【Objective】Grasslands in Songnen plain are propitious to the develop the mechanized-graziery due to its flat terrain and the climatic conditions, which influences by temperate continental monsoon climate. Moreover, influenced by the high plant coverage and diversity, black soil was developed, and the high quality forage species was dominated by Leymus chinensis in the past decades. However, the fragile eco-environment of Songnen grasslands could be impacted by the special topography, high salinity and shallow groundwater, and the parent material. Soil parameters are typically used to quantify the relationship between soil properties and forage yield and to identify the limiting factors of grassland productivity. Because of overgrazing and overdevelopment, the Songnen grasslands were degraded and salinized significantly. The husbandry was remarkable hindered by the low grassland productivity. The objectives of this study were: 1) to investigate the spatial variation characteristics of soil physical and chemical properties; 2) to estimate the spatial variation characteristics of vegetation community; 3) to explore the potential linkages between dry yield of alfalfa and soil physical and chemical properties. 【Method】 The artificial grassland locates at the Xinfawopeng village of Baicheng City, which covers an area of 600 m×450 m. The soil sample points were set by grid method, with the spacing of each neighbor sample points was 50 m. The soil samples were collected from alfalfa artificial grassland from by cutting rings and soil auger in October 2017 and May 2018 for detecting the soil physico-chemical properties, meanwhile, the forage samples were collected in a quadrat of 1 m2 (1 m×1 m). All the soil samples were air-dried, ground and then passed through a 2-mm sieve for chemical properties. The 1:5 of soil-water extracts were used to determining soil pH, EC, Na+, Ca2+, Mg2+, HCO3-, and CO32-. The spatial variations of soil properties, including soil mass water content (MWC), soil pH, EC, and total alkalinity (TA) were determined by classical statistics and geostatistics. Furthermore, the spatial variations of biological characteristics of artificial grassland, such as the Shannon-Wiener Index (SWI), dry yield (DY), shoot height (SH), and cover degree (CD) in the study area, were analyzed with classical statistics and geostatistics. 【Result】At the depth of 0 ~ 30 cm, very high spatial variability of the EC and TA were found, however, the pH and MWC showed moderate spatial variations; The biological properties of the grassland community showed high spatial variations. Results of multiple linear regression (MLR-step wise) suggested that the alfalfa dry biomass (DM) could be predicted by (DM) = 2699.73-276.496 pH (7.17< pH <9.76), which indicated that the soil pH influenced the alfalfa yield more than other soil properties. 【Conclusion】The above results indicated that the soil physico-chemical properties of artificial grasslands in western Songnen pain were high and moderate spatial variations, and the alfalfa could not tolerance the pH higher than 9.76.