the period, and decreased down the profile in all cases. No significant differences (p > 0.05) in soil water content at 0.10 m depth were observed between the upper treatment zone, the lower treatment zone or the buffer zone. In contrast, subsurface values (taken progressively deeper as the soil profile deepened downslope) in the buffer zone were


Table 2 Water content (mJ m"J) in the topsoil (0.10 m) and at the base of the soil profile in the upper and lower treatment zones and the buffer zone on 8 March and 8 April 1992.

8 March 8 April

Upper treatment zone: 0.10m 0.35 m 0.40 0.31 0.41 0.34

Lower treatment zone: 0.10 m 0.50 m 0.41 0.36 0.44 0.37

Buffer zone: 0.10m 0.43 0.45

0.70 m 0.40 0.43

0.0-0.10 m3 nf3 higher than the upper treatment zone throughout the experimental period.

Twomlow et al. (1990) recorded winter mean water content (field capacity) for the soil profile (0-0.80 m) of 0.40 m3 nf3 in an adjacent field. Both the winter mean water content and the duration for which it is maintained are dependent partly on the prevailing weather conditions. During this experiment there was no significant variation in water content for the period mid-February to mid-April 1992, which indicated that water contents were at winter mean values throughout the experimental period.

Soil hydraulic potential and subsurface throughflow

The matric potential at each soil depth in the treatment zones was negative throughout the experimental period, indicating that the soil water content was below saturation point and that the water was held under some degree of suction (Fig. 3). In the buffer areas, a zone of saturation rose to within 0.70 m of the surface both in mid-February and early April. Matric potentials varied little with depth through the soil profile, the only real difference being in the degree of fluctuation. This tended to decrease with increasing depth, as the effects of infiltration and any évapotranspiration became more restricted.

Rates of lateral subsurface throughflow at the experimental site were calculated from the available data on soil water content, hydraulic potential, and hydraulic

06-Feb 16-Feb 26-Feb 07-Mar 17-Mar 27-Mar 06-Apr 16-Apr

Fig. 3 Mean matric suction in the topsoil (0.10 m) and selected subsoil horizons, Zone 1, Zone 2 and the buffer zone, for the period 6 February-16 April 1992.


conductivity. This estimate was compared with the observed quantities of water draining into the throughflow collection pits. Bulk density, water content at 0 kPa and saturated hydraulic conductivity (Ksat) determined in the laboratory using soil cores are shown in Table 3. Bulk density increased from 1.28 Mgm"3 at 0.20 m depth to 1.43 MgnT3 at 0.50 m, although this increase was not statistically significant (p > 0.05). Similarly, soil water content at saturation was higher in the topsoil than the subsoil (Table 3), but the means for each depth were not significantly different (p > 0.05). Saturated hydraulic conductivity, Ksat, was spatially highly variable, with values up to 150 mm h"1 being observed. Errors can occur with the laboratory determination of Ksat, particularly due to leakage down the sides of the core if extraction from the soil profile is not carried out with care (Klute & Dirksen, 1986). However, such errors would not account for the high conductivities observed in some of the cores, which were indicative of the presence of cracks or fissures in the soil, and therefore indicated rapid macropore flow under saturated conditions. The lower conductivities of 4-8 mm h"1 probably represented rates of saturated flow through a soil matrix that did not contain cracks or fissures, and where the largest pores were transmission pores (as defined by Bache, 1990).

Table 3 Bulk density, water content at saturation and saturated hydraulic conductivity of soil samples taken from a mid-slope position on the experimental plots (core volume: 1038 cm3).

Soil depth 0.50 m (n = 3): Soil depth 0.20 m {n = 4): Mean SE Range Mean SE Range

Bulk density L28 004 1.16-1.37 L43 CUT 135-1.68 (Mg mJ) Water content at 0.52 0.04 0.45-0.60 0.42 0.03 0.37-0.45 0 kPa (m3 m'3) Ksat (mm h'1) 50 19 8-86 101 49 4-150

In making this estimate of the rate of water movement, it was assumed that given the soil water conditions from February to April 1992 and the low rainfall, most of the subsurface throughflow would occur as matrix flow. Therefore, the maximum hydraulic conductivity was taken as 8 mm h"1. Here, the relationship between hydraulic conductivity, j , and water content, x, was described by a simple exponential:

r ( e V (i) where saturated volumetric water content was set at 50% and saturated hydraulic conductivity was taken as 8 mm h"1.

In order to estimate downslope water movement rates, each zone was treated mathematically as a uniform storage block, with one mean water content and hydraulic potential. Working only on the centre replicate where the throughflow collection pits were sited, values of water content and corresponding hydraulic conductivity were calculated. Data from the tensiometers were used to calculate mean hydraulic gradients between the lower treatment zone and the buffer zone for the same dates. Assuming laminar flow in the matrix, Darcy's Law was applied, and equation (2) was used to obtain estimates of the rate of flow of water into the throughflow collection pits.

Q = kAd<p/dL (2)


where Q is flow (m day"1), k is hydraulic conductivity of the water content in the buffer zone (m day"1), A is the area of the collection pit face (m2), and dcp/dZ, is the hydraulic potential gradient between the zones (unitless).

Figure 2 shows the predicted volumes of flow and that intercepted by the throughflow collection troughs. During the drier period of mid to late March, rates of flow were overestimated. However, even following rainfall of relatively low intensity occurring in February and early April, rates of flow were underestimated. There are two possible explanations for this. The throughflow collection pits could have prevented the matrix flow of water during drier periods, through the build up of a saturated wedge at the pit face, which then drained rapidly following rainfall. In addition, the water content and hydraulic conductivity of the whole soil may have been at a critical boundary during much of the monitoring period, such that during even low intensity rainfall, substantial preferential flow occurred through the larger transmission pores and fissures. Coles & Tradgill (1985) observed this effect when working on similar soils in southwest England.

Nitrogen in soil water

For soil water samples obtained from suction samplers, the differences in N concentration between the control and the treatments were not significant at any time or space on the site (p > 0.05). Despite the application of organic manure and fertilizer treatments in mid-February and mid-March (Table 1), a general reduction in the concentration of both total N and NO3-N was observed from mid-February to the end of March (Fig. 4). This is in contrast with losses measured following rainstorm simulation experiments on the same site which led to artificially induced runoff and

19 Feb 19 Mar 26 Mar

D Ammonium and organic N

• Nitrate N

0 I ™ I ™ I ™ I ™ [ T T B B

0.10 0.35 0.10 0.35 T T B B

0.10 0.35 0.10 0.35 T T B B

0.10 0.35 0.10 0.35

Zone and soil depth, m

Fig. 4 Mean total N and N species in suction cup samples on three occasions, 13 February, 19 March and 26 March 1992, following application of manure, slurry or fertilizer (T = treated zone, B = buffer zone). Error bars indicate one standard error of the mean total N concentration.


significantly greater losses from fertilizer and manure than slurry (Heathwaite et al., 1998). In the experiment reported here, nitrogen transport under medium and low intensity rainfall led to smaller losses dominated by the nitrate-N fraction. Beckwith et al. (1998) noted that N in slurry was more prone to leaching than N in FYM, due to the higher concentration of mineral-N in the former, despite the ammonia volatilization and denitrification losses known to occur from slurry. The key role of mineralization processes was noted by Dunn et al. (1999) in modelling the relationship between agricultural practices and stream nitrate concentrations. They demonstrated that subsurface flow pathways transfer nitrate at a rate that more closely matches N mineralization than the supply from fertilizer sources. Significant increases in NO3-N were recorded both down-profile and downslope (p < 0.05), such that concentrations at 0.35 m depth in the buffer zone were higher than anywhere in the treated zones, indicating a downslope migration of N in solution (Fig. 4). Griffiths (1994) observed that preferential flow in this weakly structured soil transported significant quantities of NO3-N downslope, but that organic and NH4-N constituted a minor fraction of the N within the soil water. This effect is confirmed in Fig. 5.

It has been demonstrated in other experiments that riparian buffer strips can be effective in reducing subsurface NO3-N loads to water courses (Moorby & Cook, 1992; Haycock & Pinay, 1993). The effectiveness of these buffer strips depends generally on a zone of saturation to within a few centimetres of the soil surface and a rich supply of organic carbon in the soil, thus creating suitable conditions for the reduction of NO3-N to gaseous forms of N by denitrification. For grass vegetated buffer zones, Haycock & Pinay (1993) have observed an 84% retention of NO3-N within the first 5 m of the buffer, independent of the NO3-N load entering the buffer. This rate of retention rises to 99% for buffers vegetated with poplar, because of an increase in the amount of carbon in soil under trees. It is argued that this process is particularly effective in winter, when plant uptake of NO3-N is at a minimum and load

E

• Control

B Slurry

• Manure

D Fertilizer

Nitrate N Ammonium N Organic N

Fig. 5 Effect of treatment on N speciation in throughflow collection pits and dipwells averaged over the period March-April 1992 (n = 21). Error bars indicate one standard eiTor of the mean.


in throughflow is at a maximum. However, it is important that the buffer vegetation remains undisturbed and that the buffer is not underdrained.

For sites such as this, where steep slopes end rather abruptly at water courses, the widespread occurrence of conditions favourable for riparian zone denitrification is not guaranteed. This might be even more the case if the pasture was underdrained. Garwood et al. (1985) noted that when 400 kg N ha"1 was applied to drained and undrained sections of permanent grassland, subsurface loss of NO3-N was 216 kg ha"1

from the drained section compared with 69 kg ha"1 from the undrained section, and denitrification losses were 46 kg ha"1 and 64 kg ha"1, respectively. However, when Barraclough et al. (1992) examined the transport of NO3-N from grazed swards following applications of fertilizer N, they could not support the hypothesis that drainage increased rates of NO3-N leaching and reduced denitrification. Other factors were found to be of importance, such as the age of the grass sward. They concluded that predicting NO3-N leaching from a grazed sward following fertilizer application was extremely difficult.

Water sampled from the throughflow collection pits and the dipwells gave an indication of the concentrations of N actually flowing from the bottom of the plots. Because of low rainfall, few samples were collected, and therefore the results shown here for N speciation are the combined results from both collection troughs and dipwells for the period mid-March to mid-April 1992 (Fig. 5). Such results should be treated with caution, as they disguise the temporal reductions in N concentration that were observed here as well as in the suction samples. The N in these throughflow samples was dominated by NO3-N, concentrations being similar to those in the suction samples, and relatively low compared to the EU drinking water directive limit of 11.3 mg NO3-N dm"3. Differences between the control and the treatments were not significant (p > 0.05).

Implications for buffer zone management and water quality

The UK maximum admissible concentration (MAC) for NH4-N in public water supply is 0.5 mg dm"3 as NH4

+, and 1 mg Kjeldahl-N dm"3 for organic-N (Heathwaite et al., 1993). On the basis of these MACs, the transport of N in subsurface flow from this site during the monitoring period was not particularly high. Generally, the patterns observed in the suction cup samples were repeated in the throughflow collection pits and dipwells. Ammonium-N and organic-N may be largely immobilized in the soil, while any increase in the rate of mineralization and nitrification was more than offset by the processes of plant uptake and perhaps denitrification, thus rendering differences between treatments insignificant.

Interpretation of these results needs to include a consideration of the interaction between land use and topography before applying any knowledge gained from plot experiments about buffer zone effectiveness. The application of two key concepts in hillslope hydrology, Partial Source Area and Variable Source Area, is important here. The former (after Betson, 1964) is relevant where infiltration-excess overland flow is more likely, perhaps due to animal poaching or heavy machinery. This can increase surface transport of N (Heathwaite et al., 1990). The latter is a variable zone of saturated soil, e.g. in a hollow, which may contribute to increased rates of subsurface flow and saturation excess surface runoff (Anderson & Burt, 1990). Heathwaite et al.

72 R. J. Parkinson et al

(1998) have shown that the form of manures applied to land influences the magnitude and chemical fractionation of nutrient export by surface runoff. The work reported here indicates that the form of N applied to sloping grassland has less influence on subsurface N transport, indicating effective immobilization, retention or uptake of N. Nevertheless, an untreated buffer zone at least 10m wide is an important management tool for the control of N losses to water courses,

Acknowledgements P. Griffiths was in receipt of a Natural Environment Research Council postgraduate studentship (GT4/90/AAPS/52) during the course of this work. Technical support for this work was provided by Peter Russell, Patrick Bugg, Frances Vickery, Anne Clowes and Chris Duller. Access to the site was granted by Richard Newington, Seale-Hayne Farm Manager.

REFERENCES

Anderson, M. G. & Burt. T. P. (1990) Subsurface runoff. In: Process Studies in Hillslope Hydrology (ed. by M. G. Anderson & T. P. Burt), 365-400. Wiley, Chichester, West Sussex, UK.

Bache, B. W. (1990). Solute transport in soils. In: Process Studies in Hillslope Hydrology (ed. by M. G. Anderson & T. P. Burt), 437-462. Wiley, Chichester, West Sussex, UK.

Barraclough, D., Jarvis, S. C , Davies, G. P. & Williams, J. (1992) The relation between fertiliser nitrogen applications and nitrate leaching from grazed grassland. Soil Use Manage. 8, 51-56.

Beckwith, C. P., Cooper, J., Smith, K. A. & Shepherd, M. A. (1998) Nitrate leaching loss following application of organic manures to sandy soils in arable cropping. I. Effects of application time, manure type, overwinter crop cover and nitrification inhibition. Soil Use Manage. 14, 123-130.

Betson, R. P. (1964) What is watershed runoff? J. Geophys. Res. 69, 1541-1551. Brookman, S. K. E., Pain, B. F. & Harrod, T. R. (1994) Low rate irrigation of dilute farm wastes. Research and

Development Note 262, National Rivers Authority, Bristol, UK. Coles, N. & Trudgill, S. T. (1985) The movement of nitrate fertiliser from the soil surface to drainage waters by

preferential flow in weakly structured soils, Slapton, South Devon. Agric. Ecosys. Environ. 13, 241-259. Dunn, S., Domburg, P., Edwards, A. C. & Ferrier, R. C. (1999) A simple modelling approach to identify processes

controlling stream nitrate in an agricultural catchment. In: Impact of Land-Use Change on Nutrient Loads From Diffuse Sources (ed. by A. L. Heathwaite), 135-142. IAHS Publ. no. 257.

Environment Agency (1998) Water Pollution Incidents in England and Wales 1997. HMSO, London. Environment Agency ( 1999) Water Pollution Incidents in England and Wales J998. HMSO, London. Findlay, D. C , Colbourne, G. J. N., Cope, D. W., Harrod, T. R., Hogan, D. V. & Staines, S. J. (1984) Soils and their use in

South West England. Soil Survey of England and Wales Bull. 14, Harpenden, UK. Garwood, E. A., Ryden, J. C. & Tyson, K. C. (1985) Nitrogen losses from drained grassland. In: Proc. British Grassland

Society Occasional Synip. no. 20, 70-74. British Grassland Society, Reading, Berkshire, UK. Griffiths, P. (1994) Nutrient dynamics in manure amended grasslands. PhD Thesis, University of Sheffield, Sheffield, UK. Grossman, J. & Udluft, P. (1991) The extraction of soil water by the suction cup method: a review. J. Soil Sci. 42, 83-93. Hack-ten Broek, M. J. D., De Groot, W. J. M. & Dijkstra, J. P. (1996) Impact of excreted nitrogen by grazing cattle on

nitrate leaching. Soil Use Manage. 12, 190-198. Haycock, N. E. & Pinay, G. (1993) Groundwater nitrate dynamics in grass and poplar vegetated riparian buffer strips

during the winter. J. Environ. Qual. 22, 273-278. Heathwaite, A. L. & Burt, T. P. (1991) Predicting the effect of land use on stream water quality. In: Sediment and Stream

Water Quality in a Changing Environment: Trends and Explanation (ed. by N. E. Peters & D. W. Walling) (Proc. Vienna Symp., August 1991), 209-218. IAHS Publ. no. 203.

Heathwaite, A. L., Burt, T. P. & Trudgill, S. T. (1990) The effect of agricultural land use on nitrogen, phosphorus and suspended sediment delivery to streams in a small catchment in South West England. In: Vegetation and Erosion (ed. by J. B. Thomes), 161-178. Wiley, Chichester, West Sussex, UK.

Heathwaite, A. L., Burt, T. P. & Trudgill, S. T. (1993) Overview—the nitrate issue. In Nitrate: Processes, Patterns and Management (ed. by T. P. Burt, A. L. Heathwaite & S. T. Trudgill), 3-21. Wiley, Chichester, West Sussex, UK.

Heathwaite, A. L., Griffiths, P. & Parkinson, R. J. (1998) Nitrogen and phosphorus in runoff from grassland with buffer strips following application of fertilizers and manures. Soil Use Manage. 14, 142-148.

Heathwaite, A. L., Haygarth, P., Jarvis, S. C , Griffiths, P. & Parkinson, R. J. (1997) Phosphorus loss from grassland soils: implications of land management for the quality of receiving waters. In: Freshwater Contamination (ed. by B. Webb) (Proc. Rabat Symp., April-May 1997), 177-186. IAHS Publ. no. 243.

Hosomi, P. J. & Sudo, R. (1986) Simultaneous determination of total nitrogen and total phosphorus in freshwater samples using persulphate digestion. Int. J. Environ. Studies 27, 267-275.

Transport of nitrogen in soil water following application of animal manures to grassland lis

Johnson, P. A. & Smith, P. N. (1996) The effects of nitrogen fertilizer rate, cultivation and straw disposal on the nitrate leaching from a shallow limestone soil cropped with winter barley. Soil Use Manage. 12, 67-71.

Klute, A. & Dirksen, C. (1986) Hydraulic conductivity and diffusivity: laboratory methods. In: Methods of Soil Analysis Part 1: Physical and Mineralogical Methods (ed. by A. Klute), 687-734. American Society of Agronomy, Madison, Wisconsin, USA.

MAFF (Ministry of Agriculture, Fisheries and Food) (1986) The Analysis of Agricultural Materials. Reference Book 427. HMSO, London.

MAFF (Ministry of Agriculture, Fisheries & Food) (1998) Code of Good Agricultural Practice for the Protection of Water (second edn). MAFF Publications, London.

Moorby, H. & Cook, H. F. (1992) The use of fertiliser free grass strips to protect dyke water from nitrate pollution. Aspects Appl. Biol. 30, 231-234.

Muscult, A. D., Harris, G. L., Bailey, S. W. & Davies, D. B. (1993) Buffer zones to improve water quality: a review of their potential use in UK agriculture. Agric. Ecosys. Environ. 45, 59-77.

Parkes, M. E., Campbell, J. & Vinten, A. J. A. (1997) Practice to avoid contamination of drainflow and runoff from slurry spreading in spring. Soil Use Manage. 13, 36-42.

Parkinson, R. J. (1993) Changes in agricultural practice. In: Nitrate: Processes, Patterns and Management (ed. by T. P. Burt, A. L. Heathwaite & S. T. Trudgill), 321-339. Wiley, Chichester, West Sussex, UK.

Sharpley, A. N., Chapra, S. C , Wodepohl, R., Sims, J. T. & Daniel, T. C. (1994) Managing agricultural phosphorus for protection of surface waters: issues and options. J. Environ. Qual. 23, 437-451.

Topp, G. C , Davis, J. L. & Annan, A. P. (1980) Electromagnetic determination of soil water content: measurement in coaxial transmission lines. Wat. Resour. Res. 16, 574-582.

Twomlow, S. T., Parkinson, R. J. & Reid, I. (1990) Soil loosening and drainage of structurally unstable silty soils. J.Hydrol. 121,63-83.

Williams, J. R. & Lord, E. I. (1997) The use of porous ceramic cup samplers to measure solute leaching on chalk soils. Soil Use Manage. 13, 156-162.

Received 22 December 1998; accepted 16 September 1999

transport of nitrogen in soil water following the ...€¦ · transport of nitrogen in soil water...

Documents