4 DISCUSSION
4.1 Bacterial, fungal, and protist communities differ among climate and vegetation types
Bacterial diversity was significantly affected by MAT and vegetation type, whereas fungal and protist diversity were influenced by the vegetation type alone. Elevated temperatures in humid regions enhance SOM decomposition by accelerating microbial activity and turnover (Allison et al., 2010; Zhao et al., 2019a). The high soil exchangeable Ca content of karst soil inhibits SOM decomposition by forming stable aggregates (Briedis et al., 2012; Wiesmeier et al., 2019). Our previous study found that high exchangeable Ca during vegetation restoration improves SOC accumulation and negates the effect of temperature increases on SOC decomposition (Hu et al., 2021). The fact that MOC was higher under the plantation forest and shrubland of Guangxi than it was under those of Guizhou suggests that readily decomposable organic C was more rapidly degraded than recalcitrant C at high temperatures (Fig. S1; Samal et al., 2020). Hence, labile organic C decreased with increasing temperature. Bacterial groups could be stimulated under the condition of high labile organic C (Poll et al., 2008). Consequently, there was higher bacterial diversity under the plantation forest and shrubland of Guizhou than there was under those of Guangxi. In contrast to bacteria, mycorrhizal and other fungi are highly dependent on their mutualistic relationships with plants (Newsham et al., 2009; Schmidt et al., 2014). Fungi utilize the polymerized litter fraction such as lignin and cellulose (Fontaine et al., 2011). High plant residue input under plantation forest and shrubland increases fungal diversity by supplying abundant organic substrates (Chen et al., 2017; Waldrop et al., 2006). This phenomenon explains our observation that fungal richness was greater under plantation forest and shrubland than it was under cropland except for plantation forest at Guizhou. Conversely, the bacterial Shannon index was higher under cropland than shrubland. Bacteria have a growth advantage over fungi in highly fertile agricultural soils (Cai et al., 2018). Enhanced bacterial diversity under cropland promotes protist consumers and results in higher protist diversity under cropland than vegetation restoration after cropland abandonment (Guo et al., 2018). Overall, these results suggest that high fungal diversity under vegetation restoration may contribute to SOC accumulation by stimulating plant residue decomposition.
Temperature and vegetation type influenced bacterial, fungal, and protist community composition. Bacterial phylum Proteobacteria prefer labile organic carbon and their abundance is positively correlated with C/N (Fierer et al., 2007; Hermans et al., 2017). Here, C/N was higher at Guizhou than Guangxi (Fig. S1). Therefore, Proteobacteria had higher relative abundance at the former than the latter location. Compared with cropland, shrubland provided greater available organic C by increasing plant residue inputs. For this reason, Proteobacteria were more abundant under shrubland than cropland. Basidiomycota include symbiotic fungi while Ascomycota include free-living saprotrophs (Baldrian et al., 2011; Read and Perez‐Moreno, 2003). Basidiomycota promote plant litter decomposition (Baldrian et al., 2011; Toljander et al., 2006). Consequently, there was greater Basidiomycota abundance under plantation forest and shrubland than cropland. Moreover, Ascomycota abundance was greater under high-AP conditions (Dang et al., 2017). Thus, relative Ascomycota abundance was higher under cropland than plantation forest or shrubland because cropland had comparatively higher AP content (Figs. S1 and S3). In general, anthropogenic disturbance and environmental pressure are greater in cropland than plantation forest or shrubland (Santos et al., 2020). Protist community composition under cropland was markedly different from that vegetation restoration (Grossmann et al., 2016; Santos et al., 2020; Schulz et al., 2019). For this reason, Ciliophora, Lobosa, and Ochrophyta were more abundant under cropland than they were under plantation forest or shrubland. Certain Apicomplexa are plant parasites and were more abundant under plantation forest and shrubland than they were under cropland (Seppey et al., 2020). Taken together, the foregoing observations indicate that bacterial, fungal, and protist community composition markedly differ between cropland and vegetation restoration.
Contrary to our hypothesis, bacterial, fungal, and protist diversity and community composition under plantation forest and shrubland were similar in the subtropical karst region. The southwestern karst region is geologically fragile, and intense anthropogenic disturbance there has led to rapid nutrient loss (Wang et al., 2019; Wen et al., 2016; Xu et al., 2021; Zhang et al., 2013). In general, soil nutrient levels were limited in the early stages of plantation forest and shrubland restoration in the southwestern karst region (Hu et al., 2021). Soil available nutrient levels are key factors regulating microbial diversity and community composition. As bioavailable nutrient content was similar between plantation forest and shrubland, they did not differ in terms of microbial diversity or community composition as vegetation restoration progressed. Nevertheless, plant diversity was greater in shrubland than it was in single-species plantation forest (Hu et al., 2020). These results suggest that in the karst region with plantation forest and shrubland restoration, nutrient availability plays a more important role than the plant community in determining bacterial, fungal, and protist communities.