Previous PageTable Of ContentsNext Page

Nitrous oxide emissions in medium rainfall cropping environments: merely interesting, or actually important?

Patricia Hill1, Roger Armstrong1, Debra Partington2, Rob Harris2 and Ashley Wallace1

1Department of Primary Industries, Horsham Victoria 3400 Email
Department of Primary Industries, Hamilton Victoria 3300


There have been limited studies where measurements of nitrous oxide (N2O) have been made in the medium rainfall cropping environments of Australia, or that consider how grower management may influence emissions. This is particularly important given that N2O emissions represent an environmental threat and a loss of N from the farming system, and as such could significantly affect nitrogen use efficiency and financial viability in medium rainfall/low input systems. We tested the hypothesis that in medium rainfall environments particular management strategies that alter soil N, C and water status will alter N2O emissions profiles. To test this hypothesis, an experiment comparing the type of N used to supply a wheat crop (legume v’s cereal, plus/minus N fertiliser application) was established near Horsham, Victoria. N2O emissions were measured on a regular basis over the growing season using static in situ gas chambers. Preliminary results from a single cropping season suggest that while overall emissions are low, current nitrogen management strategies offer potential for reducing N2O emissions in the Wimmera, with significant treatment effects observed in the seven days following urea application to wheat, where emissions were up to four times higher than where urea was not applied, and at least double those where N was supplied by a pre-phase legume. These findings will underpin further research needed to provide evidential basis for identifying whether, and how, management strategy can enhance overall N use efficiency in medium rainfall cropping systems.

Key Words

Nitrous Oxide emissions, dryland cropping systems, static chambers, nitrogen source


Studies where N2O emissions have been measured in dryland medium rainfall cropping systems in Australia are limited, but indicate that emissions are generally low in these environments (e.g. Barton et al, 2008; Officer et al, 2008; Schwenke et al, 2010). However given the significance of N2O as a greenhouse gas, it is important to quantify emissions under different management scenarios to determine whether farmers are able to manage emissions from soils. As the four key factors affecting nitrous oxide (N2O) emissions from soils include soil water filled pore space (WFPS), soil N availability, soil temperature and soluble soil carbon (Dalal et al, 2003), adoption of management strategies that may affect one or more of these factors might be expected to result in a consequent effect on N2O emissions. However, both Barton et al (2008) and Officer et al (2008) have found that increasing the soil N availability by applying locally representative rates of N fertiliser compared with no N application had minimal effect on annual N2O emissions during the crop growth season. However, neither of these studies measured emissions where alternative sources of N were used (such as legumes), which may alter the seasonal emission profile. Schwenke et al (2010) observed increases in emissions in N fertilised canola compared with non-fertilised chickpea, but did not include a non-fertilised comparison to establish whether the N fertiliser application per se, or soil N availability was the main factor contributing to N losses in that study. Additionally, the relationship between soil WFPS and emissions remains largely unreported for these environments.

This study aimed to investigate the relationship between WFPS and N2O emissions, and to determine whether applying nitrogen fertiliser to a cereal crop grown the year after either a non-legume or a legume phase would affect nitrous oxide emissions in medium rainfall cropping systems. It was hypothesised that emissions would be low in these systems, but that applying N fertiliser would increase N2O emissions in the cereal phase. It was also anticipated that wheat grown after a legume phase would emit more N2O than wheat grown after a cereal phase when N fertiliser was not applied.


The experiment was established on June 28, 2010, on a sodic cracking clay soil (Australian soil classification: Vertosol) on a farm 20 km south of Horsham (36.790367 S, 142.386975 E) with average annual rainfall of 430 mm. The experiment consisted of a randomised block design. Treatments in 2010 were wheat or vetch, with no N applied. On October 21, when vetch was at 50% flowering, the vetch plots were green manured – that is, a disk implement was used to incorporate the vetch into the soil to a depth of approximately 10 cm. The wheat plots were harvested on January 4, 2011. Initial gas and soil samples were collected on May 4, 2011and on May 23, 2011, all plots were sown to wheat (c.v. Axe) with 45 kg of MAP at sowing. No further fertiliser was applied for the season, and all other agronomic management was undertaken by the farmer. Gas and soil samples were collected 1 and 2 weeks after sowing, and then approximately monthly until August 16. On that day when wheat was at GS31, the wheat treatment plots were split and urea was applied at the rate of 100 kg / ha urea (46 kg N / ha) to half of each wheat plot. Gas samples were collected within two hours prior to urea application and within two hours post urea application, and again 1, 2, 7 and 14 days after urea application and thereafter monthly until the final post-harvest collection on December 14, 2012. Plots were harvested on December 7, 2011. Daily air and soil temperature and rainfall were recorded by a Magpie weather station. Samples were also collected for soil moisture analysis at sowing, anthesis and crop maturity and for mineral N at sowing (NH4+ and NO3-) after extraction in 2M KCl within 24 hours of collection (data not presented). Extracts were frozen until analysed. Plant factors including emergence, biomass (at anthesis and crop maturity) and final yield were also measured (data not presented).

Nitrous oxide emissions were measured using PVC static chambers (325 mm high x 250 mm diameter) fitted with gas extraction septa. Twenty mL gas samples were withdrawn from the chambers using a syringe, injected into pre-flushed and pre-evacuated, vacuumed 12 mL LABCO exetainers, and the gas sample analysed by gas chromatography. On each gas sampling occasion, lids were attached to the chambers, and an initial gas sample was collected immediately. Subsequent samples were taken every twenty minutes over a one hour period, resulting in 4 separate gas samples taken from each chamber. Air temperature was recorded from one replicate chamber from each plot each time an individual gas sample was collected, and soil temperature was measured between the two replicate chambers at the start and end of each one hour gas collection period for each pair of chambers in each plot. On each sampling occasion, soil samples were also collected from the 0-5 cm, 0-10 cm, and 10-20 cm depths. These samples were split immediately on return to the laboratory, and one half was frozen. The other half was sub-sampled for gravimetric moisture analysis and for mineral N analysis. The data from the four gas samples for each gas sampling occasion were converted into a chamber flux using the method of Harris et al (2012) by converting each sample to gas density. Logarithmic (base 10) transformations were calculated to normalise the N2O flux data before Residual Maximum Likelihood (REML) analyses to identify treatment differences.

Results and discussion

Nitrous oxide emissions were generally below 2 g N2O-N / ha / day for the entire of the measurement season, with peaks of up to 9.5 g N20-N/ha/ day (Figure 1). These low emissions are in the range of those observed by Officer et al (2008) for soils in the Wimmera, who reported fluxes generally below 4 g N2O-N/ha/day and peaks of up to 13 g N2O-N/ha/day. Such low emissions might be expected when soil factors (specifically, WFPS, nitrogen and carbon supply and soil temperature) are not in the range likely to lead to high N2O production, or when high levels of each of these factors do not temporally coincide so that denitrification processes dominate. REML analysis showed significant treatment effects, but only in the week following the urea application. Urea is rapidly hydrolysed and ammonified to NH4+ in wet soils (usually within a week) (Myers, 1975; Black et al, 1987), whereas nitrification can take much longer. This suggests that N2O was primarily produced via nitrification of NH4+ in the weeks following urea application, and denitrification is unlikely to be responsible for emissions, even though the soil WFPS also peaked at this time.

The application of 100kg / ha of urea to the cereal on cereal treatments resulted in a significant increase in N2O emissions compared with the cereal on vetch and cereal –urea treatments, with predicted means for the cereal + urea treatment being significantly higher than the other two treatments within 2 hours of application, and then 48 hours and 7 days after application (Table 1). Importantly, for all treatments, peak seasonal fluxes occurred during the same time period, indicating that whilst urea application enhanced fluxes, in-soil processes were also leading to peaks in “background” emissions where no urea was applied.

Figure 1. Mean temporal changes in (from top to bottom, left): Water Filled Pore Space (WFPS) (0-5 cm depth), Soil mineral N (0-5 cm depth), and (from top to bottom, right) N2O emissions, and soil temperature and rainfall at the N source experiment (Taylor’s Lake, Victoria)

Table 1. Table of REML predicted means for N source experiment. Highlighted values indicate days on which significant treatment effects were observed, with values followed by different letters being significantly different at the 10% level (s.e.d. = 1.8876)


Cereal -urea

Cereal + urea






















16/08/2011 (pre app’n)




16/08/2011 (post app’n)



2.699 a






2.059 a

9.663 b

4.971 a


2.336 a

7.138 b

2.337 a

















The WFPS (0-5 cm) rarely exceeded 60 %, and was generally below 50 % throughout the season. Nitrification, rather than denitrification, therefore seems to be the likely process by which N2O emissions are produced in these soils, which would explain the generally low emissions throughout the season. However, when soil WFPS was above 60 %, small corresponding emission peaks were observed, but only immediately after addition of urea during a rainfall event (Figure 1). This would correspond to a temporal coincidence of high soil WFPS and high soil N supply, resulting in short-term enhancement of N2O emissions, and denitrification processes may have partially contributed to this peak.

Stevens et al (1998) found that large increases in N2O emissions were not observed until at least 80 % WFPS, although emissions did increase at lower WFPS (i.e. 60 %) when the soil pH was at least neutral or higher, as is the case with the soils in the present study. Sehy et al (2003) also observed a lower threshold of 60 % WFPS, but the effect of WFPS was found to only be significant when neither soil C nor soil N were limiting for microbial denitrification. However mineral N rarely exceeded 10 kg N/ha in this study (Figure 1), which is likely to be limiting N2O production. Additionally, when soil WFPS peaked around July, this also coincided with the season’s lowest soil temperature observations – usually less than 10C at that time. Hence it can be seen that whilst each of the factors that are known to affect N2O emissions were occasionally above the reported thresholds for causing enhanced N2O fluxes, the thresholds for all factors rarely coincided temporally. Indeed, in the Mediterranean environment of the Wimmera district of Victoria, this is likely to be the case more often than not, as the winter-dominant rainfall (and hence peak WFPS period) usually coincides with the lowest soil temperatures in mid-winter.


Nitrous oxide emissions from a wheat crop during 2011 were low for the entire season, with short-lived emission peaks immediately after urea application. These low emissions are likely to be a consequence of low soil mineral N availability and low soil WFPS, and are likely primarily driven by nitrification processes, with possible contributions by denitrification. Additional soil C data will allow further interpretation of the results in terms of in-soil processes. The preliminary data suggests that N2O emissions from these environments are of minor consequence for greenhouse gas mitigation or for on-farm economics, but that management strategies may have potential for reducing emissions in some instances.


Barton, L., Kiese, R., Gatter, D., Butterbach-Bahl, K., Buck, R., Hinz, C., and Murphy, D.V. (2008). Nitrous oxide emissions from a cropped soil in a semi-arid climate. Global Change Biology, 14, 177-192.

Black, A.S., Sherlock, R.R., and Smith, N.P. (1987). Effect of timing of simulated rainfall on ammonia volatilization from urea, applied to soil of varying moisture content. European Journal of Soil Science, 38(4) 679-687.

Dalal, R.C., Wnag, W., Robertson, G.P., and Parton W.J. (2003) Nitrous oxide emission from Australian agricultural lands and mitigation options: a review. Australian Journal of Soil Research 41(2) 165 – 195

Harris, R. Officer, S. Hill, P., Armstrong, R. and Partington, D. (2012). Can agronomy mitigate N2O loss from high rainfall cropping systems? Proceedings, 16th Australian Agronomy Conference, October 14-18 2012, Armidale, Australia

Myers, R.J.K. (1975) Temperature effects on ammonification and nitrification in a tropical soil. Soil Biology and Biochemistry, 7(2) 83–86

Officer, S.J., Phillips, F., Armstrong, R., and Kelly, K. (2008) Nitrous oxide emissions from dry-land wheat in south-eastern Australia. "Global Issues. Paddock Action." Edited by M. Unkovich. Proceedings of 14th Agronomy Conference 2008, 21-25 September 2008, Adelaide, Australia.

Schwenke, G., Haigh, B., McMullen, G., and Herridge, D. et al, (2010) Soil nitrous oxide emissions under dryland N-fertiliser canola and N2-fixing chickpea in the northern grains region, Australia. World Congress of Soil Science, Soil Solutions for a Changing World, August 1-6 2010, Brisbane, Australia.

Sehy, U., Ruser, R., and Munch, J.C. (2003) Nitrous oxide fluxes from maize fields: relationship to yield, site-specific fertilization, and soil conditions. Agriculture, Ecosystems & Environment, 99(1–3), 97–111

Stevens, R.J., Lauglin, R.J., Maone, J.P. (1998) Soil pH affects the processes reducing nitrate to nitrous oxide and di-nitrogen. Soil Biology and Biochemistry, 30 (8–9), 1119–1126

Previous PageTop Of PageNext Page