Abstract

Key Words

Water filled pore space, compaction, mineral N.

Introduction

The application of high rates of N fertiliser to vegetable crops poses a significant environmental risk through N leaching and runoff to water bodies and emissions of N₂O (Greenwood et al. 1989). In New Zealand, N fertilizer rates for intensive potato production are typically 300 to 450 kg N/ha. High rates of nitrate leaching have been measured from these crops (Francis et al. 2003), but N₂O emissions have not been measured. In addition to high N application rates, potato crops are often irrigated. This may further enhance N₂O production by stimulating nitrification and denitrification (Dobbie and Smith 2003; Linn and Doran 1984).

The rate of N₂O emission increases as both N fertiliser rate (Bouwman 1996) and water-filled porosity increases. High values of water-filled pore space (WFPS) has been identified as a key indicator of N₂O emissions since moisture content will affect diffusion of oxygen through the soil matrix (Linn and Doran 1984). Largest N₂O emissions have been observed when the WFPS is high, between 70 and 90% (Dobbie et al. 1999) when its production is dominated by denitrification. Below 60% WFPS nitrification, an aerobic process, is the dominant process producing N₂O (Linn and Doran 1984).

Other management effects, such as soil compaction by tractor traffic in potato crops (Flessa et al. 2002; Ruser et al. 1998), other row crops (Dobbie et al. 1999; Ruser et al. 2001; Ryden and Lund 1980), and arable crops (Ball et al. 1999b; Hansen et al. 1993) have also been shown to greatly enhance N₂O emissions.

Tillage practices can also affect emissions of N₂O by affecting soil aeration through changes to soil gas diffusivity and the water and air holding capacity of soil pores (Ball et al. 1999a; Yamulki and Jarvis 2002).

This field experiment examined the influence of N fertiliser rate, tractor compaction in furrows and WFPS on N₂O emissions from an irrigated potato crop.

Materials and methods

The experiment was established in late spring 2002 on a well-drained soil (Typic Haplustepts). After cultivation, fertiliser (as calcium ammonium nitrate) was broadcast on to each plot at a rate of 0, 225 or 450 kg N/ha. In the previous year cereal silage had been grown on the area. The experiment was a split-plot, randomised block design with five replicates. Plots were 5 m long by 2.3 m wide (3 rows). Potatoes (cv. Nadine) were planted on 26 November at 55 cm spacings in rows 78 cm apart and then all plots were machine ridged such that each plot contained both compacted (by tractors) and uncompacted furrows. Fertiliser application, potato planting and ridging occurred on the same day.

Mineral N

Soil mineral N was measured at each sampling position in each plot on seven occasions during December to February. In the furrow the samples were taken from 0-10 cm and in the ridge from 0-20 cm. Soil mineral N was extracted from soil samples using 2M potassium chloride and analysed using a Rapid Flow Analyser (Astoria-Pacific Inc., Clackamas, Oregon).

N₂O measurements

N₂O fluxes were determined using a closed chamber technique (Hutchinson and Mosier 1981). Chambers (10 cm depth) were made from 30 cm diameter PVC pipe with welded lids. Bases (15 cm depth), made from the same diameter PVC pipe, were inserted into the soil (5 cm depth) to enable gas fluxes to be measured at each of three sampling positions: the ridge (between tubers), the compacted furrow and the uncompacted furrow (Fig. 1). A water-filled channel at the top of each base produced a gas tight seal with the chamber during measurements. Within each subplot, N₂O concentrations were measured in the chamber headspace at three time intervals (0, 20 and 40 minutes) after chamber closure. Gas samples were analysed using a gas chromatograph (GC-17A, Shimadzu Corporation, Kyoto) fitted with a ⁶³Ni-Electron capture detector. Measurements were made on 24 occasions from 27 November to 8 March.

Other soil measurements

Soil bulk density was measured at each sampling position in the nil fertiliser treatment plots during December. The samples were taken at the same depths as for mineral N. Volumetric soil moisture content (0-20 cm) was measured daily in each plot using TDR probes (Model -Trase, Soil Moisture Equipment Corp., Santa Barbara, California). WFPS was calculated from measured bulk density and water content and an assumed particle density (2.65 g/cm³): WFPS = volumetric moisture content/(1-(soil bulk density/soil particle density)). Soil temperature (at 5 cm depth) was recorded hourly using thermocouples. Daily rainfall and spray irrigation (applied five times) were measured throughout the experiment.

Statistical analyses

The effects of fertiliser and position on N₂O flux and mineral N were analysed using split-plot analysis of variance. The N₂O flux and mineral N data required log transformation to make the variance more homogeneous; the results presented in this paper have been back transformed. A mixed model fitted using REML (Residual Maximum Likelihood) analysis tested for the effects of fertiliser rate, sampling position, WFPS and mineral N on N₂O emissions. Comparison of the N₂O flux and mineral N means between fertiliser treatments and position is made using the least significant ratio (LSR). The LSR is the smallest ratio between two back-transformed means (largest mean/smallest mean) such that the larger mean is significantly greater than the smallest mean. For all the statistical analyses a significance level of 5% was used to test for treatment effects. Analyses were performed using the GenStat (version 7) software package.

Figure 1. PVC cylinders in the compacted furrow, ridge and uncompacted furrow for measurement of N₂O surface flux.

Results and discussion

Following planting and ridging, the mean bulk density for the ridges, uncompacted furrows and compacted furrows were 0.96, 1.08 and 1.18 g/cm³, respectively. Mineral N contents before fertiliser application were 70 to 90 kg N/ha (data not shown). Following fertiliser application, most of the fertiliser N was in the ridges and least in the furrows (Fig 2a). The mineral N contents in the potato ridges were approximately 280 and 740 kg N/ha for the 225 and 450 kg N/ha treatments respectively.

Rainfall received during the experiment (65 mm) was supplemented with 285 mm of irrigation (Fig. 2b). The addition of rainfall or irrigation increased the WFPS in all instances. Throughout the experiment, WFPS varied greatly with sampling position in the order: compacted furrow > uncompacted furrow > ridge. The compacted furrow had a maximum measured WFPS of about 75%, whereas the ridge had a maximum WFPS of less than 50%, even after the application of a large amount (85 mm) of irrigation.

Figure 2. Results from the compacted furrow, ridge and uncompacted furrow in the 450 kg N/ ha treatment for: a) ammonium (red solid bars) and nitrate (hatched bar) concentrations; b) WFPS (black line), rainfall (red solid bars) and irrigation amounts (hatched bars); and c) N₂O surface flux (red shading).

Sampling position had a strong effect (P < 0.001) on N₂O emissions in the order: compacted furrow > ridge > uncompacted furrow (Fig 2c). This was despite the location of most of the mineral N in the ridges. This pattern has been found in other potato crops, where the highest N₂O emissions have been observed from compacted soil (Ruser et al. 1998).

The highest N₂O emissions occurred when WFPS was high (>55%), following rain or irrigation (Fig. 3). The lowest N₂O emissions occurred from the uncompacted furrows that had lower WFPS than the compacted furrows and relatively low mineral N concentrations.

Figure 3. Relationship between WFPS and N₂O surface flux from the (a) compacted furrows, (b) uncompacted furrows and (c) ridges

The highest hourly N₂O fluxes measured in the compacted furrow are most likely to have been produced from denitrification when WFPS was highest following rainfall or irrigation. Denitrification is expected to be the dominant process for producing N₂O when WFPS is greater than 70% (Dobbie et al. 1999). Typically, N₂O emissions from wet soils are also much greater than from dry soils. Nitrification is more likely to be the dominant N₂O production process in the ridge and uncompacted furrow as maximum nitrification rates occur when WFPS reaches approximately 60% (Linn and Doran 1984). Increases in WFPS due to rainfall and irrigation events may therefore have differentially stimulated nitrification (in the ridge and uncompacted furrows) and denitrification in the compacted furrow (Fig 2b and c).

Figure 4. Mean N₂O emissions from tractor-compacted furrows (circles), uncompacted furrows (triangles) and potato ridges (squares) for plots with (a) no added N fertiliser, (b) 225 kg N/ha and (c) 450 kg N/ha. Data are means of five replicates. The error bas is the LSD (P < 0.05). Note the log scale for N₂O flux.

N₂O flux was significantly affected by sampling position and time of sampling (P <0.001). In contrast, fertiliser rate had no significant effect on N₂O fluxes (Fig. 4 and Table 1). Although there was a non-significant trend towards greater N₂O emissions from the fertilised plots, calculated cumulative emissions during the four month experiment did not differ between the 225 and 450 kg N/ha treatments (Table 1). Cumulative N₂O emissions comprised only 0.6 and 0.3% of the applied fertiliser in the 225 and 450 kg N/ha treatments respectively.

Table 1. Calculated cumulative N₂O surface flux emissions during the four month experiment. The paddock average values are calculated from the surface area occupied by each sampling location. Ridges occupied approximately 50% of the soil area, whereas the tractor compacted and uncompacted furrows each occupied 25% of the area.

	Cumulative N2O surface flux (kg N/ha)
Fertiliser rate (kg N/ha)	Uncompacted furrow	Ridge	Compacted furrow	Paddock average
0	0.39	0.85	2.42	1.01
225	0.43	1.05	2.88	1.20
450	0.46	1.20	2.48	1.15

For comparison between different positions LSR = 3.36 (23 d.f.)

For comparisons within fertiliser levels, LSR = 2.81 (24 d.f.).

While we found that there was no significant effect of fertiliser level on the N₂O emissions, there was a significant overall effect (P < 0.05) of mineral N content on N₂O emissions over time. The mineral N data tended to be highly variable, particularly in the ridges. There were no differences between the amounts of ammonium and nitrate between the 225 and 450 kg N/ha treatments, except in early December (Fig. 5). Mineral N contents in compacted furrows were greater than in the uncompacted furrows. It is not clear why this has occurred, but it may possibly be related to the mechanical action of the tractor wheels and potato ridge moulding that may have affected the movement of broadcast fertiliser.

Figure 5. Nitrate and ammonium contents in the compacted furrows (a and d), uncompacted furrows (b and e) and ridges (c and f) for 0 (circles), 225 (triangles) and 450 kg N/ha (squares) fertiliser plots. Data are means for five replicates. The larger error bar (red) in Figures (a) to (c) is the LSD for comparison of ammonium contents between the 0, 225 and 450 kg N/ha fertiliser plots and the smaller error bar (black) is the LSD for comparison of ammonium contents between the 225 and 450 kg N/ha plots. The LSD in figures (d) to (f) is for comparison of nitrate from plots with different fertiliser levels.

Our estimated N₂O emissions during the growth of the crop were similar, if not slightly lower, than those found in other studies. In Scotland, mean N₂O emissions were between 3 and 4.7 kg N/ha (Dobbie et al. 1999), and in Germany the mean emission ranged from 1.6 to 2 kg N/ha (Flessa et al. 2002) and 2.4 to 4.1 kg N/ha (Ruser et al. 2001). The total N₂O emissions we estimated from the tractor compacted furrows (2.5 to 2.9 kg N/ha) were similar to those found over two growing seasons (2.5 and 2.67 kg N/ha) by Flessa et al.(2002), and 3.2 to 5.5 kg N/ha (Ruser et al. 2001).

While we only measured N₂O emissions during the growing season, there may be further N₂O losses associated with the following soil management of the crop. Ruser et al.(2001) estimated mean N₂O losses over winter were 2.7 to 3.3 kg N/ha. These losses were attributed to high soil nitrate contents after harvest and post harvest N mineralisation that would be affected by residue management and post harvest tillage (Flessa et al. 2002).

N₂O emissions may be reduced through the use of one or several of the following management practices:

• Reduction of soil compaction by avoiding the cultivation or trafficking of soils when wet
• Application of irrigation at application rates and in small amounts to avoid high WFPS
• Splitting the application of N fertiliser to better match crop demands (Freney 1997)
• The use of nitrification inhibitors to reduce soil nitrate concentrations (Freney 1997)

Conclusion

The influence of tractor compaction on N₂O emissions was more important than the rate of applied N fertiliser. Both tractor compaction and irrigation affected N₂O emissions by increasing WFPS, leading to enhanced denitrification. However, the effect of irrigation was less important in the well aerated, uncompacted fertilised ridges and furrows. Consequently, management that limits compaction in row crops will help reduce N₂O emissions.

Acknowledgements

This research was funded by the Foundation for Research Science and Technology.

References

Ball BC, Parker JP, Scott A (1999a) Soil and residue management effects on cropping conditions and nitrous oxide fluxes under controlled traffic in Scotland 2. Nitrous oxide, soil N status and weather. Soil & Tillage Research 52, 191-201.

Ball BC, Scott A, Parker JP (1999b) Field N₂O, CO₂ and CH₄ fluxes in relation to tillage, compaction and soil quality in Scotland. Soil & Tillage Research 53, 29-39.

Bouwman AF (1996) Direct emissions of nitrous oxide from agricultural soils. Nutrient Cycling in Agroecosystems 46, 53-70.

Dobbie KE, McTaggart IP, Smith KA (1999) Nitrous oxide emissions from intensive agricultural systems: Variations between crops and seasons, key driving variables, and mean emission factors. Journal of Geophysical Research-Atmospheres 104, 26891-26899.

Dobbie KE, Smith KA (2003) Nitrous oxide emission factors for agricultural soils in Great Britain: the impact of soil water-filled pore space and other controlling variables. Global Change Biology 9, 204-218.

Flessa H, Ruser R, Schilling R, Loftfield N, Munch JC, Kaiser EA, Beese F (2002) N₂O and CH₄ fluxes in potato fields: automated measurement, management effects and temporal variation. Geoderma 105, 307-325.

Francis GS, Trimmer LA, Tregurtha CS, Williams PH, Butler RC (2003) Winter nitrate leaching losses from three land uses in the Pukekohe area of New Zealand. New Zealand Journal of Agricultural Research 46, 215-224.

Freney JR (1997) Strategies to reduce gaseous emissions of nitrogen from irrigated agriculture. Nutrient Cycling in Agroecosystems 48, 155-160.

Greenwood DJ, Stone DA, Burns IG, Rowse HR (1989) Improving the efficiency of N-fertiliser use for vegetable crops. Journal of the Science of Food and Agriculture 48, 125-127.

Hansen S, Maehlum JE, Bakken LR (1993) N₂O and CH₄ fluxes in soil influenced by fertilization and tractor traffic. Soil Biology and Biochemistry 25, 621-630.

Hutchinson GL, Mosier AR (1981) Improved soil cover method for field measurement of nitrous oxide fluxes. Soil Science Society of America Journal 45, 311-316.

Linn DM, Doran JW (1984) Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and non-tilled soils. Soil Science Society of America Journal 48, 1267-1272.

Ruser R, Flessa H, Schilling R, Beese F, Munch JC (2001) Effect of crop-specific field management and N fertilization on N2O emissions from a fine-loamy soil. Nutrient Cycling in Agroecosystems 59, 177-191.

Ruser R, Flessa H, Schilling R, Steindl H, Beese F (1998) Soil compaction and fertilization effects on nitrous oxide and methane fluxes in potato fields. Soil Science Society of America Journal 62, 1587-1595.

Ryden JC, Lund LJ (1980) Nature and extent of directly measured denitrification losses from some irrigated vegetable crop production units. In 'Soil Science Society of America Journal')

Yamulki S, Jarvis SC (2002) Short-term effects of tillage and compaction on nitrous oxide, nitric oxide, nitrogen dioxide, methane and carbon dioxide fluxes from grassland. Biology and Fertility of Soils 36, 224-231.