| Urban soil system is a most important part of urban ecosystem constitution and the key to the sustainable development of urban ecosystem. As the process of urbanization continues to accelerate, urban environment and soil quality is faced with a common problem of quality degeneration, hence reinforce the research on the relationship between plants and soil quality, deepen the understanding of mechanism how plant communities improve soil quality and make good use of the research result on urban construction have gained more concern across the whole society. Different types of plant communities has been selected in Beijing Olympic Forest Park as to study the correlation between plant growth and soil quality improvement. Methods as data collection, field investigation and laboratory analysis were applied as to measure 19 indicators on soil physical, chemical and biological proprieties. Through the mathematical comparison of surface soil properties between natural and landscape plant communities, differences and correlations were found. In general, the mechanisms and results of plant communities’impact on the variability of soil characteristics in a newly built urban park were discussed, suggestions were proposed to the design and management of landscape & park. Main conclusions are as follows:(1) Main species composition of park’s plant community was native species, basicly formed a good plant landscape, though excessive planting exists. Secondary succession phenomenon were observerd in part of the plant community. A total of 83 plant species belonging to 32 families and 64 generas were recorded in 100 herbaceous samples during investigation, in which Asteraceae, Poaceae, and Fabaceae family occupied the majority. Urban Park pedogenesis was affected by human interference from the source and during the later development period. Human interference from the source was construction waste and artificial backfill municipal waste burying; and the later development influences concluding tourists’satmp, over conservation such as fertilization and defoliation cleanup.(2) Mean bulk density of park samples was 1.48 g/m3, while mean maximum water holding capacity and capillary capacity were 337.25 g/kg and 220.75 g/kg, and mean capillary porosity was 31.9%. Water capacity of soils under landscape plant communitites presented a limited improvement compared with soils from the surface of bare land, but as compared with the soils under natural communities, garden soils still showed significant degeneration in water holding capacity. Multiple comparison among five landscape plant communitites as theropencedrymion, coniferous forest, broadleaf forest, shrub and grass land on soil physical characteristics showed forests superior to other plant types, and in which coniferous forests persented as the best.(4) Park soil sample presented as alkalescent, mean conductivity was 102 us/cm. Mean nitrogen, phosphorus, potassium and organic matter content of park soil was 0.104%, 0.175%,0.375%,1.16%; the element content showed relatively a lack of organic matter, nitrogen and potassium, but an excessive of phosphorus. The soil C:N ratio was 6.43. Differenences were not significant between samples in electrical conductivity, total potassium and total phosphorus. Compared with adobe soils, pH under plant communitites tended transfer to neutral, total nitrogen and organic matter content were significantly increased, but still can’t reach the level of natural soils. In the comparison of different plant community types, organic matter and nitrogen content of theropencedrymion and coniferous forest were found significantly higher than others.(4) Soil bacteria, fungi and actinomycetes quantities in landscape samples were 1.3×105 cfu/g,0.87×102 cfu/g,1.73×104 cfu/g, microorganism activities were mainly determined by bacteria biomass; The average activity level of urease, alkaline phosphatase, invertase and catalase were 1.48 mg·g-1·24h-1,1.86 mg·g-1·2h-1,0.02 mg·g-1·24h-1,1.87 ml·g-1· 20min-1. Biological indicators of landscape communities all showed a rise than adobe soil, but inferior to natural communities. In the comparison of different forest types, bacteria biomass, urease activity, invertase activity was found that arbor> shrub> meadow, which had some similarities with changes in organic matter and nitrogen; other indicators had no obvious regularity in multiple comparison.(5) Correlation analysis of garden soil samples’physical, chemical and biological indicators generally showed significantly or extreme significantly correlated between same kind indexes. And soil organic matter content was correlated with the most of indicators. Principal component analysis were used to get a soil quality index (SQI), which indicates integrated soil quality ranking as natural forest> coniferous forest> theropencedrymion> broadleaf forest> grassland> shrub> surface adobe. Clustering analysis results were similar. According to the membership function analysis and sensitivity analysis for different indicators contribution value, main indicators of garden soil quality were organic matter content, soil bulk density, soil bacteria, soil pH and soil enzyme activity. Binding canonical correspondence analysis (RDA) were used to assess the interaction betweent plant diversity and soil quality, results showed that high tree diversity was in favor of soil physical properties improvement. Herbal diversity is easily influenced by environment, which showed a negative correlation with soil bulk density, pH and other soil properties.This paper studied basic questions of ecological landscape design theory from the aspect of soil subject, which has a good reference value upon the new ecological landscape theory. |