摘要
he influence of soil chemical properties and soil nutrition on cotton yields was studied by means ofestablishing mathematical models. The nultivariate quadratic regression equations developed by a stepwiseregression method not only presented the single effect of soil factors but also displayed the interaction (syn-ergistic or antagonistic) of soil nutrients. The effect of individual factor and the way of nutrient interactionwere further analysed by the path analysis method. The results showed that among major factors affectingcotton yields, there existed the interactions between macronutrients(available P × available K), and betweenmacronutrients and microelements (N×Zn, P×Mo, P×Cu, P × Zn, K×Mo) besides the single effect of soilpH, total P, available Cu and available Zn.
