4 Discussion
Despite some fertilizer inputs to the cropland in this study, significant losses of SOC and STN occurred in cropland compared with grassland, which suggested that the nutrient pools may be vulnerable to grassland conversion over time. This was consistent with previous studies that demonstrated that the conversion of grassland to cropland usually induces SOC and STN loss (Don et al. 2011; Wang et al. 2011; Ding et al. 2013). It is mainly because the removal of plant biomass reduces nutrient cumulate in soil, which in turn decreases the microbial biomass because less energy is available from soil organic matter decomposition (Schnitzer et al. 2011). Moreover, the breakdown by tillage results in large aggregates transforming into fine aggregates (Figure S2), which makes them more susceptible to erosion by wind and water that will lead to soil nutrient loss (Six et al. 2000). However, soil C:N ratios were significantly different between grassland and cropland sites. This indicates that the decrease in nitrogen could keep up with the pace of soil carbon loss after grassland cultivation.
In this study, grassland conversion consistently reduced the total soil microbial biomass, Gram+, Gram-, Fungi, AMF and Act. The greater microbial biomass under grassland soil has been previously reported (Jangid et al. 2011) as there are more favorable soil environmental conditions for microbes under grassland resulting in larger microbial biomass. The decreasing soil microbial biomass induced by the grassland conversion probably resulted from the declines in the concentrations of SOC and STN, which provide energy sources for microbe turnover. Crop growth during growing season (August is the mid-growing season in this ecotone ecosystem) significantly influences soil microbial biomass because crops compete with microorganisms for substrates (Zhang et al. 2012), and the higher vegetation biomass in cropland soils than in grassland soils during the growing season suggests that crops require more available nutrients, which might lead to the rapid depletion of labile substrates without inputs from plant residues and rhizodeposition (Bever et al. 2010; Zhang et al. 2012). In addition, the soil microclimate is cooler and moister under grassland soils compared to the drier and warmer in cropland soils. Loss of soil nutrients and water has been linked to increased susceptibility to other stresses (soil pH and soil aggregates) (Bever et al. 2010).
We found that neither F:B ratio nor relative abundance of bacteria and fungi changed with grassland conversion. This indicated that grassland conversion not only reduced the soil microbial activity but also altered the soil microbial community structure. One possible interpretation might be that grassland conversion affect microbial composition primarily by altering the soil nutrient (soil organic carbon, total nitrogen), as indicated by the correlation analysis. Soil microbial community may acclimate grassland environments and sustained a relatively stable structure. A higher F:B ratio in topsoil of grassland is associated with higher decomposition efficiency and greater carbon storage potential in soil (Ding et al. 2013). Moreover, this indicated soil fungi is likely to more sensitive and easily degradation to grassland reclaim (Poeplau et al. 2011), implying that the turnover rate may increase in whole microbial communities (Six et al. 2000). Previous studies have suggested that soil fungi often dominates the decomposition of soil organic matter because the lower nutrient demands and metabolic activities than bacteria in a low nutrient content (Jangid et al. 2011; Ma et al. 2015; Moon et al. 2016). Furthermore, the translocation of nutrients can be promoted by the hyphae of soil fungi (Klein et al. 2004), which would decrease significantly in response to physical disturbances (Helgason et al. 2008; Drenovsky et al. 2010), such as plowing under cultivation, so grassland soils are more favorable for the formation of fungal hyphal networks that play an significant role in the cycling of soil carbon and nitrogen (Hu et al. 2014).
The site (S), land use (LU) and soil depth (SD) have significantly effect on soil microbial biomass (Table 3). However, there was no statistically significant interactions between site (S) and the others factors (LU and SD). It indicated that decrease in soil microbial biomass after grassland cultivation is a prevailing phenomenon among different site. Moreover, soil depth has been observed to be a determinant of SOC and STN concentrations and soil microbial community composition in grassland soils. Soil depth provided varied environmental conditions as reflected by changes in soil characteristics and PLFA biomass, and the decreases in soil nutrient with increasing soil depth may be a major reason for the pronounced decrease in soil microbial biomass in lower grassland soil layers (Guo et al. 2002; Ding et al. 2013; Moon et al. 2016). In contrast, the decreases in soil microbial biomass, SOC and STN were less pronounced in cropland because tillage practices homogenize the soil substrates and resources across the plow layer (Drenovsky et al. 2010), thus leading to higher soil microbial biomass, SOC and STN in 0-30 cm soil layer in cropland.
Our results revealed that grassland cultivation affected microbial biomass mainly through enhanced soil nutrient resources rather than and soil pH, moisture and aggregation. This finding is consistent with recent studies reporting that resource availability controlled the responses of the plant soil system to land use change (Lange et al. 2015).These findings suggest that soil microbes are highly vulnerable to grassland cultivation and that this vulnerability is determined by the disruption of feedback processes between soil nutrient properties and soil microbes due to grassland conversion. It is essential to effectively evaluate soil properties before grasslands are converted to cropland (Jangid et al. 2011), especially in agro-pastoral ecosystems that are particularly sensitive to environmental changes and are difficult to restore.