Printer Friendly

Mitigation of nitrous oxide emissions from furrow-irrigated Vertosols by 3,4-dimethyl pyrazole tetra-methylene sulfone, an alternative nitrification inhibitor to nitrapyrin for direct injection with anhydrous ammonia.

Introduction

Over the past four decades, breeding and improved management practices have increased irrigated cotton lint yields in Australia by >40% (Liu et al. 2013), but crop demand for nitrogen (N) has increased even more (Rochester and Bange 2016). Although much of the N taken up by the crop is sourced from the soil (Rochester and Constable 2015), additional N fertiliser is required for optimum crop growth in most situations (Rochester 2011). According to regular surveys of Australian cotton producers, the average amount of N fertiliser applied to irrigated cotton has increased from 125 kg N/ha in 1997, to 298 kg N/ha in 2016 (range: 18-519 kg N/ha) (Roth 2017; CRDC 2018). Much of the Australian cotton crop is grown on slow-draining, alkaline, medium-heavy clay soils (Vertosols) (Isbell and National Committee on Soil and Terrain 2016). On these soils, >50% of the applied fertiliser N may be lost from the soil as gaseous nitric oxide (NO), nitrous oxide ([N.sub.2]O) and di-nitrogen ([N.sub.2]) (Freney et al. 1993), with the majority lost as [N.sub.2] during denitrification (Rochester 2003).

Such large-scale losses of fertiliser N as [N.sub.2] are an important agronomic and economic issue for cotton producers, but the concurrent losses of [N.sub.2]O are a significant environmental concern as [N.sub.2]O is a greenhouse gas with a global warming potential 298 times that of carbon dioxide (Forster et al. 2007). The increasing concentration of [N.sub.2]O in the atmosphere also contributes to stratospheric ozone depletion (Myhre et al. 2013). Release of [N.sub.2]O from agricultural soils increases linearly with the rate of N fertiliser applied to a crop (Breitenbeck and Bremner 1986a) until the N rate exceeds the crop's requirements, where the increase in [N.sub.2]O emitted can become exponential (Grace et al. 2016). In addition to reducing excess N fertiliser use, other methods are required to reduce the production of [N.sub.2]O without impacting profitable agricultural production.

Most fertiliser N applied to irrigated cotton is banded into the soil several months before sowing, either as urea (66% of farms surveyed) or anhydrous ammonia (AA) (41% of farms surveyed) (Roth 2017). Some researchers have found AA to emit proportionally more [N.sub.2]O than other N fertiliser products (Bouwman et al. 2002), but this may be more to do with differences in application placement than formulation, i.e. banding AA vs broadcasting urea (Breitenbeck and Bremner 19866). Only one previous [N.sub.2]O study in Australian cotton systems has used AA as an N source (Grace et al. 2016), but did not compare with [N.sub.2]O emissions from other N products. Both AA and urea initially produce ammonium (N[H.sub.4.sup.+]) that has mostly nitrified to nitrate (N[O.sub.3.sup.-]) by the time of sowing. Nitrous oxide is produced in soils during the processes of nitrification (aerobic conversion of N[H.sub.4.sup.+] to N[O.sub.3.sup.-]) and denitrification (anaerobic conversion of N[O.sub.3.sup.-] to NO, [N.sub.2]O and [N.sub.2]) (Bremner 1997; Rochester 2003). For several months until the period of rapid crop N uptake, the large pool of fertiliser-derived N[O.sub.3.sup.-] in the soil is vulnerable to denitrification losses, particularly during prolonged wet weather and initial irrigations (Schwenke and McPherson 2016). Inhibiting the nitrification process would retain the applied fertiliser N in N[H.sub.4.sup.+] form during the pre-plant and early crop growth phase and thus prevent much of the [N.sub.2]O produced from denitrification (Singh and Verma 2007; Chen et al. 2008).

The use of various nitrification inhibitors, either co-applied with N[H.sub.4.sup.+]-producing fertilisers or as a coating on urea has been extensively researched (Freney et al. 1993; Chen et al. 2008), including for their effects on reducing [N.sub.2]O emissions (Chen et al. 2010; Tian et al. 2015; Scheer et al. 2016). However, almost all nitrification inhibitor investigations with AA have focused on nitrapyrin (2-chloro-6-trichloro methyl pyridine), which has been in use since the 1960s (Zerulla et al. 2001; Wolt 2004). Bremner et al. (1981) found that the addition of nitrapyrin to AA being applied to fallowed soil reduced [N.sub.2]O emissions by 63-87%, depending on the time of year of application, which Gomes and Loynachan (1984) ascribed to accumulated heat units (time x temperature) reducing the effectiveness of the inhibitor. In another study, [N.sub.2]O emissions from nitrification were reduced by nitrapyrin applied as a solution to AA bands in fallow plots, but only in one of two soils trialled by Magalhaes et al. (1984). The ineffectiveness was thought to be due to biological deactivation of the nitrapyrin in the more organic soil. More recent research has demonstrated the effectiveness of nitrapyrin applied directly with AA in retaining more mineral N in the soil and increasing crop yields as a result (Goos and Johnson 1999; Randall et al. 2003; Wolt 2004; Degenhardt et al. 2016), although yield effects were not always positive (Blackmer and Sanchez 1988). Parkin and Hatfield (2010) found that fall-applied AA + nitrapyrin reduced the [N.sub.2]O emissions during fall-spring, but not when totalled over an entire year. Nitrapyrin is commercially available for use with AA in USA (as Nserve[TM], Dow AgroSciences), but not in Australia. Recent reports of nitrapyrin in stream water samples in USA (Woodward et al. 2016) may have implications for its continued use, so the range of nitrification inhibitors compatible with AA needs to expand.

Of the many other nitrification inhibitors trialled with urea, Chen et al. (2008) recommended 3,4-dimethyl pyrazole phosphate (DMPP) due to its documented benefits on both crop yield and N loss, and its stability and lack of movement in soil. DMPP has been available since 1999. Coating DMPP on urea (ENTEC[R]) has been shown to reduce [N.sub.2]O emissions compared with untreated urea (Chen et al. 2008; Schwenke et al. 2016), and has been found to be at least as effective as other inhibitors, including nitrapyrin (Weiske et al. 2001; De Antoni Migliorati et al. 2014; Scheer et al. 2016; Yang et al. 2016). DMPP has not previously been applied with AA. Our preliminary investigations found that DMPP formed a precipitate on contact with AA, thus preventing its co-application. In response, Incitec-Pivot Fertilisers (Australia) developed a new formulation containing 3,4-dimethyl pyrazole tetra-methylene sulfone (DMPS), which is miscible with AA (Australian Patent AU 2015227487 B1).

This research aimed to compare the efficacy of the nitrification inhibitors, DMPS and nitrapyrin, when directly injected into the AA stream during pre-plant N application for commercial irrigated cotton production in two contrasting regions, central Queensland and northern New South Wales. Our primary focus was on the potential for these inhibitors to mitigate soil [N.sub.2]O emissions arising from the use of AA in irrigated cotton systems. A secondary aim was to compare the impacts of these inhibitors on soil mineral N and cotton crop N uptake.

Methods

Experimental sites

This research was conducted at two commercial irrigated cotton farms in north-east Australia, near the towns of Emerald (Queensland) (23.5S, 148.3E) and Gunnedah (New South Wales) (31.OS, 150.3E), 860 km apart. The soil type at both sites was a Black Vertosol (Isbell and National Committee on Soil and Terrain 2016). At Emerald, the soil pHw range was 8.9-9.1 (0-0.9 m), organic carbon (C) was 0.56% (0-0.15 m) and soil texture was 42% clay, 44% sand and 14% silt (0-0.3 m). At Gunnedah, the soil pHw range was 8.8-9.3 (0-0.9 m), organic C was 0.93% (0-0.15 m) and soil texture was 65% clay, 9% sand and 26% silt (0-0.3 m). Both soils contained 0.4-0.5% inorganic C (0-0.15 m). Both soil profiles had a total of 34 kg N/ha of mineral N (N[O.sub.3.sup.-] + N[H.sub.4.sup.+]) to 0.9 m depth before N application.

Paddock management and weather

All N application, sowing and picking were done by the co-operators using commercial equipment. At the Emerald site, AA was applied pre-planting on 10 August 2016. Cotton (Gossypium hirsutum L.) var. Sicot 746B3F was planted on 20 August 2016. Cotton picking at Emerald was on 13 February 2017. Pre-plant AA was applied at the Gunnedah site on 29 July 2016. Cotton planting with var. Sicot 748B3F occurred on 6 October 2016, and picking was on 9 May 2017. At both sites, rainfall and air temperature were continuously logged with an automated weather station (WM3000, Environdata, Warwick, Queensland, Australia).

Experimental treatments

The experimental design at both sites was a randomised complete block design with three treatments (T1 = AA, T2 = AA + DMPS and T3 = AA + nitrapyrin (N-serve[TM])) and three replications. Each plot was 8 m wide (i.e. eight plant rows) by the full paddock length (1060 m at Emerald, 650 m at Gunnedah). There was no nil-N treatment due to the potential for significant financial loss in these large-scale plots. However, as the applied AA was banded, there were significant areas of each plot with no applied N, which, through discrete sampling gave results typical of unfertilised areas of soil.

At both sites, the farmer's 8-m wide A A application rig was used to apply the treatments. A Raven SideKick Pro[R] direct injection system (Raven Applied Technology, Sioux Falls, South Dakota, USA) was used to apply nitrification inhibitor into the AA stream at the super-cooler unit before the distributor. Both products were applied at the manufacturer's recommended rate of 2.5 L/ha.

At Emerald, the pre-plant N (AA) was applied at 150kgN/ ha, followed by an in-crop application of 150 kg N/ha as a urea side-dress on 17 October 2016, just before the second irrigation. We used normal urea to side-dress treatments T1 and T3, and DMPP-coated urea (ENTEC[R]) to side-dress T2. Nitrapyrin-coated urea was not available to use in T3. Pre-plant AA was injected through a knife-point tine into the plant bed to a depth of 0.30 m at a distance of 0.25 m from the plant row on both sides of the plant bed. The side-dressed urea was applied at a depth of 0.05 m on only the irrigated side of the plant bed, then immediately incorporated by cultivation. After side-dressing, measurements on the non-irrigated side of the plant bed allowed continued scrutiny of the pre-plant inhibitor treatments, whereas measurements on the irrigated side showed the effects of the side-dressed urea, with (T2) or without (T1 and T3) inhibitor.

At Gunnedah, 300 kg N/ha was applied pre-plant N as AA. No in-crop N application was planned, but the experimental area was top-dressed with 75 kg N/ha as aerially-broadcast urea late in the vegetative growth period. The pre-plant AA was applied with a delta-T applicator centred on the middle of the non-irrigated furrow. Approximately 85% of the AA was applied 0.09 m either side of the mid-line (to a depth of 0.45 m below the plant bed) as super-cooled (liquid) ammonia, with the remaining ~15% of the N applied as vapour in the centre of the furrow (to a depth of 0.15 m below the furrow surface).

[N.sub.2]O emissions--manual chambers

The [N.sub.2]O emissions were measured during five (Emerald) and six (Gunnedah) separate 7-14 day-long sampling campaigns with samples collected 1, 2, 4, 7 and sometimes also 10 or 14 days after a significant rainfall or irrigation event. All gas emissions measurements took place during the first 4 months after the pre-plant N fertiliser application because seasonal [N.sub.2]O emissions from irrigated cotton in this region typically peak during early crop growth (Schwenke and McPherson 2016). All gas sampling took place in a mid-field transect across all nine plots. There were four manual chambers within each plot situated on the irrigated furrow (IF), the non-irrigated furrow (NF), the irrigated side of the plant bed (IH) and the non-irrigated side of the plant bed (NH). Results from the four chambers in a plot were averaged to give an emission rate per hectare for each plot. At Emerald, the pre-plant AA was applied to both sides of each plant bed, so both IH and NH chambers were located above banded N fertiliser. All chambers were temporarily removed during side-dressing, when urea was applied to the irrigated side of the plant bed, so chamber IH covered the side-dress urea and pre-plant AA, while chamber NH covered only pre-plant AA. This allowed post side-dress measurements to discriminate between effects from the original fertiliser and inhibitor application, and those from the more recent side-dress application. At Gunnedah, AA was applied to the non-irrigated furrows and the side of the hill adjoining, so chamber NH was located on a pre-plant fertiliser band but chamber IH was not. Chamber IH was located higher up on the side of the hill than NH.

Manual chambers were cylindrical (0.15 m diameter and 0.15 m height) and were secured at the time of sampling to a base pushed 0.1 m into the soil. On each sampling day, starting at 1000 hours (AEST), chambers were secured to their base and headspace samples (25 mL) were collected immediately, and again after 60 min. Each sample was injected into a pre-vacuumed 12-mL glass exetainers (Labco, Lampeter, Wales, UK) for later analysis of [N.sub.2]O by gas chromatograph. The difference in [N.sub.2]O concentration between the 60-min and 0-min samples was divided by the measurement time to give the slope. This slope was used to calculate the [N.sub.2]O flux using gas flux formulae corrected for air temperature and barometric pressure (Schwenke and Haigh 2016). The hourly flux rate measured at this time of day was assumed to approximately the daily average flux, and was used to calculate cumulative [N.sub.2]O-N/ha loss for each individual chamber for the duration of each measurement campaign by interpolation between measurement days.

[N.sub.2]O emissions--semi-automated chambers

The semi-automated system featured four polycarbonate chambers (0.5 m x 0.5 m x 0.15 m) secured to a stainless steel base pushed 0.1 m into the soil. At the programmed time of sampling (1000 hours), the lid of each chamber automatically closed and headspace sampling commenced from each chamber using a pump and a syringe injecting the sampled gas into 12-mL exetainers for later analysis of [N.sub.2]O by gas chromatograph. Each chamber was sampled three times during a closure of 1 h, after which the lids automatically reopened. The slope of concentration change versus time, integrated to time zero, was used to calculate [N.sub.2]O flux as above.

The semi-auto chambers were located on planted hills at Emerald and included cotton plants. The chamber height was increased to 0.65 m to accommodate plant growth, with plants trimmed to fit. The four chambers sampled two replicates of treatments T1 and T2 bi-weekly from AA application until December 2016. At the Gunnedah site, two sets of semi-auto systems were used to measure emissions from two replicates of all three treatments. Because the pre-plant AA application centred on the non-irrigated furrow, the semi-auto chambers were also centred on this furrow and did not include growing plants. There were a total of 33 sampling days at Emerald and 48 at Gunnedah.

Soil mineral N and water

At both sites, surface soil (0-0.1 m) was sampled using a core sampler (0-0.05 m diameter) in line with each of the 36 manual chambers on six occasions (at the completion of each gas sampling campaign). At each chamber location, three cores were composited, then sub-sampled for moisture content determined by weight change after oven drying at 105[degrees]C for 48 h, and mineral N (N[H.sub.4.sup.+] and N[O.sub.3.sup.-]) concentration, determined by flow injection analysis of 1 M KCI extracts done using field-moist samples (Rayment and Lyons 2010). Volumetric soil water content was also determined on each day of gas sampling using a standing wave ratio probe (MP306, ICT International, Armidale, New South Wales, Australia) calibrated for use at each site. Water-filled pore space was calculated (Linn and Doran 1984) from these data using estimates of bulk density taken during the soil sampling.

The Gunnedah experiment was soil sampled to 0.9 m depth immediately before cotton planting, using a hand-held jackhammer and a 0.05 m diameter corer. In each plot, we took five cores, three to 0.3 m depth (0-0.15 and 0.15-0.3 m) and another two cores to 0.9 m depth (0-0.15, 0.15-0.3, 0.3-0.6 and 0.6-0.9 m). All cores from a plot were composited by depth before analysis for moisture and mineral N analysis as above.

Plant measurements

Prior to crop defoliation, whole aboveground plant samples were collected from three locations within each plot at each site. At each location, all plants in a 1.0 m length of row were cut at ground level, weighed in the field, sub-sampled and the sub-samples oven-dried to determine moisture content. From this we calculated biomass dry matter. These samples were then separated into vegetative matter (leaves, stems and boll husks) and lint, which was then ginned to calculate the clean lint yield and seed weight. All fractions were then finely ground and analysed for total N concentration using a combustion analyser (EA1112, Thermo Finnigan, San Jose, California, USA).

The middle six rows of cotton from each paddock-length plot were harvested by a commercial cotton picker. All raw cotton from each plot was baled by the picker and each bale was weighed in the field using truck scales. Ginning and cotton quality results from each bale were obtained from a nearby commercial gin and lint testing service.

Statistical analysis

Treatments were compared using analysis of variance in Genstat (18th Edition, VSN International, Hemel Hempstead, UK). Mean differences were judged to be significant where P [less than or equal to] 0.05, unless otherwise indicated. Skewed [N.sub.2]O flux data were log-transformed before analysis, with back-transformed means used in presentation of the results. The impact of manual chamber position within a plot on results was determined using a split-plot model.

Results

Weather and water

At Emerald, 10 mm of rain fell during the 10-day pre-plant period between N application and sowing, with another 183 mm between sowing and harvest. A total of 106 mm of this fell within the gas emissions measurement period ending in early December (Fig. 1a). A total of 7.75 ML/ha of irrigation water was applied over eight events, including once before the pre-plant N application. Manual gas measurement campaigns followed the first four irrigations post planting, and following a rainfall event of 22 mm during September. Soil moisture contents declined slowly post-irrigation/rainfall early in the season, but more quickly during October-November. Average daily air temperature for the whole growing season was 23.8[degrees]C (range: 12.0-33.2[degrees]C), with 5 days recording maxima in excess of 40[degrees]C. Fig. 2a shows the daily maxima and minima for the measurement period at Emerald.

A total of 280 mm of rain fell during the 69-day pre-plant period until sowing at the Gunnedah site. A further 428 mm fell between sowing and picking, with 84 mm of this during the gas emissions measurement period up until early December (Fig. 1e). The crop was irrigated a total of eight times. Because of the wet conditions pre-season, no pre-plant or planting irrigation was required. During the growing season, a total of 7.0 ML/ha of irrigation water was applied over eight events. Manual gas chamber measurements were triggered by rainfall events of 38, 30, 44, 19 and 36 mm and by the first irrigation event. Several significant rainfall events also occurred during the sampling campaigns. Average daily air temperature for the growing season was 23.4[degrees]C (range: 11.1-34.8[degrees]C), with 16 days with maxima in excess of 40[degrees]C between early December 2016 and mid-February 2017. Fig. 2b shows the daily maxima and minima for the measurement period at Gunnedah.

Daily [N.sub.2]O emissions

The [N.sub.2]O emissions after the first irrigation at Emerald were initially higher in T1 in the furrows but, after a few days, were highest from the sides of the hill. Hillside emissions had not reached baseline levels after 10 days post-irrigation (Fig. 1 b). In contrast, [N.sub.2]O emissions from T2 and T3 did not exceed 10 g [N.sub.2]O-N/ha/day at any location during this period (Fig. 1c, d). Similarly, the semi-auto chamber results showed a large initial response to the first irrigation from T1, but much less from T2 (Fig. 2a). High variation between replicates meant that apparent treatment differences between T1 and T2 were often not statistically significant, but there was still a significant treatment difference found one month after the first irrigation was applied. Rainfall at the end of September had minimal impact on [N.sub.2]O emissions from any treatment. The third measurement-trigger event was the second in-crop irrigation, applied directly after the urea side-dressing application. Daily [N.sub.2]O flux peaked on the fourth day after irrigation in the irrigated side of the hill, where the side-dressing had occurred, in the two treatments where ordinary urea was side-dressed (T1 and T3) (Fig. lb, d). Fluxes returned to baseline levels by the tenth day post-irrigation. The semi-auto chambers showed a similar pattern, with large and highly variable fluxes from T1 compared with a smaller response to the side-dressing from T2 (Fig. 2c). The next two sampling campaigns produced several very high daily [N.sub.2]O fluxes, mainly from the irrigated side of the hill position, but there were no further N treatment differences.

At Gunnedah, [N.sub.2]O emissions measured during the first manual chamber sampling campaign were initially highest (>20 g [N.sub.2]O-N/ha/day in T1 and T3) from the irrigated furrow position but declined rapidly during the week and were replaced by higher emissions from the non-irrigated furrow and hill positions (Fig. 1f-h). In T2, most emissions declined to <10 g [N.sub.2]O-N/ha/day by the end of the first week, whereas emissions from T3 declined more slowly and those from T1 continued to increase. By the end of the second campaign, daily [N.sub.2]O flux from the (fertilised) non-irrigated position in T1 was >50g [N.sub.2]O-N/ha/day and continued to fluctuate with further rainfall events over the following month at rates up to a maximum of 914 g [N.sub.2]O-N/ha/day (Fig. If). During this period, emissions from T2 and T3 also increased, but the maximum daily rates were 36 and 69 g [N.sub.2]O-N/ha/day for T2 and T3 respectively (Fig. 1g, h). The fifth sampling campaign, in October, led to increased daily emissions from the non-irrigated furrow and, to a lesser extent, from the non-irrigated side of the hill in all three N treatments, although the fluxes from T2 were significantly less than those from T3. In each case, the peak emissions rate occurred on the second day after the rainfall, but fluxes remained high at the end of the week-long campaign. Following the first in-crop irrigation, the final sampling campaign led to high [N.sub.2]O losses from the non-irrigated furrows, and to a lesser extent, from the non-irrigated hillside position of all three N treatments. However, the peak emissions recorded on the second day post-irrigation were significantly higher in T2 and T3 than in T1. Results from the semi-auto chambers (Fig. 2b) at Gunnedah show very similar trends in treatment differences to those from the manual chambers (Fig. 1f-h). Fluxes were greater from T1 than either T2 or T3 for much of the first two months after pre-plant N application. After mid-October, fluxes from the three N treatments were generally similar, although [N.sub.2]O increased more from T2 than the other two treatments in response to the first in-crop irrigation (Fig. 2d).

Cumulative [N.sub.2]O emitted

While not whole-of-season, our cumulative NzO emissions encompassed the period of influence of the nitrification inhibitors and therefore provide a useful tool for treatment comparisons.

At Emerald, the total [N.sub.2]O emitted was segregated between that emitted before the side-dress urea was applied, and that emitted after. Fig. 3a shows how N treatment and chamber position affected [N.sub.2]O loss. Pre side-dress, [N.sub.2]O loss was significantly greater from T1 than either T2 or T3 (P=0.005), with significantly higher losses from the two hill-side positions than from either furrow position (P = 0.064). After side-dress, [N.sub.2]O loss was similar from T1 and T3 because both were treated with normal urea, but less from T2, which had DMPP-urea (P= 0.074). This was also the case for total losses across the whole post side-dress measurement period (P = 0.068). Emissions were highest from the irrigated hill-side positions and lowest from the irrigated furrow across all N treatments (P= 0.004).

The semi-auto chambers at Emerald gave higher total [N.sub.2]O losses (Fig. 3b) than those from the manual chambers because (i) the manual chamber totals did not include [N.sub.2]O losses that occurred between the sampling campaigns, whereas the semiauto chamber totals covered the entire measurement period and (ii) the semi-auto chambers covered only the plant bed, where fluxes were highest, and not the furrows, where fluxes were typically low. There was no significant treatment difference in cumulative [N.sub.2]O loss up until the side-dress operation, but post side-dress losses were significantly greater from T1 than T2 (P = 0.04), as were total losses for the whole measurement period (P = 0.016).

At Gunnedah, total [N.sub.2]O losses measured with manual chambers were T1 >> T3 > T2 (P = 0.004), with losses strongly affected by chamber position, i.e. NF > NH >> IF > IH (P < 0.001) (Fig. 3c). The semi-auto chambers also showed significant N treatment effects, with T1 > T2 = T3 (P = 0.019) (Fig. 3d). The greater magnitude of cumulative [N.sub.2]O loss compared with the manual chamber results is explained by the same factors listed above for the Emerald experiment.

Soil mineral N

The N[H.sub.4.sup.+],-N measurements in the surface soil (0--0.1m) at Emerald (Fig. 4a-c) showed no effects of the AA applied below (0.3 m). In contrast, N[O.sub.3.sup.-]-N concentration was high in the surface soil, particularly in the non-irrigated and irrigated hill-side positions (Fig. 4d-f). The N[O.sub.3.sup.-]; concentrations declined in early October after planting, but increased in both hill-side positions of all N treatments after the side-dress urea was applied in mid-October and followed by irrigation.

At Gunnedah, the AA was applied closer to the surface, so the 0-0.1 m soils in both the non-irrigated furrow and non-irrigated hill-side positions showed high N[H.sub.4.sup.+] concentrations when sampled soon after the N was applied (Fig. 5a-c). In T1 (Fig. 5a), N[H.sub.4.sup.+] concentrations declined to low levels by the end of September, but the N[H.sub.4.sup.+] persisted at high concentrations until October in T3 (Fig. 5b) and November in T2 (Fig. 5c).

At Gunnedah, soil coring to 0.9 m immediately before sowing showed treatment differences in N[H.sub.4.sup.+] (Fig. 6a), N[O.sub.3.sup.-] (Fig. 6b) and NH[H.sub.4.sup.+] + N[O.sub.3.sup.-] combined (Fig. 6c). There was no significant treatment difference in total mineral N at the soil surface (0-0.15 m), but both inhibitors clearly retained most of the applied N as N[H.sub.4.sup.+], with much less MV available, compared with the control (T1). Below the surface (0.15-0.3 m), N leaching was apparent, with greater N[H.sub.4.sup.+] (and mineral N) in T3 than T1 or T2, and greater N[O.sub.3.sup.-] in T1 than T2. The latter effect extended down to the 0.3-0.6 m sampling interval.

Plant biomass, N uptake and lint yield

At maximum crop biomass, we recorded a higher plant density in T2 than T1 or T3, but this was balanced by a reduced number of bolls per plant in T2 compared with the other treatments at Emerald (Table 1). Despite this, there were no treatment effects on lint yield or seed N exported, but T2 did have more N concentrated in the lint itself. The commercial harvest yield for the experimental field averaged 2472 kg/ha with no significant treatment effects. Commercial pickers do not recover all lint, particularly from smaller immature bolls. The hand cuts were also done while much of the lint was immature and had a thicker fibre.

There were no treatment effects on any measured plant parameter at Gunnedah (Table 1), where commercial lint yields (1908 kg/ha) were also below those measured using hand cuts at peak biomass stage.

Discussion

General

The need for effective nitrification inhibitors in high-input agriculture is being driven by competing issues of (a) rapidly increasing agricultural production requiring greater N inputs (Roth 2017) and (b) the rising atmospheric concentration of [N.sub.2]O, much of which originates from N fertilisers (Forster et al. 2007). Nitrification inhibitors are a logical strategy for mitigating [N.sub.2]O production from irrigated cotton systems where large rates of N fertiliser (>300 kg N/ha) are typically applied several months ahead of crop N uptake. This is especially the case for Vertosols, the dominant soil type used in the Australian irrigated cotton industry, which are medium-heavy clay, slow-draining soils where denitrification is likely to be a major mechanism of N loss (Freney et al. 1993).

A range of nitrification inhibitors have been researched for their potential to increase crop yields and reduce N losses (including [N.sub.2]O) from applied fertiliser (Chen et al. 2008). However, only nitrapyrin, an organo-chlorine compound, is commercially available for use with AA (in USA), and its future use may be impacted by its recent detection in streams (Woodward et al. 2016). Our research is the first to demonstrate that the newly-developed nitrification inhibitor DMPS can also be applied directly with AA during pre-plant N application to irrigated cotton fields. Although not a focus of this study, it is logical that nitrification inhibitors would be more effective when mixed directly within the AA stream during application rather than applied separately (Ruser and Schulz 2015).

[N.sub.2]O reduction by DMPS and nitrapyrin

Compared with untreated AA, DMPS reduced [N.sub.2]O emitted from Vertosols by 86% and 77% at Gunnedah and Emerald respectively, and the [N.sub.2]O reduction due to nitrapyrin was 65% and 77% during the 2-4 month measurement periods of the two experiments (2 months at Emerald until the side-dress N application). These reductions in [N.sub.2]O emissions due to nitrapyrin were above the average of 51% from USA research with AA application (Wolt 2004). There are no previous comparisons of these inhibitors when applied with AA, but research with other fertiliser formulations has found that the effectiveness of DMPP either outlasted that of nitrapyrin (Scheer et al. 2016; Duncan et al. 2017) or was similar (Chen et al. 2010; Wang et al. 2016). Both compounds are thought to inhibit nitrification through chelation of copper, which is a co-factor in the enzyme responsible for ammonia oxidation (Singh and Verma 2007; Ruser and Schulz 2015). Therefore, the observed differences in their effectiveness are likely due to their longevity in the soil (Duncan et al. 2017), with nitrapyrin known to be potentially volatile and leachable, whereas DMPP binds to soil particles and can persist in soils for several months (Weiske et al. 2001).

Total [N.sub.2]O emissions at the Emerald experiment were 10 times less than at Gunnedah, most likely due to drier conditions at the former and very wet conditions at the latter. The period between pre-plant application and sowing was also much shorter at Emerald due to experimental constraints in this study and its earlier sowing time due to lower latitude. It is well known that production of [N.sub.2]O in soils is highly dependent on the effects on nitrification and denitrification of climate type, which affects soil moisture content, temperature, texture, pH and organic C content (Stehfest and Bouwman 2006). Soil texture, temperature and moisture content are well known to influence the effectiveness of nitrification inhibitors (Barth et al. 2001; Irigoyen et al. 2003; Ali et al. 2008; Chen et al. 2010; Menendez et al. 2012; Ruser and Schulz 2015). As both experiments were on similar-textured soils, the hot (Fig. 2a) and dry (Fig. la) conditions at Emerald appear to have reduced the duration of inhibitor effectiveness compared with the warm (Fig. 2b) and wet (Fig. 1e) conditions at Gunnedah. At Emerald, both inhibitor products reduced emissions during the first two measurement periods, but their effectiveness had dissipated by the time of the second irrigation (Fig. 1c, d) as shown by the lack of treatment effects on [N.sub.2]O emissions from the non-irrigated side of the plant bed (not side-dressed) following the second irrigation (Fig. 1). The [N.sub.2]O emissions from the side-dressed side of the plant bed in T1 and T3, compared with T2, showed that additional inhibitors need to be applied with in-crop-applied N to maintain a low emissions strategy for the whole crop.

At Gunnedah, the soil was wet at the time of pre-plant AA application and then received three months of frequent heavy rainfall (Fig. 1e). Nitrification inhibitors were designed to retain mineral N from applied N fertiliser in the soil during periods of excessive rainfall that may otherwise lead to large losses of N[O.sub.3.sup.-]-N through leaching or denitrification (Chen et al. 2008). We found indications of downward mineral N movement in soil core samples taken at sowing in untreated and nitrapyrin-treated plots, although the mineral N had not moved beyond the root zone (Fig. 6). For heavy clay soils such as at the Gunnedah site, denitrification of soil N[O.sub.3.sup.] during periods of anaerobic soil conditions is the most likely major N loss pathway and therefore also the principal source of [N.sub.2]O emission (Freney et al. 1993; Rochester 2003). Our results highlighted the substantial mitigation of [N.sub.2]O emissions during this period due to the addition of both nitrification inhibitors. The DMPS appeared to remain active for longer in the soil than the nitrapyrin, as shown by the greater cumulative [N.sub.2]O reduction with the DMPS (Fig. 3c) and the temporal trends in soil N[H.sub.4.sup.+] concentrations (Fig. 5b, c). The N[H.sub.4.sup.+] in T3 treatment had largely disappeared (nitrified) by the end of October, whereas N[H.sub.4.sup.+] concentration in T2 was still high at this time. Similar trends in soil N[H.sub.4.sup.+] and N[O.sub.3.sup.-] concentrations with time were reported by Scheer et al. (2016) and (Freney et al. 1993) after inhibitors were applied with urea to similar soils as in our study.

Plant response

Nitrification inhibitor effects on plant N uptake and lint yield were also examined, but few treatment effects were observed. Previous research in irrigated cotton (Freney et al. 1993) has demonstrated increases in lint yield with the use of nitrification inhibitors with urea, at all but the highest rate of N application (180kgN/ha). The lint yields in their large-scale field study peaked at 1650 kg/ha when 90 kg N/ha was applied. No treatment effect was found at the higher N rate because the yield was already maximised for that variety in that season. Twenty-seven years later in our study, 300 and 375 kg N/ha were applied to the Emerald and Gunnedah experiments, for commercial picker yields of 2472 and 1908 kg lint/ha respectively. Rochester and Bange (2016) have shown that lint yields in this range require >150 kg N/ha fertiliser application, so even with a reduction in N loss by the inhibitors, there was still sufficient N available to prevent N-deficiency limiting the lint yield.

Future directions

We demonstrated the efficacy of both DMPS and nitrapyrin to retard nitrification and significantly reduce [N.sub.2]O emissions in both dry and wet conditions, but further research is required to discover whether the use of these products is economically beneficial for growers. To justify the additional cost of applying a nitrification inhibitor with AA requires either a boost in yield or a cost saving. Unless extremely wet conditions during the pre-season or early crop growth have significantly reduced the mineral N available to the plant from the soil, a yield boost is unlikely, particularly as most Australian cotton producers typically over-apply N fertiliser (Rochester 2011). We were unable to demonstrate a significant saving of soil mineral N pool at sowing due to the use of the inhibitors, perhaps due to the large variability encountered when attempting to quantify N applied in highly concentrated bands (Fig. 6).

Cost savings with inhibitor use require a reduction in the N fertiliser rate applied (without any decline in lint yield). Crop yield responses are not always seen from inhibitor usage (Lester et al. 2016). One suggestion for this is the preference by some plant species for N[O.sub.3.sup.-] over N[H.sub.4.sup.+] forms of N, so that growth may actually be retarded if the crop is delayed access to sufficient N[O.sub.3.sup.-] during key growth stages (Sahrawat and Keeney 1984). We observed no crop production or lint yield impacts of the inhibitor-induced delay in fertiliser N nitrification, although the DMPS treatment at Emerald did appear to boost plant establishment numbers (Table 1), albeit at the expense of bolls/plant.

Our soil and [N.sub.2]O emissions measurements covered the expected period of nitrification activity in the soil at both sites, not the whole growing season. Our previous work, and that of others (e.g. Macdonald et al. 2015), showed most [N.sub.2]O emissions occur in the first few months of the cotton growing season, unless the crop receives in-crop N fertiliser applications followed by irrigation or rainfall. At Emerald, no further N fertiliser was applied after the side-dress application and the in-crop rainfall was modest for the remainder of the cropping season. At Gunnedah, no in-crop N applications were planned but the grower broadcasted additional urea N much later in the growing season, beyond the influence of the pre-season nitrification inhibitors.

Previous research by our team (unpublished) has shown, however, that even without in-crop N application, exceptionally wet conditions can result in large [N.sub.2]O emissions, even near the end of the season when most mineral N has been utilised by the crop. Such conditions occurred for the Gunnedah experiment and likely resulted in further [N.sub.2]O emissions after December 2015 when the nitrification inhibitors had worn off (Fig. 5). Fig. 5 also shows that, since the inhibitor had worn off, nitrification of the pre-plant AA was still proceeding during December, so losses of [N.sub.2]O from both nitrification and denitrification (after irrigation and heavy rainfall) would likely have been higher from treatments T2 and T3 than from the untreated AA in T1 during December 2015 to January 2016. However, by this stage, the growing crop's increasing demand for mineral N would have been rapidly depleting the soil of N[O.sub.3.sup.-] (and N[H.sub.4.sup.+]), thus reducing the likely overall emissions of [N.sub.2]O.

Conclusion

A newly formulated nitrification inhibitor, DMPS, can be directly applied with AA during pre-plant N fertiliser application for irrigated cotton production. Until now, nitrapyrin has been the only nitrification inhibitor available for use with AA. When applied with AA, DMPS was just as effective (Emerald) or more effective (Gunnedah) than nitrapyrin in reducing soil [N.sub.2]O emissions from nitrification/denitrification processes stimulated by rainfall and irrigation events. At Gunnedah, retention of the applied fertiliser N in the N[H.sub.4.sup.+] form lasted longer with DMPS than with nitrapyrin. Differences in inhibitor activity at this site were likely related to greater stability of DMPS in the soil over time, but more specific research with this new formulation is required to support this observation because previous research used DMPP. Although the ecological advantages of these inhibitors on [N.sub.2]O emissions are clear, future research is needed to examine the potential effect of DMPS at a range of N application rates to establish its agronomic benefits and its economic potential for use in Australian irrigated cotton farming systems.

Conflicts of interest

The authors declare no conflicts of interest.

https://doi.org/10.1071/SR18114

Acknowledgements

This experiment was part of the project 'Determining optimum nitrogen strategies for abatement of emissions for different irrigated cotton systems', AOTG14013 2013-17, jointly funded by NSW Department of Primary Industries (NSW DPI) and the Australian Department of Agriculture and Water Resources, and administered by the Cotton Research and Development Corporation. Thanks to Ross Burnett and Rod Smith for providing the experimental sites and for applying the anhydrous ammonia and urea in the field. Thanks to Amanda Noone for field work at Emerald, and Pete Perfrement, Tim Grant and Wayne McPherson at Gunnedah. Technical assistance with the application of both inhibitors was provided by Craig Burke and the Incitec-Pivot Big-N team. Incitec-Pivot (Charlie Walker) provided the DMPS inhibitor and ENTEC[R]-urea, and Dow Agrosciences (Richard Jackman) provided the N-serve[R] inhibitor. All soil and plant N analyses were carried out by Clarence Mercer in the IS09001accredited laboratory at Tamworth Agricultural Institute, NSW DPI.

References

Ali R, Iqbal J, Tahir GR, Mahmood T (2008) Effect of 3,5-dimethylpyrazole and nitrapyrin on nitrification under high soil temperature. Pakistan Journal of Botany 40(3), 1053-1062.

Barth G, von Tucher S, Schmidhalter U (2001) Influence of soil parameters on the effect of 3,4-dimethylpyrazole-phosphate as a nitrification inhibitor. Biology and Fertility of Soils 34(2), 98-102. doi:10.1007/s003740100382

Blackmer AM, Sanchez CA (1988) Response of corn to nitrogen-15-labeled anhydrous ammonia with and without nitrapyrin in Iowa. Agronomy Journal 80(1), 95-102. doi: 10.2134/agronj 1988.00021962008000010022x

Bouwman AF, Boumans LJM, Batjes NH (2002) Emissions of [N.sub.2]O and NO from fertilized fields: summary of available measurement data. Global Biogeochemical Cycles 16(4), 6-1-6-13.

Breitenbeck GA, Bremner JM (1986a) Effects of rate and depth of fertilizer application on emission of nitrous oxide from soil fertilized with anhydrous ammonia. Biology and Fertility of Soils 2(4), 201-204.

Breitenbeck GA, Bremner JM (19866) Effects of various nitrogen fertilizers on emission of nitrous oxide from soils. Biology and Fertility of Soils 2(4), 195-199.

Bremner JM (1997) Sources of nitrous oxide in soils. Nutrient Cycling in Agroecosystems 49(1-3), 7-16. doi:10.1023/A:1009798022569

Bremner JM, Breitenbeck GA, Blackmer AM (1981) Effect of nitrapyrin on emission of nitrous oxide from soil fertilized with anhydrous ammonia. Geophysical Research Letters 8(4), 353-356. doi: 10.1029/ GL008i004p00353

Chen D, Suter H, Islam A, Edis R, Freney JR, Walker CN (2008) Prospects of improving efficiency of fertiliser nitrogen in Australian agriculture: a review of enhanced efficiency fertilisers. Soil Research 46(4), 289-301. doi: 10.1071/SR07197

Chen D, Suter HC, Islam A, Edis R (2010) Influence of nitrification inhibitors on nitrification and nitrous oxide ([N.sub.2]O) emission from a clay loam soil fertilized with urea. Soil Biology & Biochemistry 42(4), 660-664. doi: 10.1016/j.soilbio.2009.12.014

(CRDC)(2018) '2017 Grower Survey.' (Cotton Research and Development Corporation: location) pp. 54. Available at https://www.crdc.com.au/ publications/growersurvey [verified 9 October 2018].

De Antoni Migliorati M, Scheer C, Grace PR, Rowlings DW, Bell M, McGree J (2014) Influence of different nitrogen rates and DMPP nitrification inhibitor on annual [N.sub.2]O emissions from a subtropical wheat-maize cropping system. Agriculture, Ecosystems & Environment 186, 33-43. doi: 10.1016/j.agee.2014.01.016

Degenhardt RF, Juras LT, Smith LRA, MacRae AW, Ashigh J, McGregor WR (2016) Application of nitrapyrin with banded urea, urea ammonium nitrate, and ammonia delays nitrification and reduces nitrogen loss in Canadian soils. Crop Forage & Turfgrass Management 2(1), 1-11.

Duncan EG, O'Sullivan CA, Simonsen AK, Roper MM, Peoples MB, Treble K, Whisson K (2017) The nitrification inhibitor 3,4,-dimethylpyrazole phosphate strongly inhibits nitrification in coarsegrained soils containing a low abundance of nitrifying microbiota. Soil Research 55(1), 28-37. doi:10.1071/SR15359

Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey DW, Haywood J, Lean J, Lowe DC, Myhre G, Nganga J, Prinn R, Raga G, Schulz M, Van Dorland R (2007) Changes in atmospheric constituents and in radiative forcing. In 'Climate change 2007: the physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the fntergovemmental Panel on Climate Change.' (Eds S Solomon, D Qin, M Manning, Z Chen, M Marquis, KB Averyt, M Tignor, HL Miller). (Cambridge University Press: Cambridge, UK and New York, USA)

Freney JR, Chen DL, Mosier AR, Rochester IJ, Constable GA, Chalk PM (1993) Use of nitrification inhibitors to increase fertilizer nitrogen recovery and lint yield in irrigated cotton. Fertilizer Research 34(1), 37-44. doi: 10.1007/BF00749958

Gomes SL, Loynachan TE (1984) Nitrification of anhydrous ammonia related to nitrapyrin and time-temperature interactions. Agronomy Journal 76, 9-12. doi:10.2134/agronj1984.00021962007600010003x

Goos RJ, Johnson BE (1999) Performance of two nitrification inhibitors over a winter with exceptionally heavy snowfall. Agronomy Journal 91(6), 1046-1049. doi:10.2134/agronj1999.9161046x

Grace P, Shcherbak I, Macdonald B, Scheer C, Rowlings D (2016) Emission factors for estimating fertiliser-induced nitrous oxide emissions from clay soils in Australia's irrigated cotton industry. Soil Research 54(5), 598-603. doi: 10.1071/SR16091

Irigoyen I, Muro J, Azpilikueta M, Aparicio-Tejo P, Lamsfus C (2003) Ammonium oxidation kinetics in the presence of nitrification inhibitors DCD and DMPP at various temperatures. Soil Research 41(6), 1177-1183. doi: 10.1071/SR02144

Isbell RF, National Committee on Soil and Terrain (2016) 'The Australian soil classification.' (CSIRO Publishing: Melbourne)

Lester DW, Bell MJ, Bell KL, De Antoni Migliorati M, Scheer C, Rowlings D, Grace PR (2016) Agronomic responses of grain sorghum to DMPP-treated urea on contrasting soil types in northeastern Australia. Soil Research 54(5), 565-571. doi: 10.1071/SR15337

Linn DM, Doran JW (1984) Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and nontilled soils. Soil Science Society of America Journal 48(6), 1267-1272. doi:10.2136/sssaj1984.03615995004800060013x

Liu SM, Constable GA, Reid PE, Stiller WN, Cullis BR (2013) The interaction between breeding and crop management in improved cotton yield. Field Crops Research 148. 49-60. doi: 10.1016/j.fcr. 2013.04.006

Macdonald BCT. Rochester IJ, Nadelko A (2015) High yielding cotton produced without excessive nitrous oxide emissions. Agronomy Journal 107(5), 1673-1681. doi: 10.2134/agronj 14.0369

Magalhaes A, Chalk P, Strong W (1984) Effect of nitrapyrin on nitrous oxide emission from fallow soils fertilized with anhydrous ammonia. Fertilizer Research 5(4), 411-421. doi: 10.1007/BF01049121

Menendez S. Barrena I, Setien I, Gonzalez-Murua C, Estavillo JM (2012) Efficiency of nitrification inhibitor DMPP to reduce nitrous oxide emissions under different temperature and moisture conditions. Soil Biology & Biochemistry 53(Supplement C), 82-89. doi: 10.1016/ j.soilbio.2012.04.026

Myhre G, Shindell D, Breon F-M, Collins W, Fuglestvedt J, Huang J, Koch D, Lamarque J. Lee D, Mendoza B, Nakajima T, Robock A, Stephens G, Takemura THZ (2013) Anthropogenic and Natural Radiative Forcing. In 'Climate Change 2013: the Physical Science Basis. Contribution of Working Group I to the Fifth Assessment. Report of the Intergovernmental Panel on Climate Change.' (Eds TF Stocker, D Qin, G-K Plattner, M Tignor, SK Allen, J Boschung, A Nauels, Y Xia, V Bex, PM Midgley. (Cambridge University Press: Cambridge, UK and New York, USA)

Parkin TB, Hatfield JL (2010) Influence of nitrapyrin on [N.sub.2]O losses from soil receiving fall-applied anhydrous ammonia. Agriculture, Ecosystems & Environment 136(1-2), 81-86. doi:10.1016/j.agee.2009.11.014

Randall GW, Vetsch JA, Huffman JR (2003) Corn production on a subsurface-drained Mollisol as affected by time of nitrogen application and nitrapyrin. Agronomy Journal 95(5), 1213-1219. doi: 10.2134/ agronj2003.1213

Rayment GE, Lyons DJ (2010) 'Soil chemical methods-Australasia.' (CSIRO Publishing: Melbourne)

Rochester IJ (2003) Estimating nitrous oxide emissions from flood-irrigated alkaline grey clays. Australian Journal of Soil Research 41(2), 197-206. doi: 10.1071/SR02068

Rochester IJ (2011) Assessing internal crop nitrogen use efficiency in high-yielding irrigated cotton. Nutrient Cycling in Agroecosystems 90(1), 147-156. doi: 10.1007/s 10705-010-9418-9

Rochester IJ, Bange M (2016) Nitrogen fertiliser requirements of high-yielding irrigated transgenic cotton. Crop and Pasture Science 67(6), 641-648. doi: 10.1071 /CP15278

Rochester IJ, Constable GA (2015) Improvements in nutrient uptake and nutrient use-efficiency in cotton cultivars released between 1973 and 2006. Field Crops Research 173, 14-21. doi: 10.1016/j.fcr.2015.01.001

Roth I (2017) Cotton growing practices 2016: findings of the CRDC's survey of cotton growers. (Roth Rural: Narrabri, Australia)

Ruser R, Schulz R (2015) The effect of nitrification inhibitors on the nitrous oxide ([N.sub.2]O) release from agricultural soils--a review. Journal of Plant Nutrition and Soil Science 178(2). 171-188. doi:10.1002/jpln.201400251

Sahrawat KL, Keeney DR (1984) Effects of nitrification inhibitors on chemical composition of plants: A review. Journal of Plant Nutrition 7(9), 1251-1288. doi:10.1080/01904168409363280

Scheer C, Rowlings DW, De Antoni Migliorati M, Lester DW, Bell MJ, Grace PR (2016) Effect of enhanced efficiency fertilisers on nitrous oxide emissions in a sub-tropical cereal cropping system. Soil Research 54(5), 544-551. doi: 10.1071/SR15332

Schwenke GD, Haigh BM (2016) The interaction of seasonal rainfall and nitrogen fertiliser rate on soil [N.sub.2]O emission, total N loss and crop yield of dryland sorghum and sunflower grown on sub-tropical Vertosols. Soil Research 54(5), 604-618. doi: 10.1071/SR15286

Schwenke G, McPherson A (2016) Alternative N application strategies for reduced [N.sub.2]O emissions in flood-furrow irrigated cotton. In 'Proceedings of the 2016 International Nitrogen Initiative Conference, "Solutions to improve nitrogen use efficiency for the world'", 4-8 December 2016, Melbourne, Australia (www.ini2016.com)

Schwenke GD, Herridge DF, Scheer C, Rowlings DW, Haigh BM, McMullen KG (2016) Greenhouse gas ([N.sub.2]O and CH4) fluxes under nitrogen-fertilised dryland wheat and barley on subtropical Vertosols: risk, rainfall and alternatives. Soil Research 54(5), 634-650. doi:10.1071/ SR15338

Singh S, Verma A (2007) The potential of nitrification inhibitors to manage the pollution effect of nitrogen fertilizers in agricultural and other soils: a review. Environmental Practice 9, 266-279. doi: 10.1017/ S1466046607070482

Stehfest E, Bouwman L (2006) [N.sub.2]O and NO emission from agricultural fields and soils under natural vegetation: summarizing available measurement data and modeling of global annual emissions. Nutrient Cycling in Agroecosystems 74(3), 207-228. doi: 10.1007/s 10705-0069000-7

Tian Z, Wang JJ, Liu S, Zhang Z, Dodla SK, Myers G (2015) Application effects of coated urea and urease and nitrification inhibitors on ammonia and greenhouse gas emissions from a subtropical cotton field of the Mississippi delta region. The Science of the Total Environment 533, 329-338. doi:10.1016/j.scitotenv.2015.06.147

Wang B, Ye L, Wan Y, Qin X, Gao Q, Liu S, Li J (2016) Modifying nitrogen fertilizer practices can reduce greenhouse gas emissions from a Chinese double rice cropping system. Agriculture, Ecosystems & Environment 215(Supplement C), 100-109. doi:10.1016/j.agee.2015. 09.008

Weiske A, Benckiser G, Herbert T, Ottow JG (2001) Influence of the nitrification inhibitor 3,4-dimethylpyrazole phosphate (DMPP) in comparison to dicyandiamide (DCD) on nitrous oxide emissions, carbon dioxide fluxes and methane oxidation during 3 years of repeated application in field experiments. Biology and Fertility of Soils 34(2), 109-117. doi: 10.1007/s003740100386

Wolt JD (2004) A meta-evaluation of nitrapyrin agronomic and environmental effectiveness with emphasis on com production in the Midwestern USA. Nutrient Cycling in Agroecosystems 69(1), 23-41. doi: 10.1023/B:FRES.0000025287.52565.99

Woodward EE, Hladik ML, Kolpin DW (2016) Nitrapyrin in streams: the first study documenting off-field transport of a nitrogen stabilizer compound. Environmental Science & Technology Letters 3(11), 387-392. doi: 10.1021/acs.estlett.6b00348

Yang M, Fang Y, Sun D, Shi Y (2016) Efficiency of two nitrification inhibitors (dicyandiamide and 3, 4-dimethypyrazole phosphate) on soil nitrogen transformations and plant productivity: a meta-analysis. Scientific Reports 6, 22075. doi:10.1038/srep22075

Zerulla W, Barth T, Dressel J, Erhardt K, Horchler von Locquenghien K, Pasda G, Radle M, Wissemeier A (2001) 3,4-Dimethylpyrazole phosphate (DMPP)--a new nitrification inhibitor for agriculture and horticulture. Biology and Fertility of Soils 34(2), 79-84. doi:10.1007/ S003740100380

Handling editor: Siobhan Staunton

Graeme Schwenke (iD), (A,B) and Annabelle McPherson (A)

(A) NSW Department of Primary Industries, 4 Marsden Park Road, Calala, NSW 2340, Australia.

(B) Corresponding author. Email: graeme.schwenke@dpi.nsw.gov.au

Received 26 April 2018, accepted 30 August 2018, published online 11 October 2018

Caption: Fig. 1. Soil water, rainfall and irrigation (a, e), and back-transformed daily [N.sub.2]O flux measured with manual chambers at four positions within the three N treatments, T1 (b), T2 (c), T3 (d) at Emerald (left column) and T1 (f), T2 (g), T3 (h) Gunnedah (right column). Solid bars in top graphs indicate rainfall, hollow bars indicate irrigation events during the gas measurement period. Significant differences (P< 0.05) in daily [N.sub.2]O flux are indicated by letters above the sampling date in (c) and (g) graphs (P = chamber position, n = fertiliser treatment, x = P x N interaction) and apply to all three [N.sub.2]O flux graphs from the same site. T1, AA; T2, AA + DMPS; T3, AA + nitrapyrin.

Caption: Fig. 2. Daily maximum and minimum air temperatures at Emerald (a) and Gunnedah (b), [N.sub.2]O flux measured with semi-auto chambers within T1 and T2 at Emerald (c) and T1, T2, and T3 at Gunnedah (d). Significant N treatment differences in (c) and (d) are indicated by

1.s.d. bars (P [less than or equal to] 0.05, * P [less than or equal to] 0.1). T1, AA; T2, AA + DMPS; T3, AA + nitrapyrin.

Caption: Fig. 3. Cumulative [N.sub.2]O flux measured with manual chambers at (a) Emerald, and (c) Gunnedah, and semi-auto chambers at (b) Emerald and (d) Gunnedah. Data are back-transformed means from each chamber position within each N treatment. Significant treatment differences are described in the text. T1, AA; T2, AA + DMPS; T3, AA +nitrapyrin; IF, irrigated furrow; NF, non-irrigated furrow; IH, irrigated side of the plant bed; NH, non-irrigated side of the plant bed.

Caption: Fig. 4. Back-transformed means of ammonium concentration in (a) T1. (b) T2, (c) T3, and nitrate concentration in (d) T1, (e) T2, (f) T3 at four positions in the surface soil (0-0.1 m) at the Emerald site. Letters above a sampling date indicate significant N treatment (N; P< 0.10) and chamber position (P; P< 0.05) differences and apply to all three graphs in a column. T1, AA; T2, AA + DMPS; T3, AA + nitrapyrin.

Caption: Fig. 5. Back-transformed means of ammonium concentration in (a) T1, (b) T2, (c) T3, and nitrate concentration in (d) T1, (e) T2, (f) T3 at four positions in the surface soil (0-0.1 m) at the Gunnedah site. Letters above a sampling date indicate significant (P < 0.05) N treatment (N), chamber position (P) effects, and N x P interactions (X), and apply to all three graphs in each column. T1, AA; T2, AA + DMPS; T3, AA +nitrapyrin.

Caption: Fig. 6. Ammonium (a), nitrate (b) and combined mineral N concentrations (c) in the soil profile (0-0.9 m) of the three N treatments at the Gunnedah site at sowing. Significant N treatment differences at a depth are indicated by 1.s.d. bars (P [less than or equal to] 0.05, * P<0.1). T1, AA; T2, AA + DMPS; T3, AA + nitrapyrin.
Table 1. Measured plant properties from pre-desiccation
sampling at the two experimental sites

T1, AA; T2, AA + DMPS; T3, AA + nitrapyrin; significant
N treatment differences (P < 0.05) are indicated by
different letters following treatment means

                                      Emerald

Measure                       T1         T2        T3

Population (plants/ha)      52 200a   77 800b    63 300a
Boll number (bolls/plant)     26b       17a        20a
Dry matter * (kg/ha)         8847       9127      8690
Lint yield (kg/ha)           3401       3631      3072
Seed yield (kg/ha)           2757       2815      2372
Dry matter * N (kg N/ha)      136       144        146
Lint N (kg N/ha)             9.2a      11.7b      9.0a
Seed N (kg N/ha)              122       123        105

                                      Gunnedah

Measure                       T1         T2        T3

Population (plants/ha)      132 500   135 200    131100
Boll number (bolls/plant)      9         9          9
Dry matter * (kg/ha)         7170       6950      7162
Lint yield (kg/ha)           2470       2472      2180
Seed yield (kg/ha)           2328       2330      2077
Dry matter * N (kg N/ha)      124       122        128
Lint N (kg N/ha)              4.7       5.2        4.7
Seed N (kg N/ha)              104       102        94

* All aboveground plant material excluding lint and seed.
COPYRIGHT 2018 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Schwenke, Graeme; McPherson, Annabelle
Publication:Soil Research
Article Type:Report
Date:Oct 1, 2018
Words:9668
Previous Article:Biostimulation of the activity of microorganisms and soil enzymes through fertilisation with composts.
Next Article:Management of the major chemical soil constraints affecting yields in the grain growing region of Queensland and New South Wales, Australia--a review.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters