4.1 Nitrous Oxide Emission from the Natural and Rice Paddy Wetlands
Within the natural wetland, N2O fluxes did vary significantly across vegetation communities. Similar results have been reported from other studies elsewhere, such as North Dakota, USA (Phillips and Beeri, 2008), Denmark (Audet et al. 2014) and Mexico (Hernández and Junca-Gómez 2020). Some studies have shown that vegetation community affects N2O fluxes from wetlands. Liu et al. (2014) showed that emissions of N2O from two vegetated coastal wetland zones, one dominated by Spatina spp. and the other dominated by Phragmites spp. differed significantly, with the Phragmites spp. zone showing higher N2O flux, even though both zones were at the same altitude. This observation was explained by differences in biomass productivity, which affects the resultant organic matter and nitrogen input into soil. This is further supported by Piñeiro-Guerra et al. (2019), whose study across several sites in Argentina showed that N2O emission increased with primary productivity. Wang et al. (2008) investigated the impact of plant species on N2O emission, using constructed wetlands. Results indicated that the wetland planted with Zizania latifolia had a higher emission of N2O compared to those planted with Phragmites australis and Typha latifolia. They attributed the results to differences in root structure of the plant species, where the root structure of Zizania latifolia was favored by ammonia oxidizing bacteria for N2O formation. Abalos et al. (2018) investigated plant functional traits that may affect N2O emission and noted that specific leaf area and root length density were key traits that did not only regulate N2O emission but also biomass productivity.
Wetland conversion/land use change, from natural to rice paddy wetland has been shown to enhance N2O emissions from wetlands (Liengaard et al. 2013; Owino et al. 2020). This is attached two reasons: i) water table drawdown through drainage favors oxygen availability, thus increasing the mineralization of soil organic nitrogen and nitrification processes (Ajwang’Ondiek et al. 2021; Liengaard et al. 2013), and ii) nitrogen fertilization of rice paddy fields to enhance rice yields increases nitrogen availability in soil, a substrate for nitrification/denitrification processes (Owino et al. 2020). Liengaard et al. (2013) reported that partial soil wetting by light rain in a tropical South American freshwater wetland resulted into high N2O emission compared to long-term soil waterlogging following heavy rain. In China, Yang et al. (2013) observed that N2O emission from a natural marsh wetland increased by 120% as the water level reduced from + 14 cm to -11 cm. Liu et al. (2020) have estimated that agricultural drainage of European (EU-28) peatlands has caused annual N2O emissions from these managed peatlands of 145 Gg N yr− 1. They further indicated that rewetting of all drained European peatlands could cut the cumulative N2O emissions by 70%. Similarly, a recent synthesis of several studies on the effect of wetting and drying cycles on N2O emission from freshwater sediments has reported N2O pulses following sediment drying and rewetting events, with exposed sediments being active spots for N2O emissions during dry phases (Pinto et al. 2021).
Owino et al. (2020) reported that conversation of Papyrus wetlands into nitrogen fertilized rice paddies in Kenya significantly increased N2O emission (4.37 ± 3.18 µg m− 2 h− 1 from the fertilized fields against − 3.59 ± 2.56 18 µg m− 2 h− 1 from the unfertilized fields). Zhang et al. (2014) also reported a positive correlation between the amount of nitrogen fertilizer application and N2O emissions from rice paddies, irrespective of the rice growth stage. Therefore, the lack of significant difference in N2O fluxes from the natural and rice paddy wetlands in this study could be explained as follows: i) both the natural and rice paddy wetlands were continuously flooded throughout the sampling period, so water table could not significantly affect N2O fluxes from both wetlands, as was indeed shown by lack of a significant correlation between water level and N2O flux (Table 3), and ii) fertilization of the rice paddy fields studied is not done, as rice cultivation is only reliant on the natural fertility of the soil. Undoubtedly, soil nitrogen contents of natural and rice paddy wetlands were not different, as was seen in Table 1.
Seasonal variations did not affect N2O fluxes from both wetlands. Seasonal variation in N2O emissions, especially associated with temperature differences (summer vs winter seasons) have been reported in temperate wetlands (Czóbel et al. 2020; Jørgensen and Elberling 2012). Warmer soil temperatures during summer seasons have been shown to enhance microbial activities, resulting into increased mineralization of organic nitrogen in soils. However, in tropical regions, it is shown that temperature is unlikely to exert a minimal control on GHGs fluxes from wetlands because temperatures are relatively stable, irrespective of season (Sjögersten et al. 2014). However, in many tropical wetlands, seasonal changes are usually associated with changes in water table e.g. drying and flooding cycles during the dry and wet seasons, respectively (Bernal and Mitsch 2013). Therefore, in such cases (drying and wetting cycles), seasonal N2O flux from those wetlands is likely to be driven more by changes in the water table depth rather than soil temperature. However, it has also been indicated that seasonal N2O fluxes based on wetting and drying cycles are more pronounced in wetland systems influenced by anthropogenic input on nitrogen (Hernandez and Mitsch et al. 2006). In the present study, seasonal changes (dry vs wet) were not associated with drying and wetting cycles as water level was above the soil surface in both seasons, and neither were the wetlands affected by anthropogenic nitrogen input. Thus, this could explain why N2O emission did not vary between the dry and wet seasons.
This study showed great variations in individual N2O fluxes even within the same wetland/vegetation community/season, with values ranging from negative (indicating N2O consumption) to positive (indicating N2O emission) (Figs. 3 and 4). This has also been reported by other studies (Ajwang’Ondiek et al. 2021; Audet et al. 2014; Owino et al. 2020), where it is attributed to wetland soils acting as both N2O sinks and sources. Schlesinger (2013) and Wu et al (2013) have explained that soils act as sinks of N2O when N2O is consumed during either nitrification or denitrification, largely under conditions of limited nitrate (NO3−).
In this study, we observed no significant correlation between soil physico-chemical characteristic and N2O from the wetlands. Audet et al. (2014) explored the factors controlling N2O uptake and emission from several natural wetlands. They also established that no significant correlation between N2O emission and soil physico-chemical characteristics. Instead, they observed that ammonium concentration in the groundwater was the only parameter that exerted a significant control on N2O fluxes from the wetlands. Richards and Craft (2015) have explained that emission of GHGs, including N2O from wetlands is an interplay between several factors, with the key controlling factor(s) varying depending on the existing conditions.
In Uganda, no study has evaluated the totality of N2O emissions from the country’s natural and rice paddy wetlands. Therefore, to provide a basis for future studies, we used our study findings to roughly estimate total annual N2O emissions from Uganda’s natural and rice paddy wetlands. To achieve this, we made a simple assumption that other natural and rice paddy wetlands in the country present more or less similar conditions to our study wetlands. Natural wetlands in the Uganda cover about 26,165 km2 compared to 150 km2 occupied by rice paddy wetlands (Were et al. 2021b). We obtained mean annual N2O emission from the natural and rice paddy as 4.4 ± 13.1 mg m− 2 yr− 1 and 6.1 ± 17.8 mg m− 2 yr− 1 (Table 2). Therefore, total N2O emissions from Uganda’s natural and rice paddy wetlands are estimated at 115.1 ± 342.8 T yr− 1 (CO2e = 30,501.5 ± 90,842 T yr− 1) and 0.9 ± 2.7 T yr− 1 (CO2e = 242.5 ± 707.6 T yr− 1), respectively.
4.2 Implication of Permanently Flooded Rice Cultivation under no Fertilization on Climate Change Mitigation
To mitigate climate change and its impacts, the IPCC (2014) has emphasized increasing carbon sequestration while at the same time minimizing emission of GHGs into the atmosphere. Some of the suggested measures to increase carbon sequestration include protection and conservation of natural ecosystems such as forests, grasslands, and wetlands. Wetlands are unique ecosystems, where water is the terminal parameter that influence the development of soil and plant characteristics different from other ecosystems. These unique characteristics have a great influence on carbon and nitrogen cycling in these ecosystems (Mitsch et al. 2013). Several studies have acknowledged that conversion of natural wetlands in farmed wetlands compromises climate change mitigation by enhancing carbon and nitrogen emission (Ajwang’Ondiek et al. 2021; Owino et al. 2020; Were et al. 2020b). As earlier explained, water drawdown in rice paddy has been attributed to enhanced carbon emissions from these ecosystems (Were et al. 2019; Were et al. 2021b), while water table drawdown alongside nitrogen fertilizer application are responsible for high nitrogen fluxes from these ecosystems (Owino et al. 2020). With the increasing human population globally, increased food demand implies natural tropical freshwater wetlands will remain under high pressure of conversion into rice paddies. Indeed, Davidson and Finlayson (2018) have reported an average increase in rice paddy acreage of 0.62% per year. As result, climate change mitigation efforts need to explore ways of optimizing carbon sequestration while minimizing emission of GHGs such as N2O from rice paddy wetlands. Basing on the results of this study, therefore, rice cultivation under continuous flooding and no nitrogen fertilizer application enhances climate change mitigation as it minimizes N2O fluxes into the atmosphere. However, continuously flooded conditions in wetlands are also known to enhance CH4 emission to the atmosphere (Were et al. 2019). Nevertheless, it has already been shown that undrained wetlands represent net carbon sinks of about 830 T C yr− 1 (Mitsch et al. 2013), indicating that the emitted CH4 is compensated by the CO2 sequestered.