The short-lived inhibitory effect of Brachiaria humidicola on nitrous oxide emissions following sheep urine application in a highly nitrifying soil

Background: Brachiaria humidicola (Bh) has the ability to produce biological nitrification inhibitors (NIs) and release NIs from the root to the soil. Aims: TocomparetheeffectsofgrowingBhwith Brachiariaruziziensis (Br,whichisnotable toproduceNIs)onsoilnitrogen(N)dynamics,Ngasesandcarbondioxide(CO 2 )emissions and nitrifiers and denitrifiers following sheep urine application, a laboratory incubation was conducted in a He/O 2 continuous flow denitrification system (DENIS). This incubation was conducted in the absence of light. Hence the measured effects of Bh and Br on N cycling were the residual effect of biological NIs released into the soil prior to the incubation and released via root death. Methods: The treatments were: (1) Bh with water application (Bh + W); (2) Bh with sheep urine (Bh + U); (3) Br with water application (Br + W); (4) Br with sheep urine (Br + U). Results: Results showed that soil NO 3– concentration increased significantly in the soil with sheep urine application after the incubation. Soil nitrous oxide (N 2 O) and nitric oxide (NO) emissions increased immediately after the sheep urine application and peaked twice during the incubation. Cumulative emissions for the first peak were significantly lower from the Bh + U treatment (0.054 kg N ha –1 ) compared with the Br + U treatment (0.111 kg N ha –1 ), but no significant differences were observed in the total cumulative N 2 O and NO emissions between the Bh + U and Br + U treatment at the end of the incubation. Sheep urine addition did not affect the AOA, nirS and nosZ gene copies, but significantly increased the AOB gene copies after the incubation. Conclusions: We conclude that the residual effect of Bh to mitigate N 2 O emissions in a highly nitrifying soil is short-lived.


INTRODUCTION
Nitrification and denitrification are key processes of the soil nitrogen (N) cycle. Nitrification is a two-step microbially mediated process carried out by chemo-autotrophic nitrifying bacteria, first oxidising ammonium (NH 4 + ) to nitrite (NO 2 -) which is further oxidised to nitrate (NO 3 -) (Firestone and Davidson, 1989). During the nitrification and subsequent denitrification, other gaseous forms of N are produced and lost from agricultural soils, such as nitrous oxide (N 2 O), nitric oxide (NO) and dinitrogen (N 2 ). N 2 O has been attributed to nitrification, denitrification and nitrifier denitrification processes depending on the soil environmental conditions, such as water-filled pore space (WFPS), O 2 availability, soil pH and temperature (Bateman and Baggs, 2005;Lai et al., 2019;Loick et al., 2016;Wrage et al. 2005). Some studies present NO emitted from soils during nitrification process (Caranto & Lancaster, 2017;Kang et al., 2020;Wang et al., 2016). However, denitrification can also be a major source of NO from soils at high water content and/or under the presence of a carbon (C) source (Ji et al., 2020;Loick et al., 2016;Wu et al., 2017), while N 2 is the final product of denitrification (Knowles, 1982).
Previous studies have focused on the effects of pure inhibitory compounds identified from the pasture grass or the root exudates of Bh on soil NH 4 + transformation and N 2 O emissions (Gopalakrishnan et al., 2009;Meena et al., 2014;Subbarao et al., 2008). While experiments have been conducted to explore nitrification inhibition and N 2 O emissions from soil planted with Brachiaria grasses, including pasture that receive bovine urine deposition (Byrnes et al., 2017;Simon et al., 2020),  (Jones et al., 2004) and not necessarily as a typical tropical soil where Brachiaria species would be grown.

Establishment of BH and BR
To investigate the residual effect of Bh and Br on soil nitrification, greenhouse gas emissions (GHG, N 2 O and CO 2 ), NO and N 2 emissions and nitrifiers and denitrifiers after sheep urine application, two varieties of Brachiaria species (Bh and Br) were sown separately in pots containing the field soil. Seeds of Bh and Br were germinated on wetted tissue paper in an incubator (20 • C). Then 1.7 kg field fresh soil was added to each pot (diameter: 15 cm; depth: 15 cm) at the same bulk density as the soil at the field site (1.6 g cm -3 ) (Marsden et al., 2016), and 10 germinated seeds were placed onto the soil surface before covering with 100 g soil. There were 12 pots in total, six pots were grown with Bh and six pots with Br. To stimulate grass growth, the plants were cut to 2 cm above the soil level on day 33 and day 75. At the same time, the equivalent of 25 kg N ha -1 as (NH 4 ) 2 SO 4 was added to each pot 3 days after each cut to promote the release of the inhibitory compounds (Subbarao, Wang et al., 2007). Note that 50 mL of tap water was added to each pot twice per week to maintain plant growth prior to the incubation experiment. The establishment of Bh and Br was from the beginning of July to the end of November. To stimulate the growth of the tropical grasses, the lights above the plots in the greenhouse were on from October until the end of the cultivation. On day 150 after sowing, the plants and soils were harvested for the incubation experiment (described below).

Experimental setup
The 23-day incubation experiment was conducted in the Denitrifica- urine used in this experiment had been collected from six Welsh Mountain ewes that had been grazing a permanent pasture at the same site the soil was collected from. The urine had been frozen immediately after collection to avoid N losses during storage. The sheep urine was defrosted the day before application to the soil cores, and the individual urine samples (n = 6) were pooled and mixed to generate one urine source (total C, 25.3 g L -1 ; total N, 11.7 g L -1 , NH 4 + -N, 1.09 mg L -1 ; NO 3 --N, 3.09 mg L -1 ) of which 670 mg N kg -1 dry soil (equivalent to 374 kg N ha -1 ) was added in the treatments.
The incubation experiment followed a similar approach to previous experiments using this DENIS (Loick et al., 2016;Wu et al., 2017).
Briefly, to remove the native N 2 from the soil cores and the headspace, the soil cores were flushed from the base at a flow rate of 30 mL min -1 for 48 h using a mixture of He/O 2 (80:20), with the outlet flow from each chamber directed to a number of gas detectors. Once the N 2 , N 2 O and NO concentrations had reached very low levels, the airflow was decreased to 12 mL min -1 to measure the baseline emissions before being switched from the flow through the base to a flow over the soil surface. The sheep urine and water amendments were contained in sealed stainless-steel vessels above the lid of each incubation vessel. In previous protocols, these amendment vessels are usually flushed with He/O 2 (80:20) to remove N 2 (Cárdenas et al., 2003).
However, in this experiment, the vessels containing the urine and water were not flushed with He/O 2 , to avoid the N losses (via NH 3 volatilisation) from the sheep urine. After the urine and water had attained room temperature, the amendments were applied to the soil by opening the ball-valve connecting the two vessels. At the start of the soil incubation, the soil moisture content was increased to 65% WFPS to optimise conditions for nitrification (Mosier et al. 1996), taking the volume of the urine or water amendments into account. The temperature of the vessels was maintained at 15 • C during the flushing phase and the 23-day incubation period after the urine and water applications.

Soil sampling and analysis
During the incubation, the system was totally sealed, with all the soil gases displaced initially via mix of He/O 2 (80:20) passed through the soil from below and the outlet flow from each chamber was directed to a number of gas detectors. Thus, fresh soil samples were only collected for analysis before the sheep urine application and at the end of the incubation period. Soil characteristics before sheep urine application and the after the incubation are presented in Table 1. Soil moisture content was measured after oven drying (105 • C, 24 h), and the soil organic matter was determined by loss on ignition of dried soil in a muffle furnace (450 • C, 16 h) (Ball, 1964). Total soil C and N concentrations were determined on milled oven dried soil samples using

Gas sampling and analysis
The airflow from each vessel was automatically directed to a valve that

Statistical analysis
One-way analysis of variance (ANOVA) followed by the LSD test at 5% confidence was used to determine the effect of Bh and Br on soil NH 4 + and NO 3 concentrations, cumulative gas emissions (N 2 O, NO, N 2 and CO 2 ) and gene abundance (AOA, AOB, nirK, nirS, nosZ) at the start (day 0) and end (day 23) of the incubation, respectively. All statistical analyses were performed in SPSS Statistics 25.0 (IBM Inc., Armonk, NY).

Soil ammonium and nitrate concentrations
At the start of the incubation, there were no significant differences between all the treatments (Bh + W, Bh + U, Br + W, Br + U) for the soil NH 4 + and NO 3 concentrations, with average concentrations of 3.1 (ranging from 2.7 to 3.3 mg kg -1 soil) and 2.7 (ranging from 1.8 to 3.7 mg kg -1 soil) mg kg -1 soil, respectively (Table 1). In the Bh + W and Br + W treatments, after the 23-day incubation the NH 4 + concentration decreased (Bh + W, 3.3 to 1.3 mg kg -1 soil; Br + W, 3.1 to 0.15 mg kg -1 soil) and NO 3 increased (Bh + W, 3.7 to 16.0 mg kg -1 soil; Br + W, 2.8 to 17.3 mg kg -1 soil). Note that 23 days after the sheep urine application, there was a small increase in the NH 4 + concentration in the urine treatments (Bh + U, from 2.7 to 3.2 mg kg -1 soil; Br + U, from 3.3 to 3.6 mg kg -1 soil) and a large increase in the NO 3 concentration in the same treatments (Bh + U, from 1.8 to 235.7 mg kg -1 soil; Br + U, from 2.6 to 213.9 mg kg -1 soil).

Gas emissions
3. was significantly higher than that in the Bh + U (0.05 kg ha -1 ) treatment, although no significant differences were observed in the cumulative N 2 O emissions for the entire 23 days incubation between the Bh + U and Br + U treatments (Table 3). The cumulative N 2 O emissions in the Bh + W and Br + W treatments were significantly lower than that from both urine treatments during both the first peak period and the whole incubation period.

Nitric oxide
The pattern of NO emissions was similar to the N 2 O emissions for all treatments during the 23 days incubation, with the exception that the maximum NO fluxes in the sheep urine application treatments occurred during the second peak on day 14-16 ( Figure 1b). The first peak of NO emissions appeared 7.0 and 10.6 h after the urine application in the Bh + U and Br + U treatments, respectively, which was a little later than the peak time of maximum N 2 O emissions (3.6 and 5.3 h, respectively) reaching values up to 3 g N ha -1 d -1 . Cumulative NO emissions in the treatments with the sheep urine application including the two peaks (Bh + U, 0.114 kg N ha -1 ; Br + U, 0.103 kg N ha -1 ) were significantly higher than those in the water only treatments (Bh + W, 0.007 kg N ha -1 ; Br + W, 0.003 kg N ha -1 ). Nevertheless, no significant differences in NO emissions were observed between the Bh + U and Br + U treatments, or the Bh + W and Br + W treatments during the first peak period or for the whole incubation period. The second NO peak was broader than the initial one (reaching up to ≈ 8 g N ha -1 d -1 ) and had not reached background values at the end of the incubation, but clearly showed fluxes were decreasing from day 16 onwards.

Nitrogen gas
Fluxes of N 2 were low and decreased continuously from the start of the incubation (data not shown), indicating incomplete flushing of the vessels with contribution of the N 2 that entered the DENIS when nonflushed (He/O 2 ) urine and water were applied to the soil. Soil-borne N 2 emissions were not observed during the incubation, as expected, as soil moisture conditions were managed to favour nitrification (65% WFPS) ).

Carbon dioxide
In the Bh + U and Br + U treatments, the CO 2 emissions increased rapidly and peaked at 10.8 h after the urine application (similar to the NO peak in the urine treatments), with the maximum fluxes of 207.2 and 198.9 kg C ha -1 d -1 , respectively (Figure 1c). The CO 2 emissions decreased afterwards and remained stable (less than ca. 30 kg C ha -1 h -1 ) from day 3.5 to end of the incubation in the Bh + U and Br + U treatments. After the incubation, the cumulative CO 2 emissions in the soil under Br treatments were significantly lower than those in the soil under Bh treatments, following the series: Br + W < Bh + W < Br + U < Bh + U, with the cumulative fluxes of 333.5, 428.5, 654.6, 768.5 kg C ha -1 , respectively (Table 3).

Nitrifiers and denitrifiers gene copies
At the start of the incubation (day 0), there were no significant differences in the AOA, AOB, nirK, nirS and nosZ gene copies between the TA B L E 3 Cumulative emissions of N 2 O and NO (in kg N ha -1 ) and CO 2 (in kg C ha -1 ) after 23 days incubation and during the first peak period Values represent means ± SEM. Different letters indicate a significant difference between treatments (n = 3, p < 0.05). different treatments (Figure 2). After the incubation (day 23), no significant differences were observed in the AOA, nirS and nosZ gene abundance between the treatments with the sheep urine application and without urine application (Figure 2A, D and E). The sheep urine application increased the soil AOB and nirK gene copies at the end of the incubation ( Figure 2B, C). The AOB gene copies in the Bh + U treatment (7.7 × 10 6 copies g -1 soil) were significantly higher than that in the Br + U treatment (4.7 × 10 6 copies g -1 soil). The nirK gene copies in the Br + W (2.1 × 10 4 copies g -1 soil) were significantly lower than other treatments, but no significant differences were observed in the nirK gene copies between the Bh + W, Bh + U and Br + U treatments (3.3 × 10 4 , 5.0 × 10 4 , 3.7 × 10 4 copies g -1 soil, respectively).

Effect of Bh and Br on soil NH 4 + -N and NO 3 --N concentrations
The decrease of NH 4 + and increase of NO 3 in the treatments without sheep urine application was caused by the nitrification of residual soil NH 4 + . In the treatments with sheep urine application, the slight increase of NH 4 + and marked increase in NO 3 -(over 200 mg N kg soil -1 ) were caused by the hydrolysis of urea and further nitrification of the NH 4 + from the urine-N applied (Byrnes et al., 2017). It was expected that soil with Bh retained significantly higher NH 4 + and lower NO 3 concentrations than soil with Br after the incubation, due to the biological NIs released from its (Bh) root to suppress the transformation of NH 4 + to NO 3 - (Gopalakrishnan et al., 2009;Nuñez et al., 2018;Subbarao, Rondon et al., 2007). However, no significant differences were observed in the soil NH 4 + and lower NO 3 concentrations between the Bh and Br treatments in this study (Table 1).
Previous studies reported that soil applied with different amount of root exudates or compounds (which have been identified as biological NIs) from Bh retained higher soil NH 4 + and lower NO 3 concentrations compared with the bare soil treatments (Nuñez et al. 2018;Subbarao et al. 2006Subbarao et al. , 2008. Ma et al. (2021) found that soil applied with different concentrations of biological NIs (linoleic acid and linolenic acid) only decreased soil NO 3 concentration but did not affect the soil NH 4 + concentration due to the nitrification inhibition and/or N immobilisation. As for the effects of different Brachiaria species on soil nitri-fication, Castoldi et al. (2013) suggested that the levels of NH 4 + and NO 3 determined in the soil were similar among the Brachiaria species.
This is consistent with the results in this study and also supported by the study by Castoldi et al. (2017), in which no significant differences were observed in the soil NH 4 + and NO 3 concentrations between  (Gopalakrishnan et al. 2009;Ma et al. 2021;Sun et al. 2016). The low stability of biological NIs released from Bh may be also one reason for the unexcepted results in this study. Ma et al. (2021) confirmed that biological NIs (linoleic acid and linolenic acid) identified from the shoot tissue of Bh were much more rapidly mineralised than synthetic NIs (such as DCD, less than 5% of mineralisation rate even after 40 days incubation) in a highly nitrifying soil, reaching 40% in about 10 days incubation.

Effect of Bh and Br on soil N-gas and CO 2 emissions
N 2 O and NO are known products of both nitrification and denitrification processes, which dominate under different optimal soil environment conditions such as soil moisture (Loick et al., 2016;Wu et al., 2017), pH (Robinson et al., 2014), temperature (Lai et al., 2019), O 2 availability (Senbayram et al., 2019;Zhu et al., 2013) and C availability (Miller et al., 2008;O'Neill et al., 2020). At the beginning of the incubation experiment, the initial soil water content was set as 65% WFPS which would have favoured nitrification of the NH 4 + from the hydrolysed urea in the urine treatments causing the initial observed N 2 O and NO emission peaks (first smaller peak). It is also supported by the study of Loick et al. (2021), in which nitrification was contributing the most to N 2 O emissions at 70% WFPS. In addition, the initial CO 2 peak coincided with those of N 2 O and NO, as a result of the amendment application, and provides evidence of aerobic respiration (Lee et al., 2011). The duration of this peak is similar to the first N 2 O and NO peaks. Error bars represent standard error of the mean (n = 3). Different letters indicate significant differences between treatments at day 0 (lower case) and day 23 (upper case), respectively (p < 0.05) Soil grown with Bh is assumed to have lower cumulative N 2 O and NO emissions than that with Br due to the high BNI capacity in Bh (Gopalakrishnan et al., 2007;Subbarao et al., 2008). In this study, the cumulative N 2 O in the Bh + U treatment during the first peak was significantly lower than that in the Br + U treatment, which may be due to the nitrification inhibition caused by the biological NIs released from the Bh as previously reported (Meena et al., 2014;Subbarao et al., 2006;Subbarao, Rondon et al. 2007

Effect of Bh and Br on Nitrifiers and denitrifiers
Synthetic NIs, such as DCD and DMPP, have been confirmed to inhibit the AOA and/or AOB genes copies, which play an important role in controlling the nitrification rates and dominate at different conditions (Chen et al., 2015;Li et al., 2019;Shi et al., 2016). NIs have also been shown to inhibit denitrifying microbes, nirS and/or nirK and/or nosZ and/or narG (Li et al., 2019;Shi et al., 2017;Zhou et al., 2018). The biological NI, 1,9-decanediol (identified from rice), has also been shown to suppress the nitrification through impeding both AOA and AOB, when applied at high concentrations (≥ 5 00 mg kg -1 soil) (Lu et al., 2019).
A study conducted by Gopalakrishnan et al. (2009) also suggested that biological NIs released by the roots of Bh inhibited nitrifying bacteria, but did not negatively affect other major soil microorganisms. In this study, the controls (Bh + W and Br + W), did not influence the AOA, nirS and nosZ gene copies, but soil with Bh (with high BNI capacity) with sheep urine application significantly increased the AOB gene copies (responsible for the oxidation of NH 4 + ) compared with Br ( Figure 2).
The AOA and AOB gene copies were not lower in the Bh treatments than Br treatments as excepted, which may be because biological NIs inhibit nitrification rates by reducing the cell-specific activity of AOA and/or AOB, rather than affecting ammonia oxidiser populations, as well as non-target soil microorganisms or functions (Kong et al., 2016).
In order to retain air-tight seals throughout the incubation for the measurement of soil derived N 2 emissions, soil samples were not collected during the incubation period. A greater number of time points to explore the dynamics of soil NH 4 + and NO 3 -, as well as gene copies data during the incubation, or specific stable isotope approaches (such as 15 N labelling) would have helped to confirm the sources of gaseous N from soil grown with these two grasses, and nitrification inhibition mechanism of Bh. Gopalakrishnan et al. (2009) suggested that BNI by roots of Bh varies with soil type. In addition, soil moisture content is an important factor related to the release of N-gas emissions (Loick et al. 2016;Wu et al. 2017). The effects of Bh on soil nitrification and GHG emissions under different soil moisture levels and soil types could be explored in the future studies.

CONCLUSION
In this highly nitrifying soil, N 2 O emissions dominated rather than NO emissions, from the soil sown with Bh and Br after the sheep urine application. Bh inhibited N 2 O emissions during the first peak compared with Br, however, no significant differences were observed in the cumulative N 2 O and NO emissions between the Bh + U and Br + U treatments over the entire 23 days incubation period. And there were also no significant differences in the soil NH 4 + and NO 3 concentrations between the Bh and Br treatments. We conclude that the residual biological NIs may inhibit the nitrification temporarily, but not last long enough in a highly nitrifying soil.