Abstract: Variation in soil microorganism populations has a great influence on soil nutrient cycles during the process of ecological restoration. However, little research has been conducted into variation in soil microorganism populations during the restoration of sub-humid desert land in alpine-cold areas. To investigate this topic, we planted a mixture including Avena sativa, Elymus nutans and Festuca sinensis, among others as an experiment in the restoration of sub-humid desert land. Changes in microbial biomass carbon, nitrogen, phosphorus, microbial biodiversity (phospholipid fatty acids, PLFAs), soil organic carbon, soil total nitrogen and soil total phosphate were checked in the 0-10 cm and 10-20 cm soil layers, annually after re-vegetation, and the changes in soil microorganism populations with time and the correlations between microbial biodiversity, microbial biomass and soil nutrients were analyzed. It was found that re-vegetation increased soil microbial biomass from 8.30 to 10.58 nmol·g^-1 after 4 years and the bacteria still accounted for more than 50% in the total amount of microorganisms. Microbial biomass carbon, nitrogen, and phosphate, microbial biodiversity (i.e. bacteria, fungi, actinomycetes, Gram-negative bacteria, Gram-positive bacteria), soil organic carbon, total nitrogen and total phosphate in the 0-10 cm soil layer were all more than in the 10-20 cm soil layer, throughout the re-vegetation process. Over time, all of these measures showed an initial decrease, and then a tendency to increase. For example, after 1 year’s re-vegetation, soil microbes and nutrients of both soil layers decreased significantly, but soil microbes and nutrients increased in the 0-10 cm soil layer after 2 years re-vegetation. Finally, soil microbe numbers and nutrients increased to values similar to or higher than the controls after 3 or 4 years. In particular, soil organic carbon and microbial biomass nitrogen were significantly (P<0.05) higher in restored grassland than in control plots at the end of the experiment. Soil nutrient levels had extremely positive correlations with soil microbial biomass, and microorganism and bacterial counts (P<0.01), and had positive correlations with counts of soil fungi, actinomycetes, Gram-negative bacteria and Gram-positive bacteria. In summary, artificial re-vegetation with a grass mixture increased soil nutrients and microbes in desert land, thus providing a positive contribution to restoration of a degraded ecosystem in this alpine-cold area where this study was carried out.

Key words: compound grasses, alpine-cold and sub-humid, sandy restoration, soil microorganism, soil nutrients