Introduction
Ecosystems worldwide are receiving increasing amounts of reactive nitrogen (N) through anthropogenic activities (Galloway et al. 2008; Canfield et al. 2010). The most considerable change in the ecosystem’s N cycle caused by N enrichment is rapid nitrification, defined as the ammonium conversion via nitrite to nitrate by soil microorganisms. Nitrification causes much of the reactive N to be lost in the environment by nitrate leaching and gaseous nitrous oxide production, leading to soil acidification (Liu et al. 2010; Coskun et al. 2017). In the last decades, N deposition levels have begun to decline in some parts of Europe (Sutton et al. 2011; Stevens, 2016). Since 2008, China’s cleaner technologies and policy frameworks have reduced atmospheric pollution and N fertilizer input (Liu et al. 2016). However, it is difficult to assess whether the high nitrification environment caused by long-term N enrichment can be reversed to a low nitrification environment after reduction or cessation of nitrogen enrichment due to the few studies examining the duration of the N-induced impacts on nitrification after N input cessation (Clark et al. 2009).
Within the N cycle, ammonia oxidizers are functionally essential groups of microorganisms that perform the first and rate-limiting nitrification steps. Ammonia-oxidizing archaea (AOA) and ammonia-oxidizing bacteria (AOB) are considered classical ammonia oxidizers. Many studies in various ecosystems have examined the increased exposure N inputs effects on AOA and AOB (Xiao et al. 2020). It is well established that mineral N-fertilizer application stimulates nitrifier activity and the subsequent development of high-nitrifying soil environments (Di et al. 2009; Lu et al. 2012; Ouyang et al. 2016). For example, a recent meta-analysis reported that N addition increased AOA and AOB abundances by 27% and 326%, respectively (Carey et al. 2016). Similarly, a comprehensive study from agricultural soils showed that N fertilization significantly increased archaeal amoA (31%), bacterial amoA (313%), nirK (53%), nirS (40%), and nosZ (75%) genes (Ouyang et al. 2018). However, the discovery of comammox (complete ammonia oxidation) Nitrospira , which can independently oxidize ammonia to nitrate challenged this long-held perception (Daims et al. 2015; van Kessel et al. 2015; Santoro, 2016). Comammox bacteria are differentiated into two clades (A and B) based on phylogeny of the ammonia monooxygenase α-subunit gene (amoA) (Daims et al., 2015; Pjevac et al. 2017). Recently, some studies have indicated that N input significantly increases comammox abundances, especially clade A (Li et al. 2019; Wang et al. 2019). Compared to a good understanding of the response of ammonia oxidizers to nitrogen addition, the reversibility of N-induced shifts in these groups are largely unknown.
Soil ammonia availability has been proposed to be a primary driver controlling the AOA and AOB abundances (Verhamme et al. 2011; Daebeler et al. 2015). High N accumulation occurs in regions with high mineral fertilizer application in the ecosystem. The amount of ammonia is sufficient for the development of microbial communities involved in nitrification, leading to nitrification becoming a highly significant process during the N enrichment period. But soil ammonium and nitrate concentrations were not significantly different from the control plots in the Minnesota prairie after N cessation (Clark et al. 2009). Similar results were observed at Tadham Moor (Stevens et al. 2012) and Wardlow Hay Cop (O’Sullivan et al. 2011) where nutrient additions were discontinued. These suggest soil nitrate and ammonium concentrations typically showed recovery signs after the cessation of N additions. In contrast, soil mineralization and nitrification show weak recovery because the cessation of nutrient addition did not reduce nitrification rates (Corre & Lamersdorf, 2004). Similarly, elevated N2O emissions after seven years of N addition ceassation were observed at a solling site (Borken et al. 2002). Higher net nitrification rates combined with negligible amounts of dissolved inorganic N (DIN) suggest that soil N retained in another form could influence internal cycling many years after N inputs cessation. The lack of recovery in soil processes is potentially related to changes in the soil microbial community. Consequently, long-term impacts on ammonia oxidizer communities are plausible.
Ammonia oxidizers played a key role in N cycling changes. The long-term retention of added N and its impacts on ammonia oxidizers dictate the irreversibility of N-induced changes. On the other hand, the microbial community legacy effects may also contribute to durable changes in N cycling due to N addition (Johnson, 1993; Bradley et al. 2006). We hypothesized that both retention and ammonia oxidizers properties influenced the reversibility of N-induced shifts in the ecosystem function. Here, we performed a 15 years long nutrient addition experiment (2000–2014) followed by a 7 years cessation of nutrient inputs (2015–2021) in the alpine meadow ecosystem to examine the effects of nutrient inputs on soil ammonia oxidizers. Our objective was to compare the effects of nitrogen addition and cessation effects on soil ammonia oxidation and nitrification and the underlying mechanisms during these two periods.