Exponential response of nitrous oxide (N₂O) emissions to increasing nitrogen fertiliser rates in a tropical sugarcane cropping system

Highlights

• 65–95 % of total N₂O emissions occurred within three months after N fertilization.

• Cumulative N₂O emissions of 0.29–3.06 kg N ha⁻¹ responded exponentially to N rates.

• EF ranged from 0.69 to 1.11% and exceeded the IPCC default of 1% at 250 kg N ha⁻¹.

• N₂O intensities for sugar doubled above the recommended N rate of 200 kg N ha⁻¹.

Abstract

High rainfall, irrigation and large fertiliser nitrogen (N) inputs create conditions highly conducive for nitrous oxide (N₂O) emissions in sugarcane systems. A positive non-linear response of N₂O emissions to increasing N fertiliser rates has been reported from a range of cropping systems, where N inputs exceed crop requirements. However, the response of N₂O emissions to N fertiliser rates in sugarcane remains largely unclear. This study monitored N₂O emissions in a tropical sugarcane system in Australia to examine the response of N₂O emissions to increasing N fertiliser application rates. Soil-borne N₂O emissions were measured at high temporal resolution with an automated monitoring system across four N fertiliser treatments (0, 150, 200 and 250 kg N ha⁻¹). Furrow irrigation and rainfall events triggered large N₂O emissions with peak N₂O emissions > 100 g N ha^-1 d⁻¹. Nitrous oxide emissions peaked when high soil nitrate (NO₃-) content (> 40 kg N ha⁻¹) coincided with water-filled pore space (WFPS) near saturation. Emissions of N₂O in the first three months after N fertilisation accounted for 74–96 % of total N₂O emitted over the year. Cumulative N₂O emissions increased exponentially with N fertiliser rates, ranging from 0.29 to 3.06 kg N ha⁻¹. Emission factors (EF) ranged from 0.69 to 1.11 %, exceeding the IPCC default of 1 % at 250 kg N ha⁻¹ of N rate. The amount of N₂O emitted per kg sugar yield doubled above the recommended N rate of 200 kg N ha⁻¹. The exponential response of N₂O to fertiliser N suggests that N₂O emissions from intensive sugarcane systems with high fertiliser N inputs are considerably larger than currently assumed and that a non-linear model may be more appropriate for estimating N₂O emissions from sugarcane soils. The steep increase in N₂O intensity demonstrates environmental inefficiency at high fertiliser N rates, emphasising the importance of avoiding excessive N fertiliser application in tropical sugarcane systems from both agronomic and environmental perspectives.

Introduction

Nitrous oxide (N₂O) is a long-lived atmospheric trace gas with a global warming potential 265 times higher than that of carbon dioxide (CO₂) over a 100-year period (IPCC, 2014) and is the primary ozone-depleting substance of the 21 st century (Ravishankara et al., 2009; Portmann et al., 2012). Croplands are globally responsible for 66 % of total anthropogenic N₂O emissions and these emissions are predicted to double by 2050 (Davidson and Kanter, 2014). In soils, N₂O is mainly produced by the microbial processes of autotrophic nitrification, the oxidation of ammonium (NH₄⁺) to nitrate (NO₃⁻), and heterotrophic denitrification, the reduction of NO₃⁻ to N₂O and dinitrogen (N₂) (Butterbach-bahl et al., 2013). These processes are driven by soil moisture (Friedl et al., 2016), temperature (Schaufler et al., 2010), organic carbon (C) availability (De Rosa et al., 2017; Liyanage et al., 2020; Pandeya et al., 2020), pH (Šimek and Cooper, 2002) and NH₄⁺ and NO₃⁻ availability (De Rosa et al., 2020). Fertiliser nitrogen (N) inputs therefore increase the magnitude of N₂O emissions from cropping systems and are estimated to account for 54 % of the total increase of global N₂O emissions in the recent decade (Tian et al., 2019).

The IPCC default emission factor (EF) for N₂O assumes 1 % of applied N fertiliser lost as N₂O (IPCC, 2006) and thus a linear increase of N₂O emissions with N fertiliser rates. However, non-linear responses of N₂O emissions to N fertiliser rates were demonstrated in various cropping systems such as cotton (Scheer et al., 2016), maize (McSwiney and Robertson, 2005; Song et al., 2018), rapeseed (Ruser et al., 2017) and wheat (Millar et al., 2018), and confirmed in meta-analyses across multiple cropping systems (van Groenigen et al., 2010; Philibert et al., 2012; Rees et al., 2013; Shcherbak et al., 2014). A larger global N₂O EF of 2.3 % and non-linear response to high levels of N input were demonstrated also at the global scale (Thompson et al., 2019), requiring adjustments in the current assumption of fixed EFs. The non-linear response of N₂O emissions has been attributed to excessive N applications, exceeding the N demand of the crop. Adjusting N fertiliser inputs to crop needs therefore has the potential to mitigate N₂O emissions with no or little yield penalty (McSwiney and Robertson, 2005; Davidson and Kanter, 2014). To this end, N₂O intensity as a measure of ‘yield-scaled’ N₂O emissions is a useful indicator when evaluating N₂O emissions in an agronomic context (van Groenigen et al., 2010; Kim and Giltrap, 2017). In theory, N₂O intensity is high at very low N rates due to significant yield penalty, decreases with the increasing N rates as yield increases, and then increases steeply after reaching the yield potential, indicating the optimum N rate with the lowest N₂O intensity (Kim and Giltrap, 2017; Russenes et al., 2019). However, the response curve of N₂O intensity to N fertiliser rates is uncertain for many cropping systems. Deriving the response of both N₂O emissions and intensity to N rates is essential for fertiliser N rate recommendations balancing crop production and environmental goals in agricultural systems.

Substantial annual N₂O emissions have been reported in tropical sugarcane systems, exceeding 10 kg N ha⁻¹ (Wang et al., 2012; 2016a; 2016b). Emissions of N₂O from sugarcane systems account for 2–5 % of the applied N in Australia (Thorburn et al., 2010) and 0.2–3.0 % in Brazil (Pitombo et al., 2017; Bento et al., 2018; Bordonal et al., 2018; Borges et al., 2019), which is far in excess of other rural industries (Denmead et al., 2010). The high N₂O emission can be attributed to high soil moisture regimes, soil C availability and fertiliser N rates (Dalal et al., 2003). Intensive rainfall and high temperature, which are typical in tropical regions, together with irrigation practices, lead to large N₂O fluxes after fertiliser application (Allen et al., 2010; Wang et al., 2016a). High soil organic C content can cause extremely high N₂O emissions by providing the sources of energy for denitrifying organisms (Denmead et al., 2010; Thorburn et al., 2013; Wang et al., 2016b). The common practice of using retaining crop residues (i.e. trash blankets) can also supply additional sources of labile C to the soil and retain the soil moisture, promoting larger denitrification rates and in turn, N₂O emissions (Weier et al., 1998; De Oliveira et al., 2016). Sugarcane farms typically receive large amounts of N fertiliser and growers reaction to uncertainty in crop growth has lead to over-application of N fertiliser exceeding the recommended rates (Calcino et al., 2010), reported as much as 300 kg N ha⁻¹ in intensive production regions in Australia (Reading et al., 2019). The current fertiliser N rate guidelines in the Australian sugar industry range from 140 to 180 kg N ha⁻¹ for plant crops and 160–220 kg N ha⁻¹ for ratoon crops as baseline N application rates, aligning with the district yield potential (Schroeder et al., 2010b). The farm-scale allocation of N is still determined by growers priorities, which means that individual paddocks may still receive > 220 kg N ha⁻¹. Fertiliser N inputs in sugarcane systems in India and China were reported to range from 150 to 400 kg N ha⁻¹ (Robinson et al., 2011) and 400–800 kg N ha⁻¹ (Li and Yang, 2015), respectively. The high N₂O emissions from sugarcane systems and the wide range of EFs reported highlight the need to derive a response curve of N₂O emissions across different N fertiliser rates.

Few studies investigated the response of N₂O emissions to a wide range of N fertiliser rates in sugarcane: In Australia, published field trials on N₂O emissions have covered no more than two non-zero N rates, and have been limited to160 kg N ha⁻¹ (Wang et al., 2016b) and 200 kg N ha⁻¹ (Allen et al., 2010). In Brazil, trials covered only a period of 50 days (Signor et al., 2013) or were limited to N rates < 180 kg N ha⁻¹ (Degaspari et al., 2020). Studies with only a few and lower N input levels are less likely to determine the exact shape of the N₂O response curve to N rates (Shcherbak et al., 2014), limiting our quantitative understanding of both the agronomic and environmental impact of N fertiliser use in intensive sugarcane production. Therefore, this research monitored N₂O emissions from a tropical sugarcane system in Australia across four N fertiliser rates up to 250 kg N ha⁻¹ over the entire sugarcane growing period to (i) examine potential non-linearity of N₂O emissions in response to N rates and (ii) demonstrate the response of N₂O intensity to fertiliser N rates.