Gross nitrogen ammonification and nitrification rates
In our study, the degree of the response of gross ammonification rates (GA) to N addition was lower than that of gross nitrification rates (GN) under lower levels of N addition, and GA and GN were strongly affected by the higher levels and frequencies of N addition tested, suggesting that GA and GN were closely related to the concentration of NH4+-N and NO3--N in soils. Few studies had examined how N addition affected gross N turnover in grassland ecosystem. Our results indicated that gross N turnover responded positively to N addition, which was in agreement with numerous studies (Barnard, Leadley, & Hungate, 2005; Niboyet et al., 2010). However, gross N mineralization does not always increase linearly with increasing N deposition levels. This has been observed in studies of forest ecosystems (Corre, Beese, & Brumme, 2003; Gao, Yang, Kou, & Li, 2015; Tietema, 1998) in which the rates of gross N mineralization increased up to an intermediate ambient level of N enrichment, but then dropped somewhat at N-enriched conditions higher than it in these systems. Differences between their results and ours are likely due to differences between the two different types of ecosystems studied and environmental conditions. Walecka-Hutchison & Walworth (2007) found that gross nitrification was stimulated by the lower N addition, while was inhibited by the higher N application, which disagreed with our results. Such discrepancy may be caused by the differences of soil nutrient availability, microbial diversities and soil enzyme activities. In addition, we found that soil pH decreased with increasing levels of N addition, and gross N turnover rates significantly decreased with increasing soil pH. This result was inconsistent with the findings of Cheng et al. (2013), in which GA and GN were positively correlated with soil pH in forests, which likely due to differences in the gross N turnover processes of different ecosystem types, and the major limiting factors are different in forest and grassland ecosystems. Furthermore, many studies indicated that GA and GN were influenced by some other factors aside from soil N availability, e.g., vegetation and soil types (Gerschlauer et al., 2016; Xu & Xu, 2015), soil moisture (Y. Chen, Borken, Stange, & Matzner, 2011), and other human or environmental factors(Wang, Dannenmann, Meier, & Butterbachbahl, 2014; West, Hobbie, & Reich, 2006). However, studies about the effects of N addition frequency on soil gross N turnover were very limited, so further studies are still needed.
In addition, we found that mowing had negligible effects on GA and GN under any N addition conditions, which is different from the findings of Wang et al. (2015), who found that mowing increased nitrification rate and did not affect ammonification rate. This difference between our study and theirs may be due to the possibility that the response of GA and GN to mowing was counteracted by the level and frequency of N addition. On the other hand, mowing might also affect soil gross turnover by changing the environmental conditions (i.e. soil moisture, temperature and so on) in grassland ecosystems (Bremer, Ham, Owensby, & Knapp, 1998; Han, Zhang, Wang, Jiang, & Xia, 2011; Owensby, Hyde, & Anderson, 1970). And previous studies indicated that mowing might increase soil total carbon and nitrogen by increasing plant diversity (Cong et al., 2014; Fornara & Tilman, 2008) and change the compositions of soil microbial communities (Li et al., 2017), which would directly or indirectly impact soil gross N turnover. However, few studies have evaluated the response of gross N turnover to mowing, so as a consequence, our knowledge of the response mechanism involved in such relationships is very limited.