Discussion
The relationship between microbial biodiversity and functioning under land use change
Our results provided direct experimental evidence that the association between soil microbial diversity and ecosystem functioning was primarily regulated by the keystone species under a subtropical land use change of China. As excepted, many studies have documented changes in the diversity of microbes and ecosystem functioning across land use types (de Vries & Shade 2013; Allan et al. 2015; Zhou et al.2020). A few studies have explored the biodiversity-functioning relationship varying in the natural assembled communities, which results may deviate from our familiar those in the experimental communities (van der Plas 2019). In natural ecosystems, biodiversity loss is random, and may be obscured by the abiotic factors (Turnbullet al. 2016; Veen et al. 2018). In this present study, our results emphasized that microbial richness was the principal factor in explaining the microbial respiration (mSR), and higher sensitivity of the mSR induced by changes in the microbial richness was observed in the afforested soils (Fig. 3; Fig. 4a). There were two possible explanations for these patterns. First, in phase with the positive linear diversity-ecosystem functioning relationship reported in experiments (Delgado-Baquerizo et al. 2016a; Rivett & Bell 2018), which indicated microbes contributed equally to the broad functioning (mSR). Second, it could be related to the microbial unique characteristics which might process much more rapid life cycle (e.g. hour to days) than in macro-organisms (e.g. plants, month to years) (Schimel & Schaeffer 2012; Delgado-Baquerizo et al.2016a). This in fact estimated a stable relationship when considering for relationship of microbial biodiversity-functioning across land use change. More especially, we found the negative relationship between the keystone species and mSR, suggesting the identify effect did not have a perceptible effect on the broad functioning, which occupied much resource than others to some extent (Trivedi et al. 2017b; Chen et al. 2020).
In contrast, our results showed that the biodiversity-functioning relationship of specialized functioning varied with land use types (Fig. 3d, e), possibly due to the heterogeneity across land use types. This also revealed the existence of some degree of functional redundancy in the amoA communities in afforested soils. It has been suggested that cultivation applications may act as an environmental filter selecting for those Nitrososphaera (affiliate to AOA), which are sensitive to soil organic N supply (Zhou et al. 2015; Gaoet al. 2018).