Nitrogen leaching from soil lysimeters irrigated with dairy shed effluent and having managed drainage.
Abstract: The leaching of nitrogen (N) from agricultural soils is undesirable for environmental and health reasons. We investigated the effects of adding dairy

shed effluent (DSE), irrigated on a weekly basis during the milking season, on the amounts and forms of N leached from large undisturbed soil monolith lysimeters of a Gley Soil over a period of 2 years. Drainage was managed using a weir that maintained the water table at 3 depths: 25 (high), 50 (medium), or 75 (low) cm below the soil surface. The low water table treatment represented the usual situation for the soil when drained. If undrained, it would be usual during wet periods in the field for a perched water table to form on the slowly permeable horizon at 75 cm depth.

The total amount of N irrigated onto the lysimeters in the first milking season was equivalent to a total of 511 kg N/ha.year, and up to 33.3 kg N/ha.year leached from the soil. The losses from lysimeters receiving effluent were nearly double those from lysimeters receiving an equivalent amount of water only, when the high and medium water tables were imposed. Adding effluent caused only a small increase (7 kg N/ha) in total N leached in the low drainage treatment. In the second milking season, the effluent-N loading was increased to 1518 kg N/ha.year and the pasture was managed to simulate a 'cut and carry' land treatment system. Under these conditions, up to 131.4 kg N/ha.year leached from the soil, which was nearly 100 kg N/ha more than lysimeters receiving only water. The total N leaching losses represented a similar proportion of added N (7% and 9%) for years 1 and 2, respectively. Most of the leached N (80-90%) was in organic N form.

The managed drainage treatment in which the water table was nearest the soil surface resulted in less N being leached in the nitrate-N ([NO.sub.3]-N) form ([is less than] 2.5 kg N/ha.year) than the other drainage treatments (6-12 kg N/ha.year); however, it did result in the greatest amount of organic and total N leached (33 and 131 kg N/ha for Year 1 and 2, respectively). The smaller amount of [NO.sub.3]-N leached from the high water table treatment is attributed to enhanced denitrification, and the greater amount of organic N is attributed to preferential flow. Although [NO.sub.3]-N concentrations in leachate generally remained below World Health Organisation (WHO) standards in all treatments, the large amount of N leached in organic form would suggest that inorganic N should not be the only form of N considered when measuring N leaching losses.

Additional keywords: nitrate leaching, water table management, irrigation, soil lysimeters, waste water.
Subject: Soils (Leaching)
Drainage (Management)
Dairy waste (Research)
Authors: Singleton, P. L.
McLay, C. D. A
Barkle, G. F.
Pub Date: 03/01/2001
Publication: Name: Australian Journal of Soil Research Publisher: CSIRO Publishing Audience: Academic Format: Magazine/Journal Subject: Agricultural industry; Earth sciences Copyright: COPYRIGHT 2001 CSIRO Publishing ISSN: 0004-9573
Issue: Date: March, 2001 Source Volume: 39 Source Issue: 2
Geographic: Geographic Scope: New Zealand Geographic Code: 8NEWZ New Zealand
Accession Number: 73023638
Full Text: Introduction

In Australasia, dairy farms produce considerable volumes of effluent following the cleaning of milking sheds after each milking. The effluent is a dilute organic waste containing faeces and urine, which is washed into a treatment pond or holding tank. In New Zealand, regional authorities are encouraging land-based irrigation of dairy shed effluent in preference to pond treatment systems that discharge to waterways.

One concern about increasing the effluent loading to land is the increased potential for nitrate ([MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]) contamination of ground water. Many studies have shown that effluent application to soil can increase [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching (Adams 1981; Vetter and Steffens 1981; Steenvoorden et al. 1986; Unwin 1986; Jarvis et al. 1987; Rate and Cameron 1992). However, some studies have reported relatively small [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching losses, despite large applications of effluent N (Sherwood 1986; Unwin et al. 1991; Smith and Chambers 1993; Di et al. 1998). The amount of N leached from the soil is a result of a range of factors such as soil chemical and physical conditions, climate, effluent composition, and timing and method of land application (Unwin et al. 1986).

A lack of increase in [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching following addition of effluent to soil has been attributed to increased [NH.sub.3] volatilisation (Smith and Chambers 1993), plant uptake (Williamson et al. 1998), and denitrification (Smith and Peterson 1982). Denitrification can be enhanced on poorly drained soils, to minimise [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching losses, by carefully manipulating irrigation or drainage (e.g. Skaggs et al. 1981; Rolston et al. 1982; Sexstone et al. 1985; Deal et al. 1986; Kliewer and Gilliam 1995).

Most studies of N leaching from cattle effluent have focussed on feedlot manure or injected slurries in cropping regimes. Very few studies have attempted to measure the amount and composition of N leached from soils following application of liquid effluent to soil. A recent study of dairy shed effluent (DSE) in New Zealand (Longhurst et al. 2000) indicated that the wastes contain, on average, approximately 0.9% solids and 200-400 mg N/L (of which 80% is organic and 18% in ammonium form). There are few studies from New Zealand on the leaching of N from applications of DSE to soil. Di et al. (1998) applied large DSE loading rates (equivalent to 400 kg N/ha.year) to pasture on undisturbed soil monolith lysimeters containing a coarse-textured sandy loam soil, and reported [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leaching losses of up to 25 kg N/ha.year. In their study, the DSE was applied twice in 2 large N loading events, and flood irrigation was used or rainfall was added to simulate `wet' years. Forms of leached N other than [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] were not reported. MacGregor et al. (1979) irrigated 20-30 mm of DSE every 10-15 days over a milking season (280 days) onto pasture on a poorly drained soil in the North Island of New Zealand. The annual loading rate from the effluent was extremely large (1125 kg N/ha). The total N losses in the drainage water (150 kg N/ha.year) were markedly higher than a grazed control which leached only 30 kg N/ha.year. The dominant form of N in mole and tile drainage water in winter was [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], while in spring it was organic N and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N. No reports on the use of controlling drainage to manipulate N leaching losses in the southern hemisphere have been found.

The objectives of this present study were to measure amounts and forms of N leaching from large soil lysimeters irrigated with DSE, and to investigate the effect of managing drainage, by controlling water table depths in the soil profile, on N leaching.

Materials and methods

Lysimeter collection and drainage management

The soil used in this study was a Te Kowhai silt loam and is classified as a Typic Orthic Gley Soil in the New Zealand Soil Classification (Hewitt 1992) and Typic Ochraqualf in Soil Taxonomy (Soil Survey Staff 1990).

One of the main characteristics of the Te Kowhai soil is a slowly permeable layer in the lower subsoil (at about 75 cm depth) on which water perches for several months of the year (Singleton 1991). The soil has worm burrows and interpedal partings that extend to more than 60 cm depth and contribute to preferential flow. The texture is silt loam to about 70 cm and silty clay to the base of the slowly permeable layer at approximately 120 cm. The Te Kowhai soil's impeded drainage makes it suitable for investigating the effectiveness of controlled drainage to enhance denitrificafion and decrease [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching losses under large effluent loading rates.

Twelve undisturbed soil monolith lysimeters (120 cm depth by 59 cm diam.) of Te Kowhai soil were collected based on the method of Cameron et al. (1992). Chemical, physical, and mineralogical characteristics of the site have been described by Joe (1986) and Singleton (1991). The lysimeters were sealed at their base and replaced in the ground around a buried access chamber into which drainage from the lysimeters was led for collection (Fig. 1). Drainage was managed using a weir that caused the water table to be maintained at a depth of 25 (high), 50 (medium), or 75 (low) cm below the soil surface. In this system, drainage could not occur until the water table reached the set height (Fig. 2). The low water table treatment represented the usual situation for the soil when drained in the field. When undrained, it would be normal for a perched water table to form on the slowly permeable horizon at 75 cm depth. The treatments, therefore, affect the amount of saturated soil through which effluent must pass before it can be leached from the profile. For the deepest water table (i.e. low drainage treatment), the effluent would not pass through any saturated soil, whereas the effluent would need to pass through 50 cm of saturated soil in the high drainage treatment (Fig. 2).

[Figures 1-2 ILLUSTRATION OMITTED]

Three replicates of each controlled drainage treatment were irrigated weekly with fresh DSE over the milking season (September-May). One replicate of each managed drainage treatment was irrigated with water only (which contained [is less than] 0.01 mg N/L), and acted as a water-irrigated control.

Irrigation and effluent collection

In order to minimise preferential flow of raw effluent, and therefore maximise absorption of effluent in the surface soil, an irrigation rate of 4 mm/h was used with a controlled sprinkler irrigator. The application rate was less than the saturated hydraulic conductivity in the field. A weekly application of 17 mm DSE, corresponding to about half the available water storage of the topsoil (0-20 cm), was applied.

The experiment was conducted over a period of 25 months from the start of June 1992 until the end of June 1994. Effluent was collected on the day of irrigation from an effluent tank which stored the day's dairy shed washings at Number 1 Dairy, Ruakura Research Centre, Hamilton. Effluent was irrigated weekly for 8 months in the first year (September-April), and for 9 months in the second (August-April) and corresponded to the milking season of each year.

The average effluent N concentration (90 mg/L) in the first year (Table 1) was fairly low for New Zealand DSE, which has recently been shown to typically vary between 200 and 400 mg/L (Longhurst et al. 2000). As a result of the fairly small quantities of N leached in the first year, it was decided to increase N loading in the second year to within this range (213 mg/L). The effluent N concentration was increased by adding faeces and urine (the urine had been added to water and left for several days to ensure complete hydrolysis such as occurred in the effluent storage tank) in amounts that maintained similar N proportions to that of effluent in the first season. The weekly irrigation of effluent in the second year caused N loading rates to be substantially above what regulatory authorities would prefer for grazed pastures (150-200 kg N/ha) in many areas of New Zealand, and it was decided, therefore, to change the pasture management to simulate a `cut and carry' system (e.g. Light 1995).

Table 1. Annual mean and range of concentrations of N components in irrigated dairy shed effluent

Effluent in the second year was amended to increase N concentration

n.d., not determined.

Leachate sampling and analysis

Leachate from each lysimeter was collected daily and stored at 2 [degrees] C. A weekly bulk sample was used for N measurements and obtained by proportionally bulking (based on daily drainage volumes) the daily leachate of each lysimeter and kept frozen prior to analysis. Pilot studies showed that the samples did not alter before being frozen. Nitrate and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] in the leachate were measured on the weekly bulked samples for each lysimeter. Total Kjeldahl N (TKN) was measured weekly for most of the first year, but practical limitations in the second year restricted sampling to proportionally bulked monthly samples of each replicate using frozen weekly samples. Dissolved organic N was determined after filtering through a 0.2 [micro]m cellulose acetate membrane filter using the TKN method minus any inorganic N. Inorganic N was measured following the auto-analyser methods of Blakemore et al. (1987). For analysis of variance it was necessary to log-transform leachate data to ensure homogeneity of variance.

Pasture management

The ryegrass/clover pasture on the lysimeters was cut approximately every 28 days to represent a `typical' dairy grazing rotation in the Waikato region. Approximately half the pasture sample was returned to the respective lysimeters in the first year. In the second year clippings were not returned to lysimeters irrigated with effluent. This was to simulate a `cut and carry' pasture system.

Results

Effluent N

The quantity of N in the dairy shed effluent that was irrigated onto lysimeters was 511 kg/ha in the first year, and 1518 kg/ha in the second year. The average irrigation event applied 15.4 kg N/ha in the first year and 37.9 kg N/ha in the second. Between 75% and 80% of the effluent N was organic and 20-25% was [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N (Table 1). Nitrate-N applied was either negligible ([is less than] 0.1 kg N/ha.month) or not detected and accounted for [is less than] 0.3 kg N/ ha. A dissolved organic N fraction was measured in the second year and represented approximately 16% of the total effluent N applied.

No significant differences in leachate volume were observed between the managed drainage treatments on a weekly, monthly, or annual basis. On an annual basis, the difference between the largest and smallest amounts of drainage was 40 mm in the first year and 22 mm in the second (Table 2).

Table 2. Mean depth of drainage (mm) from lysimeters in low, medium or high drainage treatments

Standard deviation is given in parentheses

Amounts and forms of N leached

Differences in amounts of N leached were observed between effluent- and water-irrigated lysimeters (Table 3), with up to twice as much N leached from lysimeters with medium and high water tables when effluent was added than when only water was added. The maximum difference between effluent-irrigated and water-irrigated lysimeters was [is less than] 20 kg N/ha, with a difference of only 7 kg N/ha recorded for the lysimeters with the low water table. No significant differences (P [is greater than] 0.05) in amounts of N leached were recorded between drainage treatments for lysimeters receiving effluent. In the second year, the lysimeters receiving effluent leached up to 10 times more total N than the lysimeter receiving water only (Table 3). This increase corresponded to the use of effluent with a higher N content (Table 1 and Fig. 3). The increase in N leaching losses in the second year was consistent for all drainage treatments. Approximately 7-9% of the N applied as effluent was leached in Years 1 and 2, respectively.

[Figure 3 ILLUSTRATION OMITTED]

[TABULAR DATA 3 NOT REPRODUCIBLE IN ASCII]

Organic N was generally the dominant form of N in leachates, comprising [is greater than] 70% of total N (Table 3). The only exception was in Year 1 for the lysimeter with a low water table which did not receive effluent, which had slightly more N leached as [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] than other forms. The amount of organic N leached from effluent-irrigated lysimeters was approximately double the amount leached from lysimeters receiving only water in the first year, and 6-10 times greater in the second year.

The proportion of mineral N ([MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N + [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N) in leachate was 10-20% of total N leached from the effluent-irrigated treatments. The proportion of mineral N in leachate from water-irrigated lysimeters varied from over half (no effluent, low water table) to only one-tenth (no effluent, high water table) of the total N (Table 3).

Quantities of mineral N leached from effluent-irrigated lysimeters increased in the second year, but remained similar in the water-irrigated lysimeters. The increased amount of mineral N leached in effluent-treated lysimeters was the result of increases in both [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leaching. For both effluent- and water-irrigated treatments, there appeared to be less [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leached when the height of the water table was raised closer to the soil surface (Table 3), although the difference was not significant (P [is less than] 0.05). The temporal pattern of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leaching is shown in Fig. 4, which also shows the monthly quantities of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leached from the effluent treatments. Although there was a large monthly variation, amounts leached were relatively small, particularly for the high water table treatments.

[Figure 4 ILLUSTRATION OMITTED]

In the first year there were no significant differences in the annual quantities of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N or total N leached between drainage treatments (Table 3). Although the amount of organic N leached in the medium water table treatment was less (P [is less than] 0.05) than the low and high water tables, there was no apparent trend with water table height. Similarly, the annual amount of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leached was slightly greater in the high than the medium water table treatment, and no trend with water table height was evident. In the second year, larger amounts of N were leached, and differences between effluent treatments were more apparent (Table 3). Differences (P [is less than] 0.05) in the amounts of mineral N leached between drainage treatments were recorded in most seasons (Table 4), with greater (P [is less than] 0.05) amounts of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N being leached and less [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N in the high water table than the lower water table treatments, especially during winter.

[TABULAR DATA 4 NOT REPRODUCIBLE IN ASCII]

Discussion

The amount and forms of N leached were affected mainly by the amount of N applied, and to a lesser extent, the managed drainage treatments. The application of additional N in the second year markedly increased the amount of N leached, with the 3-fold higher N application rate causing a 5-10-fold increase in the amount of N leached (depending on drainage treatment). Under the `cut and carry' pasture management system of the second year, the amount of N leached was equivalent to approximately 9% of the added effluent N. The main effect of managing drainage by controlling the water table height was on amounts and forms of mineral N leached in the second year, when higher loading rates of N were applied. In this case, more [NH.sub.4]-N and less [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leached from lysimeters with the high water table. Similar effects of managed drainage on N leaching from applied effluent have been previously reported in field experiments in the USA (Gambrell et al. 1975; Gilliam et al. 1979).

Most previous studies investigating N leaching from soil have focussed mainly on amounts of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N leached. However, in this study, the dominant form of N leached was organic N, which comprised 70-90% of total N leached. Nitrate-N generally accounted for [is less than] 10% of total N leached. The main exception to this was with the low water table (75 cm depth) when only irrigation water was added (i.e. no effluent), with over half of the N leached in [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]-N form.

Reports of organic N forms being leached are fairly rare in the literature (e.g. Cooke et al. 1979), although recently there have been observations of organic N leaching in British soils (D. Scholefield, pers. comm.). In this study, a large proportion of added N was in organic form. The increase in organic N that was leached following the increase in N loading rates indicates that organic N leaching should be measured if accurate rates of N leaching are to be measured or N balances calculated, especially when organic N is applied to soil. MacGregor et al. (1979) reported that the most important form of N leached from DSE in spring was organic, and inorganic N leaching was more important in winter. In the current study, organic N was generally the most important form of N leached in all seasons, even after cessation of effluent irrigation for the milking season.

A large proportion of the N in the effluent was particulate organic N ([is greater than] 0.2 [micro]m), and because of its larger size was presumably less likely to be immediately leached. We suggest that the large amounts of organic N leaching observed in the current study were due to preferential flow of dissolved organic N (c. 33% of total N in effluent) through macropores in the soil, which causes limited interaction of the effluent with the soil matrix. Di et al. (2000) have recently reported that a large proportion (approximately 80%) of total N leached from sandy soils receiving DSE was derived from organic sources of N within the effluent. Our results would support this suggestion. It is well known that in structured soils, the amount of preferential flow of solute in soils depends on the location of the solute within or on aggregates, and the timing and continuity of rainfall or irrigation (e.g. Kluitenberg and Horton 1990; McLay et al. 1991). In this study the N was applied in the irrigation water, which would cause immediate mobility of the N through soil macropores (which are common in the Te Kowhai soil), especially during wetter months when saturated flow at the soil surface occurs. Indeed, leachate which was similar to added effluent in colour was sometimes observed, suggesting that the organic N leached from the soil was due to rapid bypass through the soil. This was even observed occasionally in summer, despite the lower soil moisture contents and irrigation rates less than the saturated hydraulic conductivity of the soil. The Te Kowhai soil has halloysite as the dominant clay mineral, which is prone to cracking during dry periods and would enable macropore flow in summer. Preferential flow would also partly explain the larger amount of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and smaller amount of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leached from the high water table than the medium and low water tables, as the inorganic solutes would not be affected by soil biological processes if they are bypassing the bulk of the soil matrix. However, we have evidence that the smaller amount of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leached from the lysimeters with high water tables is due to a greater amount of denitrification that occurred (P. L. Singleton unpublished data).

Many reports exist on the fate of N added as slurry to soil, especially in the northern hemisphere. It has been suggested that the dominant form of N leached when cattle slurry is applied to grassland soil is [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (Jarvis et al. 1987). It is possible that the lower water content of slurry (90-92% water) compared with DSE ([is greater than] 98% water), and its less frequent application to soils compared with DSE, could result in less preferential flow of effluent organic N into the drainage water than occurs with frequent irrigation of DSE in the southern hemisphere.

Some organic components of the effluent retained in the soil may take several years to decay (Herron and Erhart 1965; Pratt et al. 1973; Smith and Peterson 1982). The proportion of mineralised N derived from effluent components resistant to decay may initially be small, but continued effluent additions increase the soil pool of slowly mineralisable material, and net mineralisation from this source can become important. Various decay series for N mineralisation with time have been suggested for cattle manure (Herron and Erhart 1965; Pratt et al. 1973; Smith and Peterson 1982) and indicate that about 4 years are required before N mineralisation is in equilibrium with manure additions. Consequently, the proportion of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] available for leaching may increase with time, and the full impact of effluent irrigation cannot be known until after several years of irrigation (Herron and Erhart 1965; Pratt et al. 1973).

The largest [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] concentration recorded for any of the effluent treatments or replicates (6.1 mg/L) was below the maximum permissible concentration of 11.3 mg/L as recommended by the WHO for drinking water. Based on [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] concentrations alone, irrigation of DSE to soil would not be expected to result in unacceptably high levels of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] in leachate. However, [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] concentration is only one criterion of water quality. In this soil, a large proportion of the N leached was in forms other than [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], which could potentially be mineralised and contribute to larger [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] concentration in either surface or ground water. Due to the rapid displacement of effluent into drainage water, it is also suggested that other aspects of water quality (e.g. faecal coliforms, biological, and chemical oxygen demand) may also need to be considered in some soils.

The increase in [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] leaching in winter (Table 4) corresponded to the onset of soil drainage and the leaching of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] that had accumulated in the soil over the summer period. Quantities of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] in drainage decreased within 2-3 months following the onset of drainage (Fig. 4) even though drainage volumes continued to be high. This indicated the decreased availability of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] for leaching because of removal by leaching, decreased nitrification in colder and wetter conditions, or denitrification.

Conclusions

This study has shown that the application of DSE to soils affects the amounts and forms of N leached from soils, and that the amount leached is proportional (slightly [is less than] 10%) to the amount applied. Of particular interest is the finding that organic N comprises a large proportion of the total amount of N leached. We suggest that in structured soils containing macropores, the forms of N leached beyond the main rooting depth of plants may reflect the nature of the added effluent. Therefore, inorganic N should not be the only aspect of water quality in leachate that scientists consider when studying the effects of irrigating effluent, with large N loadings, onto soils.

The use of managed drainage, to control water table depth in soil, had limited effect on the amount and forms of N leached from the lysimeters, probably due to the rapid leaching of DSE through soil macropores.

Acknowledgments

We thank the Foundation for Research, Science and Technology for their contribution to the funding of this research, and Landcare Research, NIWA, and Environment Waikato for laboratory support. The technical support of D. Maudsley and S. Park is also acknowledged.

References

Adams JA (1981) Influence of land use on nitrate movement through soil profiles in Paparua County. Unpublished Report, Soil Science Department, Lincoln College, Canterbury, New Zealand.

Blakemore LC, Searle PL, Daly BK (1987) Methods for chemical analysis of soils. NZ Soil Bureau Scientific Report No. 80, DSIR, Lower Hutt, New Zealand.

Cameron KC, Smith NP, McLay CDA, Fraser PM, McPherson R J, Harrison DF, Harbottle P (1992) Lysimeters without edge flow: an improved design and sampling procedure. Soil Science Society of America Journal 56, 1625-1628.

Cooke JG, Tillman RW, MacGregor AN, Syers JK (1979) Studies in the chemical and microbiological characteristics of soil-filtered dairy shed effluent. Progress in Water Technology 11, 19-31.

Deal SC, Gilliam JW, Skaggs RW, Konyha KD (1986) Prediction of nitrogen and phosphorous losses as related to agricultural drainage system design. Agriculture, Ecosystems and Environment 18, 37-51.

Di HJ, Cameron KC, Moore S, Smith NP (1998) Nitrate leaching and pasture yields following the application of dairy shed effluent or ammonium fertilizer under spray or flood irrigation: results of a lysimeter study. Soil Use and Management 14, 209-214.

Di HJ, Cameron KC, Moore S, Smith NP (2000) Contributions to nitrogen leaching and pasture uptake by autumn-applied dairy effluent and ammonium fertiliser labeled with [sup.15]N isotope. Plant and Soil 210, 189-198.

Gambrell RP, Gilliam JW, Weed SB (1975) Nitrogen losses from soils of the North Carolina coastal plain. Journal of Environmental Quality 4, 317-323.

Gilliam JW, Skaggs RW, Weed SB (1979) Drainage control to diminish nitrate loss from agricultural fields. Journal of Environmental Quality 8, 137-142.

Herron GM, Erhart AB (1965) Value of manure on an irrigated calcareous soil. Soil Science Society of America Proceedings 29, 278-281.

Hewitt AE (1992) `New Zealand soil classification'. DSIR Land Resources Scientific Report No. 19, Lower Hutt, New Zealand.

Jarvis SC, Sherwood M, Steenvoorden JAHM (1987) Nitrogen losses from animal manures: from grazed pastures and applied slurry. In `Animal manure on grassland and fodder crops'. (Ed. HG van der Meer) pp. 195-211. (Martinus Nijhoff: Netherlands)

Joe EN (1986) Soil water characterisation studies of 6 soils in the Waikato District, New Zealand. NZ Soil Bureau SWAMP data sheets 1984 [1-6], DSIR, New Zealand.

Kliewer BA, Gilliam JW (1995) Water table management effects on denitrification and nitrous oxide formation. Soil Science Society of America Journal 59, 1694-1701.

Kluitenberg G, Horton R (1990) Effect of solute application method on preferential transport of solutes in soil. Geoderma 46, 283-296.

Light C (1995) Taupo land treatment and disposal system. In `Land treatment systems: Design and monitoring'. Proceedings of the NZ Land Treatment Collective, No. t3, November 1995. (Eds S. Pandey, J. M. Carnus)

Longhurst RD, Roberts AHC, O'Connor MB (2000) Farm dairy effluent: Review of published data on physical and chemical characteristics. NZ Journal of Agricultural Research 43,

MacGregor AN, Stout JD, Jackson RJ (1979) Quality of drainage water from pasture treated with dairy shed effluent. Progress in Water Technology 11, 11-17.

McLay CDA, Cameron KC, McLaren RG (1991) Effect of time of application and continuity of rainfall on nutrient leaching losses in undisturbed soil. Australian Journal of Soil Research 29, 1-9.

Pratt PF, Broadbent FE, Martin JP (1973) Using organic wastes as nitrogen fertilizer. Californian Agriculture 27, 10-13.

Rate AW, Cameron KC (1992) The fate of nitrogen in piggery waste applied to a shallow stony pasture soil. In `The use of wastes and byproducts as fertilizers and soil amendments for pastures and crop'. (Eds PEH Gregg, LD Curry) pp. 314-326. (Massey University: Palmerston North, New Zealand)

Rolston DE, Sharpley AN, Toy DW, Broadbent FE (1982) Field measurement of denitrification. II. Rates during irrigation cycles. Soil Science Society of America Journal 46, 289-296.

Sexstone AJ, Parkin TB, Tiedje JM (1985) Temporal response of soil denitrification rates to rainfall and irrigation. Soil Science Society of America Journal 49, 99-103.

Sherwood M (1986) Nitrate leaching following application of slurry and urine to field plots. In `Efficient land use of sludge and manure'. (Eds AD Kofed, JH Williams, P L'Hermite) pp. 150-157. (Elsevier Applied Science: London)

Singleton PL (1991) Soils of Ruakura--a window on the Waikato. DSIR Land Resources Scientific Report No. 5, Lower Hutt, New Zealand.

Skaggs RW, Fausey NR, Nolte BH (1981) Water management evaluation for north central Ohio. Transactions of American Society of Agricultural Engineers 24, 922-928.

Smith KA, Chambers BJ (1993) Utilising the nitrogen content of organic manures on farms-- problems and practical solutions. Soil Use and Management 9, 105-112.

Smith JH, Peterson JR (1982) Recycling of nitrogen through land application of agricultural, food processing, and municipal wastes. In `Nitrogen in agricultural soils'. (Ed. FJ Stevenson) pp. 791-832. (American Society of Agronomy Inc.: Madison, WI)

Soil Survey Staff (1990) `Keys to soil taxonomy'. Soil Management Support Services Technical Monograph No. 19. (Virginia Polytechnic Institute and State University: VA)

Steenvoorden JHAM, Fonck H, Oosterom HP (1986) Losses of nitrogen from intensive grassland systems by leaching and surface runoff. In `Nitrogen fluxes in intensive grassland systems'. (Eds HG van der Meer, JC Ryden, GC Ennik) pp. 85-99. (Martinus Nijhoff Publishers: Dordrecht, The Netherlands)

Unwin RJ (1986) Leaching of nitrate after application of organic manures: lysimeter studies. In `Efficient land use of sludge and manure'. (Eds AD Kofed, JH Williams, P L'Hermite) pp. 158-167. (Elsevier Applied Science: London)

Unwin RJ, Pain BF, Whinham WN (1986) The effect of rate and time of application of nitrogen in cow slurry on grass cut for silage. Agricultural Wastes 15, 253-268.

Unwin RJ, Shepherd MA, Smith KA (1991) Controls on manure and sludge applications to limit nitrate leaching. Does the evidence justify the restrictions which are being proposed? In `Treatment and use of sewage sludge and liquid agricultural wastes'. (Ed. P L'Hermite) pp. 261-270. (Elsevier Applied Science: London)

Vanderholm DH (1984) Properties of agricultural wastes. Agricultural Waste Manual Report No. 32, NZAEI, Lincoln College, Canterbury, New Zealand.

Vetter H, Steffens G (1981) Leaching of nitrogen after leaching of slurry. In `Nitrogen losses and surface run-off from landspreading of manures'. (Ed. JC Brogan) pp. 251-269. (Martinus Nijhoff/Dr Junk: The Hague)

Williamson JC, Taylor MD, Torrens RS, Vojvodic-Vukovic M (1998) Reducing nitrogen leaching from dairy farm effluent-irrigated pasture using dicyandiamide: a lysimeter study. Agriculture, Ecosystems and Environment 69, 81-88.

Manuscript received 28 February 2000, accepted 14 July 2000

P. L. Singleton,(A), C. D. A McLay(B), G. F. Barkle(C)

(A) AgResearch, Ruakura Research Centre, Private Bag 3123, Hamilton, New Zealand.

(B) Department of Earth Sciences, University of Waikato, Private Bag 3105, Hamilton, New Zealand.

(C) Lincoln Environmental, Private Bag 3062, Hamilton, New Zealand.
Year 1        Year 2 (amended)
N component           Mean     Range      Mean    Range

Total N                90      44-186      233   108-412
Organic N              67      34-125      181    90-337
Dissolved organic N   n.d.      n.d.       35     1-92
[MATHEMATICAL          23       5-70       42     16-87
  EXPRESSION NOT
  REPRODUCIBLE IN
  ASCII]-N
[MATHEMATICAL         0.05   <0.05-0.45   <0.05   <0.05
  EXPRESSION NOT
  REPRODUCIBLE IN
  ASCII]-N


Drainage treatment         Irrigation   Rainfall
           Low       Medium      High        (mm)        (mm)

Year 1   829 (56)   789 (53)   807 (63)      586         1210
Year 2   799 (60)   777 (23)   796 (58)      697         1228
Gale Copyright: Copyright 2001 Gale, Cengage Learning. All rights reserved.