• In-season soil nitrogen (N) mineralisation can be a significant source of N to crops, but has not been well characterised for WA cropping soils. Improving the synchrony of soil N supply and crop demand could result in changes to N fertiliser inputs and better efficiency of use. 
• Loss of N fertiliser to the atmosphere via ammonia (NH3) volatilization may be as high as 29% from WA, however findings are restricted to two field-based studies using granular urea. Nitrogen leaching from WA cropping soils can also be significant during the growing season (14 –72 kg N/ha), however measurements have been confined to deep sands. 
• Further research is needed quantify soil N mineralisation and NH3 volatilisation rates under current WA farming practises, including incorporating findings into decision support tools.

Aims

To describe the supply and loss of N from soils in WA dryland cropping systems. Specifically we investigated soil N supply via net soil N mineralisation, plus N loss pathways including NH3 volatilisation and other gaseous losses, and N leaching.

Method

We reviewed the scientific literature and collated soil N mineralisation, NH3 volatilisation, N2O and N2 emission, and N leaching rates reported for WA dryland cropping systems.

Results

Ammonia volatilisation

Volatilisation is the loss of N to the atmosphere as NH3 gas from livestock urine patches and N fertiliser applied to the soil surface. All fertilisers that contain ammonium (NH4+), or can produce NH4+ (e.g. urea) are susceptible to NH3 volatilisation when applied to the soil surface (Freney et al, 1983). Ammonia volatilisation from urea requires the hydrolysis of the urea by the enzyme urease (Bremner and Douglas, 1971). The extent of the losses varies depending on fertiliser placement, soil type, soil pH and pH buffering capacity, crop residue management and environmental conditions (Sommer et al., 2004).
Ammonia volatilisation has been directly measured at one site in the WA grainbelt and indirectly calculated from another (Fillery and Khimashia, 2016). Ammonia volatilisation was directly was directly measured from an acidic soil in Merredin cropped to wheat, and following the application of urea (30 kg N/ha) to the soil surface. Losses were estimated using a field-based non-intrusive micrometeorological method. Cumulative NH3 losses after 12 days represented 29% of the N applied, which is at the upper range of values reported for Australian cropping soils (e.g. Schwenke et al. 2014). These losses were consistent with values indirectly calculated from a study site at Regans Ford.
A variety of strategies to decrease NH3 losses from cropping soils have been investigated in Australia. This has included decreasing stubble retention and using urea coated with N-(n-butyl) thiophosphoric triamide (i.e. ‘green’ urea) to inhibit the enzyme urease (Bacon and Freney, 1989; Turner et al, 2010; Schwenke et al, 2014). Applying green urea decreased NH3 losses by up to 89% in comparison to applying urea, but the results have not consistent across all Australian studies (Turner et al, 2010; Schwenke et al, 2014).

Nitrous oxide and dinitrogen gas emissions

Nitrous oxide (N2O) and dinitrogen (N2) emissions represent another gaseous loss pathway for N fertiliser applied to cropping soils. Both gases are produced in soils by microorganisms under particular soil conditions. Denitrification is the reduction of nitrate (NO3-) to N2, with N2O an intermediary gaseous product (Wrage et al, 2001). Denitrification occurs in anaerobic microsites in the soils when there is sufficient NO3- and available carbon. Nitrification converts soil NH4+ to NO3- under aerobic conditions, and like denitrification, incomplete conversion results in N2O emissions (Wrage et al, 2001). Applying N fertilisers, whether synthetic or organic, enhances soil microbial production of N2O and N2 (Davidson, 2009; Smith et al, 2012). 
Quantifying N2O and N2 emissions from soils is challenging. Losses vary spatially, and differ from day-to-day in response to factors that regulate production and subsequent emission of these gases (Butterbach-Bahl et al, 2013). Soil N2O emissions are most often measured in the field using manual (static) or fully automated chambers. Chamber measurements are short-term (e.g. hourly), repeated in intervals of hours (automated systems) to weeks (manual systems), and are in turn integrated across time to calculate annual losses. The ability of chambers methods to adequately quantify N2O losses relies on the user characterising N2O emissions during the year, in particular peak emissions following N fertiliser applications and soil re-wetting, which may contribute up to 70% of the total annual flux (e.g. Barton et al, 2008).
Fifteen annual N2O emissions have been reported for lupins and various N fertilised crops in WA (Table 1). Losses have been estimated using automated field-based chambers that measured emissions on a subdaily basis. Studies have been conducted on duplex and deep sands, with granular urea applied at a range of application rates depending on the crop and year (0–100 kg N/ha). Measurements of in situ N2 emissions have not been reported for Western Australian cropping soils; it is best measured using 15N tracer methods (McGeough et al. 2012; Kulkarni et al. 2013).
Nitrous oxide emissions reported from WA cropping soils have been small (0.02–0.27 kg N/ha/yr) and represented <0.12% of applied N fertiliser (Table 1). Including grain legumes in cropping rotations has not enhanced N2O emissions during or the growing season or between growing seasons in WA studies (Barton et al, 2011; Barton et al, 2013); unlike studies conducted in Eastern Australia and overseas (e.g. Barton et al, 2011; Grace 2015). Although increasing soil carbon increased N2O emissions at Buntine, the losses represented <0.12% of the N fertiliser applied (L. Barton, pers. comm). Developing strategies for mitigating N2O fluxes from cropping soils in our region is challenging (and possibly not warranted based on current data) as losses occur between growing seasons in response to summer rainfall rather than N fertiliser applications. Strategies that control soil N supply from nitrification following soil wetting, or immobilise excess inorganic N via soil microbial or plant uptake, would be expected to decrease the availability of N for subsequent N2O emission. Increasing the efficiency of the nitrification process by increasing soil pH (via liming) decreased N2O emissions from an acidic soil following summer rain (Barton et al. 2013).

Table 1. Annual nitrous oxide emissions (N/ha/yr) for Western Australia dryland cropping soils

Location	Crop and treatment	Soil carbon (%)	N fertiliser (kg N ha-1)	Study period and duration (days)	N2O emission (kg N/ha/yr)	Reference
Buntine	canola+OM	1.20	100a	Jun. 2012–Jun. 2013; 365	0.14	Barton (pers. comm)
Buntine	canola	0.64	100a	Jun. 2012–Jun. 2013; 365	0.02	Barton (pers. comm)
Buntine	barley+OM	1.20	100a	Jun. 2013–Jun. 2014; 365	0.27	Barton (pers. comm)
Buntine	barley	0.64	100a	Jun. 2013–Jun. 2014; 365	0.02	Barton (pers. comm)
Cunderdin	wheat	0.98	100a	May 2005–May 2006; 354	0.11	Barton et al. (2008)
Cunderdin	canola	0.98	75a	May 2007–May 2008; 356	0.13	Barton et al. (2010)
Cunderdin	lupin	0.98	0	May 2008–May 2009; 349	0.13	Barton et al. (2011)
Wongan Hills	wheat+lime	1.03	75a	Jun. 2009–Jun.2010; 371	0.05	Barton et al. (2013)
Wongan Hills	wheat	1.03	75a	Jun. 2009–Jun.2010; 371	0.06	Barton et al. (2013)
Wongan Hills	lupin+lime	1.03	0	Jun. 2009–Jun.2010; 371	0.05	Barton et al. (2013)
Wongan Hills	lupin	1.03	0	Jun. 2009–Jun.2010; 371	0.04	Barton et al. (2013)
Wongan Hills	wheat+lime	1.03	50a	Jun. 2010-Jun./2011; 364	0.04	Barton et al. (2013)
Wongan Hills	wheat	1.03	50a	Jun. 2010-Jun./2011; 364	0.07	Barton et al. (2013)
Wongan Hills	wheat+lime	1.03	20a	Jun. 2010-Jun./2011; 364	0.06	Barton et al. (2013)
Wongan Hills	wheat	1.03	20a	Jun. 2010-Jun./2011; 364	0.06	Barton et al. (2013)

^aUrea; OM, organic matter.

Nitrogen leaching 

Nitrogen leaching occurs from cropping soils when water drains through the soil and beyond the rooting zone. Any N not taken up by the crop, immobilised in soil organic matter, denitrified or volatilised to gaseous N compounds is susceptible to leaching when drainage occurs. The extent of N leaching will also be affected by the rate that dissolved N moves through the soil profile. Crop N uptake and soil biological processes often occur at greater rates in the surface- than the sub-soil. Nitrogen fertiliser practices that maximise the contact time between the dissolved nutrients and the surface soil should increase crop N uptake and soil ‘retention’, and therefore decrease the risk of N leaching. The extent of leaching beyond the surface soil will also vary depending on amount and intensity of rainfall, soil texture and structure, and the extent of the crop rooting zone. Nitrogen losses will be less in those soil types where soil drainage moves evenly through the entire soil profile (‘matrix’ flow) than in soil types were soil water moves through macropores such as down cracks, along old root channels and worm holes (‘preferential’ flow) (Barton et al., 2004). 
Nitrogen leaching from soils is best quantified directly, and throughout the year to account for seasonal changes in soil N availability (Addiscott, 1996). Measuring N leaching for an extended period will also account for any effects of establishing the experiment (e.g., soil disturbance) on N leaching. Nitrogen can be leached as NO3-, NH4+ and as organic N. It has often assumed that NO3- is more susceptible to leaching than other forms of N as the solid phase of most soils has a net negative charge, and so repels negatively charged anions such as NO3-. However, organic N leaching can also be significant from agricultural soils (Murphy et al., 2000).
Nitrogen leaching has been measured from two study sites on deep sands cropped to wheat following a grain legume or pasture, and in the absence of N fertiliser (Table 2). Nitrogen leaching losses have been estimated by measuring NO3- losses (to a depth of 1.0–1.5m) using either porous (suction) cup lysimeters in combination with soil hydrological models or soil lysimeters fitted with anion exchange resin traps at the base. Nitrate leaching losses have ranged from 14 to 72 kg N/ha during the growing season, and 14 kg N/ha in one study where losses were measured for a year. Greatest losses during the growing season have occurred from soils cropped to wheat following lupin (Table 2).  
Table 2. Nitrogen leaching (NO₃-N/ha) for Western Australia dryland cropping soils

Conclusion

Better utilisation of N fertiliser in WA cropping system requires an understanding of the timing and magnitude of N release from soil organic matter via mineralisation, as well as an understanding of the conditions whereby major losses from the soil can occur. The extent and timing of soil N supply (net N mineralisation) has been reported from a limited number of locations in the grainbelt, but demonstrates up to 0.5 kg N/ha may be released in a day during the growing season. Nitrogen leaching and NH3 volatilisation represented significant losses of soil N in the studies reviewed, however the number of field-based measurements is extremely limited. Further research is needed quantify soil N mineralisation and NH3 volatilisation under current WA farming practises, and incorporate into decision support tools.

References

Reference list available on request.

Acknowledgments

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the authors would like to thank them for their continued support.