Long-term effects of nitrogen fertilization on methane emissions in drained tropical peatland

Abstract

Nitrogen (N) fertilization improves crop productivity. However, the long-term effects of N application on methane (CH₄) emissions in drained peat soils, particularly under different hydrological conditions, remain poorly understood. Accurate quantification of CH₄ emissions from peatlands is essential for assessing carbon losses and formulating effective climate change mitigation strategies. This study was conducted to investigate the impact of N fertilization on CH₄ emissions and identify the main factors influencing CH₄ emissions from drained tropical peatlands. This study was conducted on an oil palm plantation in Sarawak, Malaysia, a randomized block design included four N fertilizer treatments: Control (0 kg N ha^− 1 yr^− 1) (T1); low (31.1 kg N ha⁻¹ yr⁻¹) (T2), moderate (62.2 kg N ha⁻¹ yr⁻¹) (T3), and high (124.3 kg N ha⁻¹ yr⁻¹) (T4). Soil CH₄ fluxes showed no statistically significant differences between treatments or across years, with emissions ranging from − 163.6 to 320.7 µg C m^− 2 hr^− 1 at T1, -86.7 to 285.8 µg C m^− 2 hr^− 1 at T2, -131.6 to 274.1 µg C m^− 2 hr^− 1 at T3 and − 125.7 to 185.9 µg C m^− 2 hr^− 1 at T4 (p > 0.05). Although ammonium sulfate fertilization did not significantly alter CH₄ emissions, its pronounced acidifying effect on soil pH, particularly at application rates above 62.2 kg N ha⁻¹ yr⁻¹ along with elevated sulfate (SO₄²⁻) inputs and nitrogen pools exceeding the critical threshold (> 400 ppm), likely suppressed methanogenic activity and constrained soil organic matter decomposition. Water-filled pore space (WFPS) influenced CH₄ emissions more than groundwater level (GWL), with the low GWL at the site limiting its impact. Increased WFPS (60–80%) reduced nitrate (NO₃⁻) through enhanced denitrification, lowering its inhibition on CH₄ production and thus increasing emissions. This study highlights the key role of soil moisture and nitrogen cycling in regulating CH₄ emissions in peatland.

Introduction

Methane (CH₄) is a potent greenhouse gas with a global warming potential 28 times that of carbon dioxide (CO₂) over a 100-year time horizon¹. Natural wetlands are among the largest contributors to global CH₄ emissions, accounting for approximately 20% of total annual emissions². Within this category, tropical peatlands, particularly those in Southeast Asia, are of significant concern due to their extensive carbon stores and sensitivity to land use change. These ecosystems are characterized by waterlogged, acidic, and nutrient-poor conditions that slow the decomposition of organic matter, leading to the accumulation of carbon-rich peat soils^3–5. In Southeast Asia, peatlands are predominantly found in Indonesia and Malaysia^6,7, where they support diverse ecosystems and play a crucial role in regional and global carbon cycling.

However, the expansion of oil palm and acacia plantations has led to significant land conversion, causing increased greenhouse gas emissions, land subsidence, altered hydrology and increased risk of peat fires⁸. Upon draining for agriculture, peat decomposition and mineralization increase, leading to higher emissions of CO₂ and nitrous oxide (N₂O)^9,10. In contrast, CH₄ fluxes can vary in drained peatlands, with peat swamp forests typically acting as net CH₄ sources and drained peatlands serving as either sources or small sinks¹¹. Draining peatland lowers the water table, promotes aerobic CH₄ oxidation and reduces CH₄ production^12,13. CH₄ dynamics in soil are regulated by two key microbial processes: methanogenesis, which produces CH₄ under anaerobic conditions, and methanotrophy, which consumes CH₄ under aerobic conditions^14,15. These processes are strongly influenced by hydrological factors, particularly groundwater level (GWL) and soil water-filled pore space (WFPS)^16–18. Among these, GWL is often the dominant control on CH₄ flux in tropical peatlands, as it directly affects redox conditions in the soil^19,20. In general, CH₄ emissions tend to be higher in areas with shallow groundwater tables that promote anaerobic conditions conducive to methanogenesis^19,20. However, in non-flooded peatland, WFPS becomes a more important factor¹⁷. For example, in oil palm plantations, WFPS has been observed to increase with plantation age, potentially enhancing CH₄ production in anaerobic microsites²¹. Under fully saturated conditions (100% WFPS), methanogenesis dominates, while at lower WFPS values (< 100%) CH₄ is produced in anoxic microsites, with some being oxidized¹⁵. While GWL and WFPS are key regulators of CH₄ dynamics in tropical peatlands, their interactive effects under varying plantation ages and management practices remain unclear, particularly as changes in environmental conditions and soil properties occur with palm growth.

While hydrological controls such as GWL and WFPS are recognized as key drivers of CH₄ emissions in tropical peatlands, nitrogen (N) fertilization may also significantly influence CH₄ fluxes. N fertilization is commonly practiced in oil palm plantations to maintain or increase productivity, yet its long-term effects on CH₄ emissions are still not fully understood. The impact of N additions on CH₄ emissions is complex, as it depends on the balance between CH₄ production and uptake. A decrease in CH₄ uptake due to suppressed methanotrophy, for instance, may result in increased net emissions²². Studies have shown that N fertilization can reduce the diversity and activity of both methanogens and methanotrophs, altering their ecosystem functions²³. Moreover, both the form and application rate of nitrogen fertilizer are important, as low rates tend to stimulate CH₄ emissions, whereas high rates may suppress them^24,25. Different types of N fertilizers can also have varying effects. For example, studies in rice fields have shown a stronger CH₄ response to urea compared to ammonium sulfate, likely due to the influence of sulfate (SO₄²⁻⁾ in suppressing methanogenesis²⁴. Increased ammonium sulfate application has been associated with decreased CH₄ emissions, likely due to competition for carbon substrates between sulfate-reducing bacteria and methanogens. Nitrogen inputs can also indirectly affect CH₄ emissions through other soil processes. For instance, methanogens compete with denitrifiers and iron reducers for labile carbon, and the presence of nitrate (NO₃⁻) or ammonium (NH₄⁺) can suppress CH₄ production by promoting these alternative pathways^26,27. Ammonium may also inhibit CH₄ oxidation by competing with CH₄ for the CH₄ monooxygenase enzyme in methanotrophs^28,29. In vegetated systems like oil palm plantations, N fertilizer can indirectly support CH₄ production by stimulating root growth and increasing organic exudates available to methanogens³⁰.

The influence of N fertilization on CH₄ emissions is also closely tied to hydrological conditions. In paddy soils, NO₃⁻ has been shown to inhibit both CH₄ uptake and production under varying GWLs^31–33. A meta-analysis of 33 studies by Banger et al.²⁴ found that N fertilization reduced CH₄ emissions in 67% of cases under intermittent drainage, largely due to enhanced CH₄ oxidation. However, under continuous flooding, this effect was much weaker, with only 16% of cases showing reduced emissions. These results highlight the importance of water management in modulating CH₄ responses to N inputs. Most studies on the effects of N fertilization on CH₄ fluxes have been conducted in paddy fields, but these findings cannot be directly applied to drained tropical peatlands, where distinct soil properties, plantation age, and management practices may influence CH₄ dynamics.

Despite growing interest in GHG emissions from tropical peatlands, there remains a significant knowledge gap regarding the long-term impacts of N fertilization on CH₄ emissions in drained peat soils, especially under varying hydrological conditions. As such, the quantification of CH₄ emissions in response to long-term N fertilization in drained tropical peatlands is critical for improving the accuracy of greenhouse gas inventories and developing effective mitigation strategies. Therefore, this study aims to investigate the long-term effects of nitrogen fertilizer application on CH₄ emissions from drained tropical peatlands, with a particular focus on how these effects are influenced by hydrological conditions such as groundwater level and soil moisture. It is hypothesized, first, that long-term nitrogen fertilizer application enhances CH₄ emissions from drained tropical peatlands by promoting conditions favorable to CH₄ production. Secondly, it is hypothesized that CH₄ emission levels increase with rising soil moisture, particularly as oil palm plantations age and WFPS increases.

Methods

Study site and experimental design

The study was conducted at an oil palm plantation in Sarawak, Malaysia (2°11’ 45.6” N, 111°51’11.7” E) (Fig. 1). The plantation was established in 2007 and planted with oil palms in 2008. Oil palms were planted in a triangular design with a density of 148 palms ha^{− 1} (8.5 m between palms). Triangular planting patterns are essential to provide for maximum penetration of sunlight, to maximize yield³⁴. The soil at the site is classified as Typic Haplofibrist according to the USDA soil classification system and has an average peat depth of approximately 9 m. The local climate in the region is humid and tropical, with air temperatures, ranging from 31.2 to 32.0 °C and relative humidity ranged from 71.8 to 74.1% (Table 1). Annual rainfall is 2,701 mm, with a wet season from November to March and a dry season from May to September in Sarawak. Further details about the study site can be found in Chaddy et al.³⁵.

Fig. 1.

(a) Location of the study site in Sibu, Sarawak, Malaysia, with the white polygon outlining the study area within an oil palm plantation on peatland. The aerial imagery of the study site was obtained from Google Earth (source: https://earth.google.com) while Sarawak and Sabah map and boundary were re-digitized from the ESRI World Imagery basemap (source: https://www.arcgis.com/home/item.html?id=10df2279f9684e4a9f6a7f08febac2a9) using QGIS software version 3.28.3 (source: https://qgis.org/). (b) Experimental design layout showing treatment plots (T1 = black, T2 = yellow, T3 = red, T4 = green)³⁵.

Table 1.

Environmental characteristics and soil properties at different N rates (mean ± SE).

Parameter	T1	T2	T3	T4
Air temperature (°C)	32.0 ± 0.3a	31.7 ± 0.3a	31.6 ± 0.3a	31.2 ± 0.2a
Relative humidity (%)	71.8 ± 1.2a	72.2 ± 1.1a	74.1 ± 1.2a	74.1 ± 1.1a
Soil temperature (°C)	28.3 ± 0.2a	28.2 ± 0.1a	28.2 ± 0.1a	28.1 ± 0.1a
GWL (cm) (below the peat surface)	53.2 ± 2.4a	54.4 ± 2.0a	60.7 ± 2.0a	60.6 ± 2.1a
WFPS (%)	77.4 ± 1.2a	77.3 ± 1.0a	77.2 ± 1.1a	76.8 ± 1.1a
Bulk density (g cm− 3)	0.25 ± 0.002ab	0.24 ± 0.002b	0.25 ± 0.002a	0.24 ± 0.002ab
Soil pH	3.43 ± 0.02a	3.41 ± 0.02ab	3.30 ± 0.02c	3.35 ± 0.02bc
NO3− (mg N kg− 1)	57.6 ± 5.6a	52.0 ± 5.7a	64.8 ± 6.0a	75.4 ± 9.2a
NH4+ (mg N kg− 1)	453.8 ± 11.3a	429.5 ± 11.2a	427.0 ± 9.4a	423.3 ± 10.8a
SO42− (mg kg− 1)	52.0 ± 9.7c	64.7 ± 9.1bc	100.1 ± 13.5ab	137.2 ± 17.5a
Total C (%)	55.8 ± 0.3a	55.5 ± 0.4a	55.0 ± 0.4a	55.4 ± 0.3a
Total N (%)	1.85 ± 0.02a	1.86 ± 0.02a	1.87 ± 0.02a	1.88 ± 0.03a
C: N ratio	30.4 ± 0.3a	30.1 ± 0.3a	29.8 ± 0.3a	29.9 ± 0.3a
LOI (%)	96.7 ± 0.1a	97.3 ± 0.1b	96.6 ± 0.2a	96.9 ± 0.2ab

Different letters (a, b,c) indicate significant differences among N rates at p < 0.05.

A randomized complete block design (RCBD) was used for this experiment, consisting of a control and three different nitrogen (N) rates, each replicated in three blocks. Each block covers an area of 1,564 m². The control (T1) received no N fertilization, while N treatments were as follows: low (T2, 31.1 kg N ha^{− 1} year^{− 1}), moderate (T3, 62.2 kg N ha^{− 1} year^{− 1}), and high (T4, 124.3 kg N ha^{− 1} year^{− 1}). In each block, 36 oil palms were designated, with gas sampling points located around four centrally positioned palms in each treatment plot. The blocks were separated by two to three rows of palms or by field drains to minimize interference. In this study, powdered ammonium sulfate was used as the N fertilizer at a baseline rate of 62.2 kg N ha^{− 1} year^{− 1} (T3) in accordance with the recommendations of Hasnol et al.³⁶. Fertilizer applications were scheduled four times a year, in March, June, September, and November, and were placed two meters away from the center of each palm. The experimental design and treatment plots were previously described in Chaddy et al.³⁵ (Fig. 1).

Soil CH₄ fluxes measurement and environmental variables

Measurements of soil CH₄ fluxes were carried out from January 2010 to December 2013 and resumed from January 2016 to December 2017. The closed chamber method described by Chaddy et al.³⁵ was used to determine the soil CH₄ fluxes. Stainless steel chambers with an internal diameter of 21 cm and a height of about 25 cm. Each chamber was equipped with a syringe connected to the outlet of the chamber lid via a silicone tube. Four chambers were installed in each block, 2 m away from each selected palm (i.e. at the fertilization point). The bottoms of the chambers were carefully inserted into the soil to a depth of about 3 cm. Gas samples were taken before closing the chamber and 10, 20 and 40 min after closing the chamber with the lid. Before closing the chamber, 20 mL gas sample from the headspace of each chamber was extracted using the syringe and transferred into an evacuated vial (20 mL) for CH₄ analysis. This first measurement was considered to time 0 min. Additional 20 mL gas samples were taken at each 10, 20 and 40 min after the chamber was closed, and transferred to evacuated vials. In the laboratory, the CH₄ concentration in soil was determined using a gas chromatograph with flame ionization detector (FID) (7890 A, Agilent). The gas flux (F; µg C m^{− 2} h^{− 1} for CH₄) was calculated using the following equation:

1

where p is the density of CH₄ (0.717 kg m^− 3), V is the volume of the chamber (m³), Sb is the area of the chamber floor (m²), ∆c/∆t is the rate of change of gas concentration in the headspace of the chamber over time while the chamber was closed (10^− 9 × m³ m^− 3 h^− 1), T is the average air temperature (°C) during the sampling period (0, 10, 20 and 40 min) and α is the ratio of the molecular weight of C to CH₄ (0.749). The annual cumulative soil CH₄ emissions were calculated from the monthly mean values using the following equation:

2

where Fi is the mean gas flux (kg C ha^− 1 day^− 1) between two sampling dates (i.e. for the time interval), Di is the number of days in the sampling interval and n is the sampling frequency.

Environmental parameters such as air temperature, soil temperature at a depth of 5 cm and relative humidity were recorded at the same time as the gas measurements. Groundwater level (GWL) below peat surface was monitored using a perforated polyvinyl chloride (PVC) pipe installed in the center of each block and precipitation was measured with a rain gauge.

Soil sampling and analysis

Undisturbed peat core samples were collected with a core ring (100 cm³) at each gas sampling point, specifically targeting depths of 0–5 cm. The actual volume (solid and liquid phases) of each undisturbed core sample was measured in the laboratory using a digital volumetric meter (DIK-1150, Daiki Rika, Saitama, Japan) to calculate the WFPS. The samples were then dried in oven at 105 °C for 48 h and reweighed to determine their bulk density (Bd).

Upon collecting core samples, additional soil samples were collected at a depth of 0–25 cm using a peat auger (Eijelkemp, The Netherlands). The soil samples were analyzed for their soil chemical properties. In the laboratory, the soil samples were air-dried, sieved through a 2-mm sieve, and stored at 4 °C for further analysis. Soil pH (H₂O) was measured for a 2:5 (w/v) soil-water suspension using a pH meter (Metrohm 827, Metrohm, Herisau, Switzerland). The air-dried soil was mixed with deionized water, filtered and the extracts were analyzed using an ion chromatograph (Methrohm 761 Compact IC) to determine soil NO₃⁻ and water-soluble SO₄²⁻. Air-dried soil was also mixed with 2 M potassium chloride (KCl) and filtered. The resulting extract was then analyzed for soil NH₄⁺ following the method proposed by Keeney and Nelson³⁷. The carbon (C) and nitrogen (N) content was analyzed with a CN analyzer (LECO TruMac 4060, LECO Corporation, Saint Joseph, MI, USA). At the same time, loss on ignition (LOI) was measured with a thermogravimetric analyzer (Leco TGA701, LECO Corporation, Saint Joseph, MI, USA). The loss on ignition (LOI) is a method used to estimate the organic matter content in soil³⁸.

Statistical analysis

A two-way ANOVA was performed with a significant level of p < 0.05 to assess differences in soil CH₄ fluxes between treatments and years. Tukey’s Honestly Significant Difference (HSD) test was used to compare the mean difference (p < 0.05) of specific variables between treatments. Linear and non-linear regression analyses were used to examine the relationships among CH₄ fluxes, environmental parameters, and soil properties. All statistical analyses were performed using SPSS Statistics software version 21 (IBM, New York, NY).

Results

Environmental variables and soil properties

Monthly variation in rainfall, GWL, WFPS and soil temperature are shown in Fig. 2. Rainfall showed significant seasonal variation, ranging from 37.0 to 389.1 mm during the dry season and from 144.8 to 482.1 mm during the wet season (Fig. 2a). The GWL pattern generally followed the rainfall pattern, ranging from 103.3 to 31.5 cm below the peat surface (Fig. 2a) and showed no significant differences between treatments (p > 0.05) (Table 1). Variation in GWL had no consistent effect on WFPS, as WFPS increased with palm age, ranging from 59% to 95.9% (Fig. 2b). There were no significant differences in WFPS between treatments (p > 0.05) (Table 1). Soil temperatures decreased with palm age in all treatments, ranging from 26.0 to 31.5 °C (Fig. 2c), with no significant differences between treatments (p > 0.05) (Table 1).

Fig. 2.

Monthly (a) rainfall and GWL below peat surface (b) WFPS, (c) soil temperature and (d) soil CH₄ fluxes of different N rates. Black arrows indicate N fertilizer application.

The physico-chemical properties of the soil under different N fertilizer applications are shown in Table 1. Among the parameters analyzed, soil pH, SO₄²⁻, and LOI exhibited statistically significant differences between treatments (p < 0.05). Significantly lower soil pH was observed at T3 (62.2 kg N ha⁻¹ year⁻¹) and T4 (124.3 kg N ha⁻¹ year⁻¹) compared with the control (T1). Relative to T1 (pH 3.43), soil pH decreased to 3.30 in T3 and 3.35 in T4. The use of ammonium sulfate as N fertilizer resulted in significantly higher soil SO₄²⁻ concentrations in T3 and T4, which were three times higher than in the control (p < 0.001). The effect of N fertilization on LOI was less pronounced, with only T2 showing a significantly lower LOI compared to the control (T1) (p < 0.05).

Soil CH₄ fluxes

Soil CH₄ fluxes measured during the study periods at T1, T2, T3 and T4 ranged from − 163.6 to 320.7 µg C m^− 2 hr^− 1 at T1, -86.7 to 285.8 µg C m^− 2 hr^− 1 at T2, -131.6 to 274.1 µg C m^− 2 hr^− 1 at T3 and − 125.7 to 185.9 µg C m^− 2 hr^− 1 at T4 (Fig. 2d). A slight decrease in soil CH₄ fluxes was observed from 2010 to 2013, ranging from − 72.8 to 191.8 µg C m^− 2 hr^− 1, followed by an increase between 2016 and 2017, with a range of values from − 163.6 to 320.7 µg C m^− 2 hr^− 1. Seasonal variations in CH₄ fluxes were not apparent, and seasonal effects on CH₄ fluxes were insignificant across all treatments (p > 0.05). As shown in Table 2, there were no significant differences in soil CH₄ fluxes between treatments and years. Although there were no significant differences between years within each treatment (p > 0.05), the soil CH₄ fluxes were generally decreasing from 2010 to 2013 and increasing between 2016 and 2017, as shown in Fig. 2d.

Table 2.

Soil CH₄ flux in different N rates (mean ± SE) and two-way ANOVA for soil CH₄ fluxes in response to N rates, year and their interaction.

Year	T1	T2	T3	T4	Average
	(µg C m− 2 hr− 1)
2010	90.2 ± 13.4	110.0 ± 11.5	77.2 ± 17.9	68.6 ± 11.0	86.5 ± 9.0
2011	65.8 ± 13.3	49.4 ± 18.4	52.1 ± 6.6	64.1 ± 14.2	57.9 ± 4.1
2012	66.2 ± 17.2	74.4 ± 19.6	39.5 ± 18.7	42.2 ± 16.8	55.6 ± 8.7
2013	22.7 ± 14.5	23.5 ± 14.9	19.5 ± 9.5	32.4 ± 17.6	24.5 ± 2.8
2016	40.6 ± 18.8	50.1 ± 17.3	51.9 ± 27.7	59.1 ± 15.4	50.4 ± 3.8
2017	68.8 ± 41.2	78.9 ± 29.8	67.2 ± 29.6	57.7 ± 29.2	68.1 ± 4.4
Mean	59.1 ± 9.7	64.4 ± 12.2	3.96 ± 8.3	4.21 ± 5.6

Source	d.f	Mean Square	F-value	p-value
Year	5	20092.91	4.17	0.001
N rates	3	2427.35	0.50	0.680
Year * N rates	15	1576.37	0.33	0.992
Error	264	4818.46

d.f. = degrees of freedom; F-value = ratio of between-group variance to within-group variance; p-value = the statistical significance of the test result. Significance at p < 0.05.

Factors influencing soil CH₄ fluxes

In this study, no clear relationship was found between CH₄ fluxes and GWL below peat surface (p > 0.05) (Fig. 3a). In contrast, a quadratic relationship was found between soil CH₄ fluxes (y, in µg C m^− 2 hr^− 1) and WFPS (x, in %) (y = 0.2853 × ² – 44.855x + 1799.8, R² = 0.47), where CH₄ fluxes decreased as WFPS increased from 60 to 80%, but increased again when WFPS exceeded 80% (Fig. 3b). Soil CH₄ fluxes (y, in µg C m^− 2 h^− 1) also showed a quadratic relationship with soil temperature (x, in °C) (y = 19.247 × ² – 1077.7x + 15136, R² = 0.42), with fluxes rising markedly once soil temperature exceeded 29 °C (Fig. 3c). It was observed that soil temperature (y, in °C) was significantly influenced by both GWL (x, in cm) and WFPS (x, in %) (Figs. 3d, e), showing a negative linear relationship with GWL (y = − 0.0215x + 29.138, R² = 0.27, p < 0.05) and a quadratic relationship with WFPS (y = 0.0032 × ² – 0.5328x + 49.836, R² = 0.50). However, no clear relationship was found between WFPS and GWL (p > 0.05) (Fig. 3f).

Fig. 3.

Relationships between soil CH₄ fluxes and (a) GWL below the peat surface, (b) WFPS, and (c) soil temperature; as well as the relationship between (d) soil temperature and GWL below peat surface, (e) soil temperature and WFPS, and (f) the relationship between WFPS and GWL below the peat surface.

As shown in Fig. 4a and b, soil CH₄ fluxes (y, in µg C m^− 2 hr^− 1) showed no significant relationship with soil NH₄⁺ (p > 0.05) but exhibited a positive linear relationship with soil NO₃⁻ (x, in mg N kg^− 1) (y = 1.3379x + 37.618, R² = 0.25, p < 0.05). Soil NO₃⁻ concentrations (y, mg N kg^− 1) were not influenced by GWL below the peat surface (p > 0.05) (Fig. 4c), but were significantly affected by both WFPS and soil temperature, showing quadratic relationships with WFPS (x, in %) (y = 0.0795 × ² – 12.904x + 531.7, R² = 0.72) (Fig. 4d) and soil temperature (x, in °C) (y = 18.953 × ² – 1050.4x + 14563, R² = 0.68) (Fig. 4e). Soil NO₃⁻ decreased when WFPS ranged from 60 to 80%, but increased when WFPS exceeded 80% (Fig. 4d), a trend consistent with the relationship between CH₄ fluxes and WFPS (Fig. 3b).

Fig. 4.

Relationships of CH₄ flux with (a) soil NH₄⁺ and (b) soil NO₃⁻, and relationships of soil NO₃⁻ with (c) GWL below peat surface, (d) WFPS, and (e) soil temperature.

Discussion

Few studies have examined the effects of N fertilization on soil CH₄ fluxes in tropical peatlands, such as sago palm plantations¹⁶ and oil palm plantations³⁹. In this study, no significant effects of N fertilization on soil CH₄ fluxes were observed across all ages of palms (Table 2), and the results do not support the first hypothesis. Similar results were reported by Watanabe et al.¹⁶, who found no effect of N fertilization on soil CH₄ fluxes in sago plantations on peat. In contrast, Melling et al.³⁹ observed an increase in CH₄ emissions following urea application in young oil palm plantations on peat. Compared to urea, ammonium sulfate had the potential to acidify the soil through a process known as nitrification and the release of hydrogen ions (H⁺)⁴⁰. Soil pH was significantly lower at T3 (62.2 kg N ha⁻¹ year⁻¹) and T4 (124.3 kg N ha⁻¹ year⁻¹) compared to T1 (0 kg N ha⁻¹ year⁻¹) (p < 0.001) (Table 1). This reduction in soil pH possibly reflects the acidifying effect of ammonium sulfate, particularly at application rates above 62.2 kg N ha⁻¹ year⁻¹, which may have enhanced peat acidification. Low soil pH caused by this fertilizer can potentially inhibit the activity of methanogens, as many methanogens are sensitive to low pH conditions⁴¹ and their activity may be suppressed in such an environment.

The use of ammonium sulfate as nitrogen fertilizer in this study resulted in a significant increase in SO₄²⁻ concentration in the soil (p < 0.001) (Table 1). The addition of sulfate-containing fertilizers to soil can reduce soil CH₄ emissions^42–44, as the presence of high SO₄²⁻ ions in soil can suppress CH₄ formation by competing with methanogens for substrates^22,43. Although there was no significant effect of N rates on soil NO₃⁻ and NH₄⁺ in this study (p > 0.05) (Table 1), but high inorganic N concentration, especially NH₄⁺, was observed even in the control treatment (T1) after peat conversion (Table 1). This high NH₄⁺ concentration probably exceeded the critical value (> 400 ppm) for soil organic matter decomposition, which could explain the insignificant effects of N fertilization on CH₄ fluxes in this study (p > 0.05). This could explain the weak relationship between CH₄ fluxes and soil NH₄⁺ in soil (p > 0.05) (Fig. 4a). These considerations would have elucidated the reasons behind the lack of notable effects of ammonium sulfate on soil CH₄ emissions in this study. Due to inconsistent results by different N fertilizers, future investigations are important to explore the effects of different types of N fertilizers on CH₄ emissions in tropical peatlands as different types of N fertilizer would give different outcomes through alterations in soil chemistry, microbial community dynamics, and redox conditions. A comprehensive understanding of these effects will facilitate the identification of fertilizer practices that mitigate methane emissions while still supporting crop growth.

As shown in Fig. 2d, CH₄ fluxes increased with increasing age of palms and WFPS, thereby supporting the second hypothesis. A quadratic relationship was found between CH₄ fluxes and WFPS (y = 0.2853 × ² – 44.855x + 1799.8, R² = 0.47) (Fig. 3b), indicating that WFPS had a stronger impact on CH₄ fluxes than GWL. This dominance of WFPS in regulating CH₄ fluxes in drained tropical peatlands was also found by Swails et al.¹⁷ in an oil palm plantation. The CH₄ fluxes decreased when the WFPS increased from 60 to 80%, (Fig. 3b) which was accompanied by a decrease in NO₃⁻ within the same WFPS range (Fig. 4d). This range of WFPS mainly occurred during the early years of palm growth (2010–2012) (Fig. 2d). The decrease in NO₃⁻ when WFPS was increased from 60 to 80% suggests active NO₃⁻ reduction via denitrification, leading to a decrease in NO₃⁻ supply while methanogenesis was still suppressed by the availability of alternative electron acceptors, thereby delaying the onset of methanogenesis^18,45. This process led to the observed positive relationship between CH₄ fluxes and soil NO₃⁻ concentration (Fig. 4b). After the reduction of NO₃⁻ due to denitrification, the inhibitory effect of NO₃⁻ on CH₄ emission would decrease, leading to an increase in CH₄ production⁴⁶, once WFPS exceeded 80% (Fig. 3b). Soil NO₃⁻ acts as a secondary control on CH₄ emissions, as its influence on methanogenesis becomes apparent only after anoxic conditions are established. Under these conditions, NO₃⁻ serves as a more favorable electron acceptor than CO₂, allowing denitrifiers to outcompete methanogens for labile carbon and thereby suppress CH₄ production^14,26.

At higher WFPS (> 80%), the larger, more extensive root biomass of older palms would produce exudates that support methanogenesis and contribute to higher CH₄ fluxes. This is because the readily available carbon inputs from root exudates are more readily utilized by soil microbial communities, as discussed by Girkin et al.⁴⁷. However, this aspect was not explored in the current study. Thus, the influence of oil palm’s roots exudates on soil CH₄ fluxes in tropical peatlands should be considered for future investigations.

The water table plays an important role in soil CH₄ production^14,51. In wetland soils, CH₄ is produced by methanogens in the anaerobic zones of submerged soils and oxidized to CO₂ by methanotrophs in the aerobic zones. The balance between production and consumption determines whether CH₄ is released into the atmosphere¹⁴. Drainage of peatlands for agricultural and forestry purposes lowers the water table to support the required level of productive growth, which can reduce methanogenesis and increase methanotrophs, leading to a significant reduction in CH₄ emissions^13,53. As shown in Fig. 5, CH₄ emissions observed in oil palm plantations in this study as well as in previous studies are lower compared to those documented in peat forests. This discrepancy is attributed to fluctuations in the GWL, which means that in most cases a higher GWL in the forest resulted in greater CH₄ emissions than in the oil palm plantation, where the GWL was managed or controlled to a certain extent. However, soil CH₄ fluxes in this study did not respond well to changes in GWL, similar to the observations of Swails et al.¹⁷ in an oil palm on tropical peatland. According to a review paper by Hatano²⁰, CH₄ emission increases significantly when GWL increases above 20 cm below peat surface. In this study, the GWL remained consistently low, ranging from 103.3 to 31.5 cm below the peat surface (Fig. 2a), which explains its insignificant direct effect on CH₄ emissions. However, soil temperature was strongly influenced by both GWL and WFPS (Fig. 3d, e) and occurred as a key factor affecting CH₄ fluxes (Fig. 3c), consistent with previous findings that elevated temperatures stimulate CH₄ fluxes^50,54. This effect was most pronounced during the early stages of oil palm development when canopy cover was minimal. Under these conditions, drainage and root activity increased the availability of labile carbon, while higher temperatures accelerated microbial metabolism, enhancing the production of methanogenic substrates such as acetate, H₂, and CO₂. At the same time, this conditions also promotes microbial oxidation of reduced compounds, regenerating alternative electron acceptors such as NO₃⁻, Fe³⁺, and SO₄²⁻, which explained the positive relationship between soil temperature and NO₃⁻ (Fig. 4e). As alternative electron acceptors (NO₃⁻, Fe³⁺, SO₄²⁻) became depleted, methanogenic archaea dominated the anaerobic processes, resulting in higher CH₄ emissions during the early establishment phase. Future research should investigate how soil temperature, GWL, and WFPS interact to influence CH₄ fluxes in drained peatlands. Identifying hydrological thresholds and their influence on redox dynamics and temperature sensitivity will be important to improving predictions of CH₄ emissions under climate change and management scenarios.

Fig. 5.

Relationship between GWL below peat surface and annual cumulative CH₄ emissions in other literature. Forest^17,48–50; Oil palm plantation^17,49,51,52.

In this study, increasing WFPS from 60% to over 80% caused a decline in NO₃⁻concentration, likely due to enhanced denitrification, followed by increased CH₄ emissions once NO₃⁻ is depleted. These trends suggest a shift in dominant microbial processes under increasingly anaerobic conditions. However, a key limitation is the absence of direct measurements of redox potential (Eh), which restricts the ability to accurately identify the redox thresholds at which denitrifiers and sulfate-reducing bacteria, potentially stimulated by ammonium sulfate application are outcompeted by methanogens. Addressing this limitation in future work would enhance mechanistic understanding of greenhouse gas fluxes and support more accurate modeling under varying hydrological, different types of N fertilizers and redox conditions.

Conclusions

This study highlights the complex interactions between nitrogen fertilization, soil chemistry, and hydrological factors in influencing CH₄ emissions from tropical peatlands. Contrary to the hypothesis, nitrogen fertilizer had no significant effect on CH₄ emissions but changes in WFPS significantly affected the emissions through changes in soil N availability and microbial dynamics. This lack of effect can be attributed to the low soil pH, higher SO₄²⁻ content and larger N pools. Our results highlight the crucial role of WFPS in regulating CH₄ emissions compared to GWL. Future studies should investigate the effects of peat decomposition stages and oil palm root exudates on the availability of labile and recalcitrant carbon pools and their impact on CH₄ emissions. Incorporating nitrogen fertilizer variability and redox potential measurements will further clarify underlying biogeochemical controls. Addressing these limitations collectively will provide a more comprehensive understanding of methane dynamics, ultimately leading to more accurate predictions and effective mitigation strategies.

Author contributions

Literature collection, data collection and analysis were performed by Auldry Chaddy, Faustina Elfrida Sangok, and Sharon Yu Ling Lau. The first draft of the manuscript was written by Auldry Chaddy. Faustina Elfrida Sangok, Sharon Lau Yu Ling and Lulie Melling revised the draft. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This research was funded by the Federal Government of Malaysia and the Sarawak State Government.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.