Daily Variation of Soil Greenhouse Gas Fluxes in Rubber Plantations Under Different Levels of Organic Fertilizer Substitution

Abstract

It has been widely recognized that replacing chemical fertilizers with organic fertilizers (organic substitution) could significantly increase the long-term productivity of the land and potentially enhance resilience to climate change. Nevertheless, there is limited information on the accurate monitoring of soil greenhouse gas (GHG) fluxes at different levels of organic substitution in rubber plantations. Before accurate estimation of soil GHG fluxes can be made, it is important to investigate diurnal variations and suitable sampling times. In this study, six treatment groups of rubber plantations in the Longjiang Farm of Baisha Li autonomous county, Hainan Island, including the control (CK), conventional fertilizer (NPK), and organic substitution treatments in which organic fertilizer replaced 25% (25%M), 50% (50%M), 75% (75%M), and 100% (100%M) of chemical nitrogen fertilizer were selected as study objectives. The soil GHG fluxes were observed by static chamber-gas chromatography for a whole day (24 h) during both wet and dry seasons. The results showed the following: (1) There was a significant single-peak daily variation of GHGs in rubber plantation soils. (2) The soil GHG fluxes observed from 9:00–12:00 are closer to the daily average fluxes. (3) Organic fertilizer substitution influenced soil CO₂ and N₂O fluxes and had no significant effect on soil CH₄ fluxes. Fluxes of soil CO₂ and N₂O increased firstly and then decreased gradually when the substitution ratios exceeded 50% or 75%. (4) Soil CO₂ and N₂O fluxes were positively correlated with soil temperature and soil moisture, and CH₄ fluxes were negatively correlated with soil temperature and soil moisture in both wet and dry seasons. The study indicated that understanding the daily pattern of GHG changes in rubber forest soils under different levels of organic fertilizer substitution and the optimal observation time could improve the accurate assessment of long-timescale observation studies.

1. Introduction

Carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) are recognized as the primary greenhouse gases (GHGs) contributing to global warming [1]. Since 1850, concentrations of CO₂, CH₄, and N₂O have increased by 47%, 156%, and 23%, respectively, significantly accelerating global warming [2]. Agriculture, forestry, and other land use are significant contributors to anthropogenic GHG emissions, with agricultural activities alone estimated to account for 30–50% of global GHG emissions [3]. Soils are important sources and sinks of GHGs and play a crucial role in regulating their emissions through various biological processes [4]. These processes are influenced by multiple factors, including land use and management practices. As a result, climate-smart land use management is increasingly emphasized as a strategy for reducing emissions, and the precise monitoring of soil GHG fluxes, particularly from agricultural soils, is essential for effective climate action.

In addition to conventional agricultural activities, it is possible to engage in climate-smart land-use practices by intentionally combining trees with crops or livestock, which has been extensively researched over the past few decades. Rubber plantations have undergone rapid expansion across montane mainland Southeast Asia in recent years, with projections indicating an additional 8.5 million hectares within the next decade [5]. China ranks sixth globally in natural rubber production, with a total output of 0.65 million metric tons, and Hainan Island is the primary production site within China [6]. In response to increasing rubber demand, large quantities of chemical fertilizers (primarily nitrogen-based) are applied to rubber plantations to boost yields. However, these practices significantly degrade soil health, contribute to environmental pollution, and exacerbate GHG emissions [7]. In recent years, the Ministry of Agriculture and Rural Affairs of China introduced the ‘two reductions’ policy, aimed at reducing chemical fertilizer use to promote sustainable agricultural development. Consequently, organic fertilizers have become an important alternative for nutrient management in rubber plantations. Organic fertilizers could offer various nutrients essential for crop growth [8], and they contain abundant organic substances [9]. Although they enhance soil fertility and benefit crop growth, organic fertilizers can also increase GHG emissions, particularly CO₂, due to organic decomposition [10]. Studies have shown that while chemical fertilizers may inhibit CH₄ emissions [11], organic fertilizers tend to promote them [12]. Similarly, N₂O emissions are often higher with organic fertilizer, mirroring the trends observed with CO₂ and CH₄ emissions [13]. Although some efforts have been made to quantify the impact of replacing chemical fertilizers with organic alternatives on soil GHG emissions, the results have varied considerably among different agricultural soils [14,15]. Assessing the impact of organic fertilizer substitution on soil GHG emissions is an important aspect of assessing the ecological benefits of organic fertilizer substitution. Therefore, it is important to understand how soil GHG fluxes vary in rubber plantations under different levels of organic substitution, as it is conducive for evaluating the ecological benefits of using organic fertilizer to replace chemical nitrogen fertilizer (organic substitution).

Static chamber-gas chromatography has been widely used for the observation of GHG fluxes due to its simplicity and cost-effectiveness. However, this method also has limitations, including its inability to capture fine-scale spatial variation and potential biases in measurement under certain environmental conditions [16]. Moreover, it is particularly important to account for the daily variation in soil GHG fluxes within the study area, as fluctuations can be significant and may influence the representativeness of the measurements [17]. Previous research has highlighted optimal observation times for various crops and ecosystems, underscoring the need for site-specific sampling protocols. For example, Zhu et al. determined that the CO₂ and N₂O fluxes from 9:00–11:00 were representative of the daily fluxes in rice soil in Ningxia Province, China [18]. Similarly, Wang et al. found that soil CO₂ fluxes at 9:00 and 21:00 closely matched the daily mean value in tobacco fields in Zhengzhou City [19]. Bruno et al. reported that 9:00–10:00 and 21:00–22:00 were the optimal sampling times for soil N₂O in farmland in the United Kingdom and grasslands in Brazil, respectively [20]. Additionally, Cueva et al. observed that 17:00–19:00 and 20:00–21:00 were suitable sampling times for soil GHG emissions, contrasting with morning sampling times used in other studies [21]. Therefore, to accurately evaluate GHG emissions, determining the daily patterns of soil GHG fluxes and optimal sampling times in different ecosystems remains necessary.

Therefore, this study analyzed daily variations in GHG fluxes from rubber plantation soil under different organic substitution treatments using the static chamber method coupled with gas chromatography. The specific objectives of this study were (1) to characterize daily GHG flux patterns in rubber plantation soils under varying levels of organic substitution; (2) to explore how GHG emissions in rubber plantation soils change under levels of different organic substitution; (3) to identify the optimal sampling time for GHG fluxes under these conditions. This study aims to enhance the accuracy of GHG flux estimates in rubber plantation ecosystems, contributing to improved emission monitoring practices.

2. Materials and Methods

2.1. Study Area

The experimental area is located at Longjiang Farm in Baisha Li autonomous county, Hainan Island (19°22′35″ N, 109°17′41″ E, average elevation 120 m). The site features flat terrain, abundant rainfall, and a tropical monsoon climate, with a distinct wet season from May to November and a dry season from December to April. The average annual temperature is 23.4 °C, and the average annual precipitation is approximately 1960 mm, with roughly 80% of rainfall occurring during the wet season. The soil texture type of the experimental area is sandy clay loam [22]. The rubber plantation was planted in 2004 and rubber tapping was started in 2011. The rubber clones are PR107, and the spacing of the plants is 3 m × 5 m.

2.2. Experiment Design

The experiment was conducted on 7 October 2023 and 7 January 2024, covering both wet and dry seasons. Six organic substitution treatments were applied to assess their effects on GHG fluxes in rubber plantation soils: a control group (CK), conventional fertilization group (NPK), and organic substitution treatments in which organic fertilizer replaced 25% (25%M), 50% (50%M), 75% (75%M), and 100% (100%M) of chemical nitrogen fertilizer. Three replicates were set up for each treatment group, and five soil respiration rings were installed per replicate, totaling 72 soil respiration rings across all treatment groups (Figure 1).

To accurately capture soil GHG fluxes, monitoring was conducted continuously over 24 h periods using the static chamber-gas chromatography (SC-GC) method. Each soil respiration ring, made of PVC plastic tubes with a diameter of 0.20 m and a height of 0.10 m, was inserted 5–7 cm into the soil. A static chamber was placed atop each soil respiration ring to capture gas emissions. Initial gas samples were collected at 9:00 a.m., with subsequent samples taken 0, 10, and 20 min after placing the static chamber. Using a 30 mL syringe with a three-way valve, gas samples were collected and then stored in labeled containers with detailed sampling records. This sampling procedure was repeated at 3 h intervals until 9:00 a.m. the following day, providing a comprehensive 24 h record of soil GHG emissions.

Rubber plantations in the region are typically fertilized with organic fertilizers composed of cow dung (67% moisture content), containing 70 g kg⁻¹ of total organic carbon, 5.76 g kg⁻¹ of total nitrogen, 0.29 g kg⁻¹ of total phosphorus, and 0.27 g kg⁻¹ of total potassium. Compound fertilizers, consist of (urea, calcium superphosphate, and potassium chloride, with an N:P₂O₅:K₂O ratio of 14:7:9). According to the farm’s standards, the annual fertilizer application for each rubber tree includes 20 kg of organic fertilizer and 2 kg of compound fertilizer. This equates to approximately 609 g of urea (46% N content), 1166 g of calcium superphosphate (12% P₂O₅ content), and 300 g of potassium chloride (60% K₂O content).

In this study, organic fertilizer was applied as a single application in June, while chemical nitrogen fertilizer was split into two applications in June and September to provide consistent nutrient availability. Based on the principle of equal nitrogen substitution, each treatment’s fertilizer application rate varied according to the intended substitution level (

Table 1. Fertilizer application in different treatment groups.

Treatment Groups	Fertilizer Application in June from a Single Fertilization Pits
	Urea (g)	Calcium Superphosphate (g)	Potassium Chloride (g)	Organic Fertilizers (kg)
CK	0	0	0	0
NPK	608.70	1166.00	300.00	0
25%M	456.52	696.00	253.00	4.70
50%M	304.35	236.00	207.00	9.30
75%M	152.17	0	160.00	14.00
100%M	0	0	110.00	19.00

). For each application, efforts were made to time fertilization after rainfall. Fertilizer pits were cleared prior to application, and fertilizer was evenly spread in the pits and then covered with a layer of soil to minimize nutrient loss and optimize fertilizer efficiency.

2.3. Measurement and Calculation of GHG Fluxes

The concentrations of CO₂, CH₄, and N₂O in the gas samples were determined using a Gas Chromatograph (Agilent 7890B, Agilent Technologies, Santa Clara, CA, USA), with a carrier gas of high-purity N₂ (21 mL min⁻¹); the column was a HayesepQ (80–100 mesh), which operated at 60 °C. CO₂ entered a nickel catalyst converter (Ni) before the detection and was reduced to CH₄ by H₂ before it was detected by a Flame Ionization Detector (FID). CH₄ was detected using an FID detector and N₂O using an Electrical Conductivity Detector (ECD) detector. Soil GHG fluxes were calculated according to Equation (1) [23]:

$$F=\rho H{\displaystyle \frac{P}{P_0}}{\displaystyle \frac{T_0}{T}}{\displaystyle \frac{\Delta C}{\Delta t}}$$

(1)

where F is the GHG emission fluxes (mg m⁻² h⁻¹); ρ is the density of the gas at standard conditions (N₂O: 1.977 kg m⁻³, CO₂: 1.997 kg m⁻³, CH₄: 0.717 kg m⁻³); H is the height of the static box (0.20 m); ΔC/Δt is the rate of change of the concentration in the static box; T₀ and P₀ are the absolute air temperature (T₀ = 273.15 K) and air pressure (P₀ = 101.325 kpa), respectively, at standard conditions; T is the absolute temperature in the static box; and P is the air pressure in the sampling location. F > 0 indicates that the gas is emitted from the soil into the atmosphere, and F < 0 indicates that the atmosphere is flowing into the soil or the soil is absorbing the gas from the atmosphere.

The correction factor for soil GHG fluxes is defined as the difference between the daily average of soil GHG fluxes and the real-time observation. The daily average of soil GHG fluxes is the sum of soil GHG fluxes at each sampling moment multiplied by the time interval of sampling (3 h) and divided by 24 h. The correction factor for soil GHG fluxes is calculated as described in (2) [24]:

$$C_i={\displaystyle \frac{F_{mean}}{F_i}}$$

(2)

where C_i is the correction factor for GHG fluxes, F_i is the real-time observation of GHG fluxes, and F_mean is the daily average of GHG fluxes.

2.4. Soil Sample Collection and Analysis

For soil sampling, 0–10 cm of soil was taken in the vicinity of each soil respiration ring. Specifically, four soil samples were taken from each replicate and mixed to form a composite sample, resulting in three composite samples per treatment group. These samples were stored in self-sealing bags at room temperature and brought back to the laboratory. Then, the composite samples were placed in trays and mixed well, and a 10-mesh sieve was used to remove stones, animals, roots, and other non-soil materials. A portion of the sieved fresh soil was then stored in the refrigerator at 4 °C in a sealed container for the determination of soil ammonium nitrogen (NH₄⁺-N) and nitrate nitrogen (NO₃⁻-N). The remaining portion was air-dried, ground, and sieved using a 100-mesh sieve. The sieved soil samples were sealed and stored under dry conditions for the determination of soil pH and soil organic carbon content (SOC).

Soil samples were collected at a depth of 0–10 cm for physicochemical property analysis (pH, SOC, NO₃⁻-N and NH₄⁺-N). To determine soil pH, 10 g of air-dried soil was mixed with 25 mL of distilled water, and pH was measured using a calibrated pH meter. The SOC content was determined using the K₂Cr₂O₇-H₂SO₄ oxidation method [7]. The soil NO₃⁻-N and NH₄⁺-N contents were extracted with KCl solution, with NO₃⁻-N contents analyzed by ultraviolet (UV) spectrophotometry and NH₄⁺-N contents determined by indophenol blue colorimetry [22]. Soil temperature and soil moisture were determined using a soil tester (Sigma PH368, Merck KGaA, Darmstadt, Germany) carried at the time of sampling, which was inserted 5–10 cm into the soil; after a period of a few minutes, values were read and recorded.

2.5. Statistical Analysis

Statistical analyses were performed using SPSS software v25.0. A one-way ANOVA was constructed using Duncan’s Multiple Polar Difference method to analyze significant differences in environmental factors (pH, SOC, NO₃⁻-N and NH₄⁺-N) among the six treatments. Statistical significance was found at the p < 0.05 level.

3. Results

3.1. Soil Physical and Chemical Properties

The daily variation of soil temperature and moisture at 5 cm depth is shown in Figure 2. The diurnal patterns of soil temperature showed a single-peak curve. In the wet season, the lowest soil temperature occurred at 3:00, and the highest soil temperature occurred at 15:00. In the dry season, the lowest soil temperature occurred at 6:00, and it reached peak values at 15:00. The diurnal patterns of soil moisture were complicated; moisture increased around midnight (0:00–3:00), reached peak values at 18:00, and subsequently decreased in the wet season (Figure 2a). However, there were no obvious daily variations of soil moisture in the dry season (Figure 2b). Both soil temperature and soil moisture were higher in the wet season than in the dry season (Figure 2). Soil temperature was 14.64% higher in the wet season than in the dry season, and soil moisture was 33.80% higher than in the dry season.

In this study, soil pH, SOC, soil NO₃^—N, and soil NH₄⁺-N were measured (Figure 3). As shown, soil pH and SOC did not vary significantly between the wet and dry seasons (Figure 3A,B), with some differences between treatment groups. In the wet season, soil pH was not significantly different between CK and NPK, 25%M, and 100%M and was significantly different from the other treatment groups. The SOC was significantly different between CK and NPK and 25%M and was not significantly different from the other treatment groups. In the dry season, soil pH and SOC of the CK control group differed significantly only from those of the 100%M treatment group and not from the other treatment groups (Figure 3A,B). Soil NO₃⁻-N and soil NH₄⁺-N contents were more variable and significantly different between the wet and dry seasons. Soil NO₃⁻-N content did not differ significantly between treatment groups in the wet season, and in the dry season the soil NO₃⁻-N content in the 25%M and 100%M treatment groups differed significantly from the other treatments, whereas soil NH₄⁺-N content did not differ significantly between treatment groups in the wet and dry seasons (Figure 3C,D).

3.2. Daily Changes in Soil GHG Fluxes

Unlike the daily changes in soil temperature and moisture, soil GHG fluxes showed different daily characteristics during the wet and dry seasons (Figure 4).

In this study, the trend of soil CO₂ fluxes showed a significant single-peak curve, which was higher during the day than at night, implying that the amount of CO₂ emitted during the day was higher than at night. There were significant differences in soil CO₂ fluxes between the wet and dry seasons and also between treatment groups. During the wet season, the maximum value of soil CO₂ fluxes occurred at 15:00 and the minimum value at 3:00, with a range of 73.42–526.17 mg m⁻² h⁻¹ (Figure 4a). In the wet season, CO₂ fluxes were higher in the 75%M, 100%M, and 50%M treatment groups than in the CK treatment group, which showed 75%M > 100%M > 50%M > CK. 25%M, and NPK treatment groups were lower than the CK treatment group, which showed CK > 25%M > NPK (Figure 4a). During the dry season, the maximum value of soil CO₂ fluxes also occurred at 15:00 and the minimum value at 3:00, with a range of 102.38–446.16 mg m⁻² h⁻¹1 (Figure 4b). CO₂ fluxes were higher in all treatment groups than in the CK treatment group, showing 50%M > 75%M > 100%M > 25%M > NPK > CK (Figure 4b). Overall, the daily changes in soil CO₂ emissions had a strong daily variation characteristic in both the wet and dry seasons, and the trend was very similar to that of soil temperature.

Compared to the daily variation of CO₂ fluxes, CH₄ fluxes were somewhat different. In this study, the soil CH₄ fluxes showed negative values, indicating that CH₄ was absorbed by the soil. The daily trend also showed a significant single-peak curve, which was lower during the day than at night. There was no significant difference in soil CH₄ fluxes between the wet and dry seasons, and the differences between treatment groups were not significant. In the wet season, the minimum value of soil CH₄ fluxes occurred at 15:00 and the maximum value at 3:00, with a variation range of −0.0202 to −0.0091 mg m⁻² h⁻¹. CH₄ fluxes were higher in all treatment groups than in the CK treatment group, which showed that 25%M > NPK > 50%M > 100%M > 75%M > CK (Figure 4c). In the dry season, the minimum value of soil CH₄ fluxes also appeared at 15:00 and the maximum value appeared at 3:00, with a variation range of −0.0247 to −0.0115 mg m⁻² h⁻¹. CH₄ fluxes were lower than those in the CK treatment group in all treatment groups, as shown by CK > 50%M > 100%M > NPK > 75%M > 25%M (Figure 4d).

In this study, the trend of soil N₂O fluxes also showed a significant single-peak curve, with the daily variation being higher during the day than at night, implying that the amount of N₂O emitted during the day was higher than at night. There were also significant differences in soil N₂O fluxes between the wet and dry seasons, with the difference being that there was no significant difference in soil N₂O fluxes between treatment groups in the wet season, while there was a significant difference in the dry season. In the wet season, the maximum value of soil N₂O fluxes also occurred at 15:00 and the minimum value at 3:00, with a range of 0.0101–0.0364 mg m⁻² h⁻¹. N₂O fluxes were higher in all treatment groups than in the CK treatment group, which showed 100%M > 25%M > 50%M > NPK > 75%M > CK (Figure 4e). In the dry season, the maximum value of soil N₂O fluxes also appeared at 15:00 and the minimum value appeared at 3:00, with a variation range of 0.0042–0.0965 mg m⁻² h⁻¹. N₂O fluxes in all treatment groups were also higher than that in the CK treatment group, as shown by NPK > 50%M > 25%M > 75%M > 100%M > CK (Figure 4f). The N₂O fluxes were similar to CO₂ fluxes in terms of overall changes.

3.3. Relationships Between Soil GHG Fluxes and Soil Temperature or Moisture

Correlation analysis showed that the response of different GHG fluxes to soil temperature was different (Figure 5). The CO₂ fluxes and N₂O fluxes showed significant positive correlations with soil temperature in the rainy season but not in the dry season. Soil CH₄ fluxes, on the other hand, showed a significant negative correlation with soil temperature in both the wet and dry seasons. Overall, all three soil GHG fluxes were more strongly correlated with soil temperature in the wet season than in the dry season.

The response of different GHG fluxes to soil moisture was also different. Regression analysis revealed (Figure 6) that soil CO₂ fluxes were positively correlated with soil moisture, as evidenced by a non-significant correlation in the wet season and a significant correlation in the dry season. The relationship between soil CH₄ fluxes and soil moisture showed a significant negative correlation in both the wet and dry seasons. Soil N₂O fluxes were positively correlated with soil moisture in both the wet and dry seasons, in contrast to soil CH₄ fluxes, and the correlation was not significant in both the wet and dry seasons.

3.4. Correction Factor for GHG Fluxes

Because of the large spatial and temporal heterogeneity of soil GHG fluxes, the most representative moments of the day should be selected for observations to enhance the representativeness and accuracy of the observations when conducting studies on longer time scales, such as different seasons. Figure 7 shows the distribution of correction coefficients of CO₂, CH₄, and N₂O fluxes with time. As can be seen from Figure 7, the correction factor of CO₂ fluxes in the wet and dry seasons is closer to one at 9:00–12:00, indicating that the period of 9:00–12:00 can be used as the best time of day to observe CO₂ fluxes. Similarly, the correction coefficients of CH₄ and N₂O fluxes are also closer to one at 9:00–12:00. To further analyze their representativeness, the fluxes of each GHG at the optimal moment in the wet and dry seasons were regressed against the total daily fluxes (Figure 8). For CO₂ fluxes, the correlation coefficients with the total daily fluxes were 0.9475 (p < 0.01) in the wet season and 0.9750 (p < 0.01) in the dry season. The CH₄ fluxes exhibited significant correlations with its total daily fluxes, and the correlation coefficients were 0.7023 (p < 0.01) and 0.6220 (p < 0.01) in the wet season and dry season, respectively. The N₂O fluxes also showed significant correlations with its total daily fluxes, and the correlation coefficients were 0.9675 (p < 0.01) and 0.9656 (p < 0.01) in the wet season and dry season, respectively. All of these reached the significant level, indicating that the fluxes at these moments have strong correlation with the total daily fluxes and can be used as the optimal observation times for the study of GHGs in rubber plantations on a long time scale.

4. Discussion

4.1. Daily Variation of Soil GHG Fluxes

In this study, the mean values of soil GHG fluxes for each treatment group were calculated. The results showed that the average daily emission of soil CO₂ was 194.1 mg m⁻² h⁻¹, the average daily uptake of soil CH₄ was −0.0155 mg m⁻² h⁻¹, and the average daily emission of soil N₂O was 0.011 mg m⁻² h⁻¹ in the CK treatment group (

Table 2. Average daily fluxes in different treatment groups.

Treatment Groups	Average Daily Fluxes (mg m−2 h−1)
	CO2	CH4	N2O
CK	194.1 ± 36.8 b	−0.0155 ± 0.0033 a	0.011 ± 0.002 c
NPK	137.0 ± 29.4 c	−0.0148 ± 0.0030 a	0.028 ± 0.004 b
25%M	216.1 ± 36.0 b	−0.0152 ± 0.0023 a	0.032 ± 0.005 ab
50%M	317.4 ± 53.2 ab	−0.0147 ± 0.0039 a	0.040 ± 0.006 a
75%M	325.5 ± 62.0 ab	−0.0158 ± 0.0035 a	0.026 ± 0.004 b
100%M	372.0 ± 70.6 a	−0.0160 ± 0.0043 a	0.025 ± 0.005 b

Note: The different lowercase letters represent significant differences among the different treatment groups.

). The soil CO₂ fluxes in this study were significantly lower than those in Xishuangbanna (soil CO₂ fluxes of tropical rainforest and rubber plantation were 359.95 mg m⁻² h⁻¹ and 351.99 mg m⁻² h⁻¹). Soil CH₄ uptake of rubber plantations in this study was also lower than that in Xishuangbanna (soil CH₄ uptake was −0.11 mg m⁻² h⁻¹ and −0.02 mg m⁻² h⁻¹ for tropical rainforest and rubber plantation, respectively) [25]. However, it was higher than that of Southeast Asia (soil CH₄ uptake of rubber plantations was −0.0104 mg m⁻² h⁻¹) [26]. Comparing with the results of other regions, we found that the emission of N₂O from rubber plantation soil in this study was lower than those of Danzhou, Hainan [27], Xishuangbanna, and Yunnan [28], which may be due to differences in hydrothermal conditions.

Daily variation of soil GHG fluxes is a common phenomenon [29]. In this study, the daily variation of soil GHG fluxes in rubber forests under different levels of organic fertilizer substitution showed a clear single-peak curve (Figure 4). CO₂ fluxes and N₂O fluxes showed a trend of high in the daytime and low in the nighttime. CH₄ fluxes showed CH₄ uptake was high in the daytime and low in the nighttime, which is consistent with the results of Li et al. [30]. The CO₂ fluxes and N₂O fluxes showed that the dry season was saw greater fluxes than in the wet season, and there was no significant change in CH₄ fluxes.

To accurately measure GHG fluxes with minimal labor, it is important to observe daily changes in GHG fluxes in different regions. It has been shown that 8:00–10:00 a.m. is the best observation time for CO₂ fluxes and 8:00–12:00 a.m. is the best observation time for N₂O fluxes in wheat fields [31]. Tai et al. showed that 8:00–10:00 a.m. was the best observation time for CO₂ fluxes and N₂O fluxes after fertilizer application in spring maize fields under shallow buried drip irrigation [32]. In this study, based on the results of the distribution of the correction coefficients over time, the observed soil GHG fluxes from 9:00–12:00 were closest to the daily average fluxes, which indicates that the fluxes of each soil GHG at this time of the day are most representative of the average fluxes over the day. It was concluded that 9:00–12:00 was the best time of day for the study of soil GHG fluxes from rubber forests in this region under different levels of organic fertilizer substitution (Figure 7 and Figure 8).

4.2. Effects of Organic Fertilizer Substitution on GHG Fluxes in Rubber Plantation Soils

Organic fertilizer application could led to alternations in soil structure and nutrient availability, subsequently influencing soil GHG emissions [33]. Compared with the CK treatment, fertilizer application increased soil CO₂ fluxes and N₂O fluxes during both dry and wet seasons (Figure 4), which was consistent with the result of Abbasi et al. and Qiu et al. [34,35]. In a four-year experiment, Abbasi et al. [34] found that organic-inorganic allotment treatments increased soil CO₂ emission fluxes in a maize-soybean crop rotation system in Ontario, Canada. Qiu et al. [35] found that fertilizer application significantly increased soil CO₂ and N₂O emission fluxes in their experiments on fir plantations in Nanping, Fujian Province. However, the application of fertilizer had minimal influence on soil CH₄ fluxes in this study.

In this study, the organic fertilizer replacement treatment significantly increased soil CO₂ fluxes compared to the NPK treatment (Figure 4a,b). This effect is primarily because the organic and inorganic fertilizers could provide organic materials and accelerate soil microbial activity, thus enhancing CO₂ fluxes [36]. Additionally, soil CO₂ fluxes in organic substitution treatments (with the 25%M, 50%M, 75%M, and 100%M) were significantly higher than those in both CK and NPK treatments. In the wet season, CO₂ fluxes gradually increased as organic fertilizer substitution increased (Figure 4a), whereas in the dry season, emissions peaked with substitution ratios ≤ 50%, then decreased as substitution exceeded 50% (Figure 4b). We speculate that this may be due to the low soil moisture during the dry season, which reduced microbial activity [37]. When the percentage of organic fertilizer substitution exceeded 50%, the soil accumulated a large amount of organic matter that failed to decompose, thus reducing soil CO₂ fluxes.

Soil CH₄ fluxes indicated CH₄ uptake in this study, which is a typical pattern in most dry-land agricultural fields [38]. Soil CH₄ fluxes are primarily affected by both methanogenic and methane-oxidizing microbial communities [39]. Additionally, the results showed that the organic fertilizer application slightly reduced CH₄ uptake, but there were not significant differences between treatment groups (Figure 4c,d). This stability is likely due to the equilibrium maintained between CH₄ production by methanogenic bacteria and CH₄ consumption by methane-oxidizing bacteria. These two methanogens are mainly affected by factors such as soil temperature, moisture, and pH [40], which were not significantly affected by the different fertilization treatments in this study, so there were no significant differences in soil CH₄ fluxes between treatment groups.

Soil N₂O emissions arise mainly from nitrification and denitrification processes, which govern N₂O fluxes in the soil ecosystem [41]. In this study, N₂O fluxes were higher in the fertilizer treatments than in the CK treatment, which may be attributed to enhanced nitrification and denitrification driven by available N contents, which in turn promotes N₂O emission [42]. Notably, N₂O emissions decreased when organic fertilizer substitution exceeded 50%M (Figure 4e,f). This is because available nitrogen (NO₃⁻-N and NH₄⁺-N contents) decreased as the organic fertilizer substitution ratios increased, which limited microbial processing efficiency and elevated N₂O emissions [43]. In addition, organic fertilizer substitution improved soil structure, enhanced nitrogen use efficiency, and stabilized microbial communities, contributing to the observed decline in N₂O fluxes [44].

The 25%M treatment group was superior to other organic fertilizer replacement treatment groups in reducing soil GHG fluxes compared to the NPK treatment group. Therefore, the 25%M treatment group is recommended as the optimum fertilizer treatment group for this study.

4.3. Effects of Soil Temperature and Soil Moisture on Soil GHG Fluxes

In this study, we examined the impact of different organic fertilizer substitution treatments on daily variation of GHG fluxes across wet and dry seasons. Soil temperature and soil moisture are key drivers of soil GHG fluxes, as these factors impact soil microbial activity [45], yet their effects vary between seasons and under organic substitution conditions.

There were positive correlations between soil CO₂ fluxes and soil temperature or moisture (Figure 5 and Figure 6a,b), suggesting that increased microbial respiration and decomposition rates in warmer, moist soils elevated CO₂ emissions [46]. Previous studies’ results also showed that, under sufficient moisture and nutrient substrate availability, the temperature primarily influences soil CO₂ fluxes [47].

Correlation analysis results showed that soil CH₄ fluxes were significantly and negatively correlated with soil temperature (Figure 5b,c), meaning that higher temperature enhanced CH₄ uptake [48]. This could be because optimal conditions for most methanotrophic microorganisms is around 35–37 °C [49]. When soil temperatures fall below this level, higher temperatures lead to greater consumption [50]. Additionally, soil moisture is also negatively correlated with soil CH₄ fluxes during the wet season (Figure 6b,c). Soil CH₄ emissions increased under flooded conditions and uptake increased under dry conditions [51].

Similarly, soil N₂O fluxes displayed a positive correlation with both temperature and moisture during the wet season (Figure 5 and Figure 6e,f). This is consistent with the results of Gao et al. and Wu et al. [52,53]. This might be caused by the fact that the nitrification-denitrification reaction was more active in the wet season [54] because soil temperature and soil moisture were higher in the wet season than in the dry season. The positive correlation between soil N₂O fluxes and soil moisture shows a positive correlation, which is consistent with the results of most studies [55,56]. In the present study, there was a significant difference in soil N₂O fluxes between the wet and dry seasons. We speculate that the main reason for this is that, in the tropics, soil temperatures and humidity are higher in both seasons, and substrate differences may be the main influencing factor. Our results showed that effective N was significantly higher in the dry season than in the wet season (Figure 3).

Considering the lack of long-term positional observations in this study, future research needs to focus on the effects of daily variability patterns on long time scales and other types of fertilization patterns on GHG fluxes from rubber forest soils. At the same time, as there are many factors that influence the GHG fluxes from rubber forest soils, future research needs to focus on considering the effects of other factors. In addition, fertilizer applications would affect soil nutrient dynamics and yield in rubber plantations. Therefore, future research also needs to focus on this aspect.

5. Conclusions

Accuracy estimation of GHG fluxes from rubber plantation soils necessitates a thorough understanding of their daily variation characteristics and the optimal sampling times. This study demonstrated that the optimal sampling window for reliable GHG flux measurement was identified as between 9:00 and 12:00 in the morning. Organic fertilizer substitution promoted soil CO₂ and N₂O emissions while having no significant effect on soil CH₄ emissions, with CO₂ and N₂O emissions rising with the increased substitution ratios and declining when substitution exceeded thresholds of 50% for CO₂ and 75% for N₂O. Soil CO₂ and N₂O fluxes were positively correlated with soil temperature and soil moisture, and CH₄ fluxes were negatively correlated with soil temperature and soil moisture in both wet and dry seasons. These results suggest that the daily variation characteristics of soil GHG fluxes should be taken into account when designing experiments or observation programs to improve the accuracy of annual GHG flux estimates in rubber plantation ecosystems.