AN APPLICATION OF POLYNOMICAL STEP-REGRESSION OPTIMUM MODEL IN THE STUDY OF PLANT-SOIL CORRELATION IN TIANTONG,ZHEJIANG,CHINA

In this paper, the polynomical step-regression method was used to study on the relationships between 20 plant species and 5 soil factors: Nitrogen, Phosphorous, Organic matter, Water content and pH value. The limiting soil factors of each species were selected to build the optimum models of plant-soil environmental correlation. The effect of soil factor on the interspecific correlation has been discussed as well. The results show that the polynomical step-regression is excellent to study on the nonliner correlation between plant species and soil variables.