Abstract: Desert steppe is an ecologically fragile zone transitioning from grasslands to deserts, which is sensitive to global change. Primary productivity in desert steppe is generally limited by both nitrogen (N) and phosphorus (P). To clarify the effects of N and P addition on soil quality and screen suitable methods for soil quality evaluation, we conducted a long-term field experiment in the desert steppe of Yanchi County, Ningxia with four treatments: control (CK), N addition (10 g N·m^-2·a^-1), P addition (10 g P·m^-2·a^-1), and combined N and P addition (10 g N·m^-2·a^-1 + 10 g P·m^-2·a^-1). Based on 12 measured soil physicochemical and biological parameters, we used principal component analysis (PCA) combined with Norm values to select the minimum data set (MDS), and calculated and evaluated the soil quality index (SQI) by both linear and nonlinear models. The results showed that: 1) The screened MDS composed of three indicators: microbial biomass carbon, N-acetyl-β-glucosaminidase, and total phosphorus, with a filtering rate of 75%, which could effectively substitute the full dataset in soil quality evaluation. 2) Compared with the nonlinear model, the linear model based on the MDS exhibited higher fitting accuracy with the full dataset (R²=0.60 vs. 0.56), with a higher coefficient of variation (17.8% vs. 13.5%), and was more sensitive to soil quality variation, indicating better applicability. 3) N and P addition significantly improved the SQI, with the combined N and P addition exhibiting a significantly higher improvement (17.9%) than individual N (4.7%) and P addition (14.7%). In conclusion, combined N and P addition was an effective measure for improving soil quality in desert steppe, and the linear SQI model based on MDS was suitable for evaluating soil quality.

Key words: desert steppe, combined nitrogen and phosphorus addition, soil quality index, minimum data set