Factors indirectly and directly regulate ecosystem functions
We also explored the direct and indirect effects of environmental factors on ecosystem functions under land use change (Fig. 6), which are pivotal for better understanding the underlying mechanisms of diversity-ecosystem functioning relationship under varying environments (Brose & Hillebrand 2016; Peters et al. 2019; Wang et al.2019). We found that the substrate supplying factors (e.g. TC, TN, and C/N) were directly associated with the mSR (Fig. 6a), highlighting the evidence that substrate quality dominated ecosystem functions (Nieet al. 2013; Zhang et al.2016). It has been well established that increased soil C and N supply could stimulate microbial activities as they were sources of nutrients and energy to microorganisms (Kaiser et al. 2016; Yang et al. 2018). Although environmental factors, especially soil pH has been considered as best predictor of microbial community composition (Fierer 2017; Yanget al. 2018) and ecosystem functioning (Delgado-Baquerizoet al. 2017; Zhou et al. 2020), here it was evidenced that the indirect environmental factors predominately controlled over ecosystem broad function by shifts in microbial community richness (Figure 6). The possible explanation for the indirect effect was that environmental factors span a narrow range, and did not rapidly respond to land use change (Fierer 2017). Thus, the positive relationship between the substrate supplying and microbial broad functions suggested that the microbial respiration was probably both energy- and nutrient- limited across land use types in our study site (Trivedi et al.2016; Zhang et al. 2016).
By comparing ammonium availability in yellow-brown soil of our study site (0.8 mg/Kg, low level) to black soil (1.89 mg/Kg) of other studies e.g. (Liu et al. 2018), we inferred the PNR was maintained by substrate supplying factor irrespective of alterations in the microbial communities. However, in our study, the NR was directly related with environmental factors (especially, soil temperature), and the PNR was directly regulated by the AOA and AOB keystone species community (Fig. 6b, c). This result accorded with a previous study that found soil temperature was the important driver of global soil nitrification rate (Li et al. 2020). Some studies have observed niche differentiation between the AOA and AOB communities across different land use types (Dai et al. 2018; Hink et al. 2018; Liuet al. 2018). The inconsistency between our results and previous hypothesis indicated that land use changes in alkaline soil did not significantly affect the AOA and AOB community diversity (Fig. 2). This founding could be explained by a previous review with neutral pH and low nitrogen availability (Shen et al. 2012).