3.1 SMB under low vegetation cover level in dryland regions
Our study summarized that in the dryland regions, the amount of SMBC and SMBN were 358.47 ± 25.45 mg kg-1 and 51.86 ± 4.59 mg kg-1 in the 0-30 cm soil profile, which were almost half of the previous terrestrial ecosystems study (SMBC 680.4 mg kg-1and SMBN 105 mg kg-1, Xu et al. 2013). It has been speculated that the discrepancy might be due to the study regions. Dryland regions are part of land and just cover about 41% of Earth’s land surface, which include four main ecosystem types: grassland, desert, cropland and forest. The total area of these four ecosystem types accounts for at least 65% of the dryland regions area; meanwhile, forests which with high vegetation cover only account for 18% of the dryland regions area. (Millennium Ecosystem Assessment, 2005). Desert and grassland have lower vegetation cover rate and plant diversity compared to tropical forest (Schuldt et al., 2019; Conti et al., 2013; Cook-Patton et al., 2011). Vegetation cover and plant diversity influence the plant quantity directly and the quality of plant litter and root exudates which in turn indirectly affect the soil condition (Drenovsky et al., 2010; Xiao et al., 2017); meanwhile, the microbial processes of carbon and nitrogen cycles were affected by soil condition due to the differences in quality and quantity of nutrient (Bargali et al., 1993; Kara et al., 2008).
Consistent with our hypothesis, the amount of SMBC and SMBN were significantly different among ecosystem types in dryland regions (Fig. 5 ). Different ecosystems develop unique patterns of adaptation, especially under extreme conditions such as drought. Desert which was the main ecosystem type in drylands region is with sparse vegetation cover, in turn the soil in desert was lacked of ample root networks and plant litter to provide enough nutrient to soil microbes (Yang et al., 2014; Luna et al., 2016; Gang et al., 2012). The reason for low amount of SMBC and SMBN in the desert was likely a deficiency available nutrient content by low degree of aboveground vegetation cover. To some extent, the greater amount of organic matter accumulates may indirectly affect soil microbial activity and abundance, then soil becomes more suitable for microbial growth (Shrestha et al., 2008).
According to the plant-soil-microbes theory supported by many studies, the response among the three parts is significant and sensitive. The roots-dominated soil usually shows a greater chance of nutrient supply and cycling, thus enhances the contents of SMBC and SMBN around the root zone especially in desert soil (Jia et al., 2017). Jia et al. (2017) found after plant restoration in an arid desert of northern China, the vegetation cover rate increased and the greater amount of plant litter enters the soil, which in turn governs the soil organic matter and nutrients, consequently the content of SMBC significantly increased. In the grassland ecosystem, grazing was one of the major and special factors to affect the ecosystem nutrient cycles. There was no doubt that microbes and animals, such as sheep, competed for ecosystem nutrient. When grazing was reduced, higher organic matter inputs to soil from root exudates and plant litter which are more conducive to microbial growth, and then the contents of SMBC and SMBN increased (Wu et al., 2014). On the contrary, some researchers found no evidence of a difference between SMB and land cover, soil quality was important instead (Moghimian et al., 2017).
3.2 stoichiometry of soil and soil microbial biomass
The stoichiometry of SMB from open-access papers and our study were together summarized in the table (Table 2 ). We indicated that the soil C/N and SMBC/SMBN in dryland regions was 10.69 and 8.73, which were different compared to other publications. Firstly, the data only included research from drylands region (aridity index < 0.65) in our study, concluding grasslands and deserts mainly (accounted for 75.4%); while the data in other studies like Xu et al. (2013) and Cleveland et al. (2007) were extracted from global region, including all ecosystem types. Differences in vegetation cover and plant diversity are the main reasons. Secondly, the sampling depth used in our study was different from that of the studies at global scale (i.e., 0-30 cm vs. 0-10 cm). There were a lot of studies that show the microbes were distributed differently in the vertical direction of soil, and the content in the top layer soil was higher than that of the deep soil (Jia et al. 2017; Xu et al., 2013).
The ratio of SMBC to SMBN was an indicator of soil stoichiometry balance and was close related with microbial C use efficiency (CUE) and N use efficiency (NUE), which was important in predicting efficiency of nutrient cycling (Mooshammer et al., 2014). The definition of CUE or NUE is a ration of nutrient allocated to growth over nutrient taken up by microbial community composition (Manzoni et al., 2012; Sinsabaugh et al., 2013). High ratio of soil C/N meant soil is rich in available C (N deficiency), microbes would have high NUE and low CUE, finally lead a low degree of SMBC/SMBN. In our study, the ratio of soil C/N was 10.69, which waslower than other previous results, indicated that in dryland regions, the soil was at a high substrate N sufficient condition. Thus, it was not surprising that microbes would be adjusted to high CUE, then lead to a high SMBC/SMBN. This inference was consistent with our research result, the ratio of SMBC/SMBN in our study was 8.73, which was higher than other results.
Providing additional evidence that the SMBC/SMBN was higher in dryland regions, we also found a previous study shown that soil with high water content usually had more available C than available N. Actually, it would result in lower SMBC/SMBN. The study reinforced our results. Dryland regions often were characterized by water deficiency, which is a plausible explanation for SMBC/SMBN in dryland regions was lower than global average. To the best of our knowledge, the mechanisms behind this relationship are not clear, and this might be also related to differences in microbial communities at different soil depths.