Discussion
The results from our field experiments show that the alpine steppe was a net source of N2O. N addition significantly increased N2O emissions (Figure 2). Most terrestrial ecosystems, especially grassland ecosystems, are widely limited by N (Geng et al., 2019; Lu et al., 2011). N enrichment increases N available in soil, even reaching N saturation, and available N directly affects N2O emissions (Peng et al., 2018). In our experiment, N addition significantly increased inorganic N in soil (Table 2). N2O emissions occurred mainly due to the supply of substrate NO3-N, independent of NH4+-N (Figure 4), which indicates that denitrification may be the predominant pathway of N2O emissions in this alpine steppe. A possible explanation for this finding is that the N supply may lead to plants and microorganisms competing for NH4+-N (as a substrate for nitrification) in N-limited grassland ecosystems. Liu et al. (2013) discovered that N input promotes plant N uptake, especially NH4+-N. In this case, nitrification might have been inhibited due to lack of substrates. We also found that changes in abiotic factors such as soil temperature and pH regulated N2O emissions. Generally, soil N cycling largely depends on soil temperature in alpine ecosystems. In particular, warming was found to drive N2O production and emissions (Griffis et al., 2017). However, rising temperatures negatively affected N2O emissions in our study (Figure 4). It is possible that higher temperatures aggravate evapotranspiration and decrease soil water availability, thereby limiting various microbial N cycling processes (Shi et al., 2012). Previous studies have also shown that soil acidification caused by N saturation limits microbial growth, thus restraining N2O emissions (Oertel et al., 2016; Treseder, 2008). In contrast, we found that lower pH contributed to N2O emissions (Figure 4). A possible explanation for this discrepancy is that even though N addition significantly decreased soil pH, the soil was still alkaline (Table 2) and therefore microbial activity was not negatively affected. It is worth noting that plant biomass is also a key driver of N2O emissions. Soil labile C via root secretion may accelerate N2O emissions because denitrification is commonly driven by high available C as a source of energy (Li et al., 2020). This phenomenon is consistent with our conclusion that the increase of belowground biomass boosted N2O emissions (Figure 4).
Changed precipitation regimes also play an important role in modulating soil N cycling (Chen et al., 2013; Cregger et al., 2014; Lin et al., 2016). Li et al. (2020) demonstrated that increased precipitation exacerbated N2O emissions in grassland ecosystems while reductions in precipitation mitigated N2O emissions. In this study, however, we observed that altered precipitation patterns did not affect N2O emissions (Figure 2). On the one hand, water addition may diminish soil N pools (soil inorganic N) by promoting plant N uptake and soil leaching, neither of which are conducive to nitrification and denitrification (Austin et al., 2004; Kruger et al., 2021; Lin et al., 2016). On the other hand, water reduction (i.e., prolonged drought treatment) had little effect on N2O emissions, possibly because the alpine steppe itself belongs to an arid grassland ecosystem and is insensitive to drought treatment (Dijkstra et al., 2013). The interaction between altered precipitation regimes and N addition did not significantly affect N2O emissions in our experiment (Figure 2). There are several mechanisms that could contribute to this finding. Ordinarily, N and water co-limitation is a typical feature of arid grassland ecosystems (Austin et al., 2004; Lü et al., 2009). The responses of grassland ecosystems to N deposition are strongly regulated by precipitation patterns (Harpole et al., 2007). Increased precipitation, particularly under the background of N addition, could increase plant access to soil inorganic N resources (Li et al., 2019), so the effect of N addition on N2O emissions may be alleviated by water addition. In addition, decreased precipitation may suppress microbial activity, leading to inefficient N assimilation, despite the presence of large amounts of N substrates in the soil (Homyak et al., 2017; Li et al., 2020). Overall, precipitation changes attenuated N2O flux responses to N addition, thus mitigating N2O emissions on the QTP.
The community composition and diversity of N cycling microbes are directly involved in N2O production and emissions. Microbial functional genes associated with N cycling encode some key oxidoreductases and are therefore used as genetic markers for nitrifying and denitrifying microorganisms (Mushinski et al., 2021). The functional genes of AOA and AOB usually regulate the rate-limiting step (ammonia oxidation: NH3 → NH2OH) in nitrification (Hu et al., 2015; Lu et al., 2015). Some studies have indicated that N2O emissions were promoted by increased abundances of both AOA and AOB (Brin et al., 2019; Linton et al., 2020). However, we found that N addition only significantly increased the abundance of AOB (Figure 3), and the functional genes of AOB rather than those of AOA dominated the N2O emissions from nitrification (Table 3). Di et al. (2009) also showed that N2O emissions are driven by AOB and not AOA in N-enriched grassland ecosystems. Previous investigations demonstrated that AOA and AOB occupy different niches. AOA and AOB play a dominant role in acidic and alkaline soils, respectively, and pH is the chief factor for niche separation (Hu et al., 2015; Tzanakakis et al., 2019). The alkaline conditions in this study may be more conducive to the activity of AOB, which further supports our conclusion that AOB controlled the N2O emissions in nitrification. The key step of denitrification (NO2 → NO) is generally mediated bynirS - or nirK -encoding nitrite reductase (Butterbach-Bahl et al., 2013). In this study, N2O emissions were not related to the abundance of nirS and nirK (Table 3). This finding can be explained by other environmental factors such as soil temperature, pH, labile carbon, and oxygen concentration dominating the underlying ecological process (Li et al., 2020). The nitrous oxide reductase encoded by nosZ promotes N2O reduction (N2O → N2), thereby reducing N2O emissions (Butterbach-Bahl et al., 2013; Hu et al., 2015). Decreased nosZ abundance is unfavorable to the reduction of N2O, thus aggravating N2O emissions (Bowen et al., 2020). We found that N addition decreased nosZabundance to some extent, and N2O flux was negatively correlated with nosZ . Thus, the lower nosZ abundance may be responsible for the increased N2O emission in denitrification. Although changes in nirS and nirK had no effect on N2O emission, the high ratios of (nirS+nirK )/nosZ induced N2O emissions (Table 3). Given that the high ratios of (nirS+nirK )/nosZrepresented a strong N2O emissions capacity (Hu et al., 2015), the nirS - or nirK -containing denitrifiers cannot be ignored in future work on N cycling.