4.1. Varietal differences in total CH4 emissions
We found significant variability in total CH4 fluxes among the 22 rice varieties, consistent with previous reports. This variability did not correlate with the difference in plant phenology, especially at the later stage (Stage 3, Fig. 2). However, at Stage 2, late-maturing varieties tended to exhibit higher total CH4 fluxes compared to early-maturing ones (Fig. 2). This may be attributed to the higher plant-mediated flux in late-maturing varieties contributing significantly to the total flux (Figs. 3 and 4). Previous studies have reported inconsistent results regarding the CH4 emissions of early- and late-maturing varieties (Gogoi et al., 2008; Baruah et al., 2010; Shin and Yun, 2010; Gutierrez et al., 2013). These discrepancies could be due to the varying contributions of plant-mediated / bubbling fluxes depending on the rice genotypes investigated. Our findings highlight the importance of separately quantifying plant-mediated and bubbling fluxes to fully understand the underlying mechanisms of the varietal differences in CH4 emissions.
4.2. Factors affecting plant-mediated and bubbling CH4 emissions
Our results indicate plant traits, rather than the CH4 pool size, may predominantly control the plant-mediated flux, considering the weak correlation between the CH4 flux and dissolved [CH4] at Stage 2 (Fig. 6b). However, at earlier growth stages, where plant trait variabilities may be less pronounced, the CH4 pool could have a relatively greater impact on plant-mediated flux, as evidenced by the significant correlation between dissolved [CH4] and plant-mediated flux at Stage 1 (Fig. 6b).
A potential candidate for the plant traits controlling plant-mediated flux at later stages is the abundance of aerenchyma in roots, as gas diffusivity is higher in the gaseous phase compared to the liquid phase. Root porosity has been observed to vary significantly across rice varieties, although the varieties in these observations did not overlap with those in our study (Mei et al., 2020). Varietal differences in aerenchyma abundance may be influenced by both phenology and genetics. Examining individual stages, negative and stronger correlations were found between the maturity index and plant-mediated flux at later growth stages (Fig. 3, Results 3.1), suggesting that plant traits influenced by maturation speeds may partially control the rate of plant-mediated emissions, especially at later stages. Early maturing varieties may experience more advanced senescence than late-maturing ones, resulting in smaller aerenchyma tissue remaining in the roots and decreased gas transfer via the plants.
However, the varietal differences in plant phenology were found to have only a marginal effect on plant-mediated flux across the three stages of measurement (Fig. 3). Even after accounting for the effect of phenology (and/or its confounding influences), several varieties with high plant-mediated emissions were observed (Fig. 4; T-test). Furthermore, varieties that exhibited low plant-mediated fluxes in the early growth stage also showed consistently low plant-mediated fluxes in subsequent stages of measurement (Fig. 7). These results indicate that plant traits determined by genetics, rather than maturity level, may substantially influence the plant-mediated flux. Further studies are required to investigate the relationships between root traits and plant-mediated CH4 emissions in conjunction with rice genotypes having different maturity speeds.
We did not observe any significant correlations between plant-mediated flux and straw weight (Fig. S3), which we considered as an alternative quantitative index of the exits of CH4 from the rice plants, suggesting that the rate at which CH4 exits from plants did not control the plant-mediated emissions. The findings align with a prior study indicating that variations in total CH4 flux among eight varieties were minimally correlated with the number of tillers and shoot weight at a maturity level of 0.57–0.66, which corresponds to Stage 1 in this study when plant-mediated flux was predominant (Watanabe et al., 1995).
Bubbling CH4 emissions, on the other hand, may be largely governed by the size of the soil CH4 pool, which increases markedly as the plant matures. This is supported by the strong and consistent relationship between the dissolved [CH4] and bubbling flux and the observed increase in dissolved [CH4] as plant matures (Figs. 6a, b). As rice plants mature, the accumulation of root exudates and/or dead roots likely increases, providing more substrates for methanogens and, consequently, enhancing CH4 production (Tokida et al., 2011).
Nevertheless, significant varietal differences remained at Stage 3 in the bubbling emission even when the difference in plant maturity level was considered (Fig. 4). Significant bubbling emission from Koshihikari with elevated soil CH4 pool is particularly noticeable (Figs. 6b). This elevated [CH4] in Koshihikari may be attributed to factors other than maturity, as dissolved [CH4] was higher than what would be expected from the maturity versus dissolved [CH4] relationship observed at Stage 2 (Fig. 6a). Potential reasons for this discrepancy include higher CH4 production in Koshihikari's rhizosphere, possibly due to factors such as root decomposability or the quality/quantity of root exudates. Another possibility is that soil CH4 could not efficiently transfer to the atmosphere through plant-mediated flux, possibly remaining more abundantly in soil due to the plant traits mentioned earlier. Indeed, the plant-mediated flux of Koshihikari was relatively low despite its highest dissolved [CH4] (Fig. 6b).
4.3. Toward CH4 mitigation in rice paddies
The total CH4 emission in rice paddies is the sum of plant-mediated and bubbling fluxes. To effectively mitigate total CH4 emissions, it is essential to adopt strategies tailored to each emission pathway, as they are influenced by different factors (discussed in section 4.2). Bubbling emission, in particular, plays a crucial role alongside plant-mediated flux in rice varieties like Koshihikari that exhibit a high bubbling contribution, especially during later growth stages (Fig. 4). The primary strategy for mitigating bubbling emissions involves reducing the CH4 pool in paddy water, given the sensitivity of bubbling flux to dissolved [CH4] (Fig. 6b). Possible methods for reducing the CH4 pool may include water management, enhanced rhizosphere oxidation, and altering root decomposability.
Plant-mediated CH4 emissions play a significant role throughout the entire growth season. It's particularly noteworthy that Koshihikari's high bubbling contribution of 0.57 was observed during the daytime at the heading stage, with contributions being lower at other growth stages (Fig. 4) and likely even lower at night. Kajiura and Tokida (2021) pointed out that the bubbling contribution to seasonal total CH4 emissions could approximate 0.26, although this estimation is based on conservative assumptions and might increase with practices such as pre-puddling straw incorporation. Given this significant role, minimizing plant-mediated emissions remains crucial for CH4 mitigation. Our findings indicate that plant traits, rather than the CH4 pool, influence plant-mediated flux considerably, especially during the later growth stages where flux variability is more pronounced (Figs. 3, 6). The consistent differences in plant-mediated flux observed across various varieties throughout all three growth stages (Fig. 7) suggest that altering plant traits, such as root characteristics, could effectively reduce CH4 emissions over the entire growth period.
Maintaining or enhancing grain yield is equally vital when breeding rice varieties for reduced CH4 emissions. The weak correlation between panicle weight and plant-mediated flux, consistent with previous reports (Hayashi et al., 2015), and the existence of varieties with higher yields and lower CH4 emissions than Koshihikari indicate the possibility of developing varieties that achieve both high yield and low CH4 emissions (section 3.2). Thus, there is potential for breeding 'win-win' rice varieties that do not compromise yield for the sake of emission reduction (Gutierrez et al., 2013; Jiang et al., 2019).
4.4. Limitations
The results of this study are based on experiments conducted in a single field obtained in a single season. To validate the general trends regarding which rice varieties emit more or less CH4, it is essential to replicate these observations under varying field conditions and environmental settings. Bubbling emission's high sensitivity to temperature (Kajiura and Tokida, 2021; Kajiura and Tokida, 2024) indicates that climate and weather condition significantly impacts CH4 emissions, particularly at later growth stages. Additionally, in fields with greater retention of rice residues in the soil, the quantity and profile of dissolved [CH4] may vary, potentially affecting the influence of root traits on plant-mediated flux differently than observed in our study. Nonetheless, the following findings are likely robust: (1) pathway-specific CH4 emissions and their contributions to the total flux differ across rice varieties, (2) Koshihikari exhibits an exceptionally high bubbling contribution, and (3) different mechanisms control pathway-specific CH4 emissions.