Soil Type, Precipitation, and Fertilizer N Effects on Corn Yields

Abstract

Consideration of the multitude of factors to maximize the efficiency of N inputs, particularly from purchased fertilizer N, suggests a systems approach. One facet in our development of a systems model for Wisconsin was an evaluation of existing data on soil and climate factors influencing corn (Zea mays L.) grain yields and yield response to applied fertilizer N. This paper reports on a study using multiple regression techniques to examine the relationships among soil type, pre-and early-season precipitation, fertilizer N rate, and corn grain yields from numerous continuous corn plot studies. Nitrogen fertilizer rates required to maximize yields were lower on rainfed silt loam and silty clay loam soils (155 to 210 lb N/acre) than on irrigated loamy sand or sandy loam soils (175 to 230 lb N/acre). Similarly, economic optimal N rates were considerably lower on silt loam and silty clay loam soils than on irrigated sandy soils, especially at low N:corn price ratios. Regression models that included fertilizer N rate, soil N, and precipitation variables were examined. Linear and quadratic fertilizer rate variables explained only 10 to 38% of the variation in grain yields on silt loam or silty clay loam soils, whereas more than 65% of the variation was explained on irrigated sandy soils. As would be expected, pre- and early-season precipitation were important factors with the rainfed soils. Other key variables included available N contributions from soil organic matter and residual profile NO3. Although not measured directly, soil physical properties (soil depth, water holding capacity) also probably contributed to yield variations on rainfed silt loam and silty clay loam soils. Importantly, we demonstrate that economic optimal N rates would often be considerably less than N fertilizer rates required to attain maximum yield. Also, grouping of soil types with similar organic matter contents, rooting depths, and drainage characteristics for N management recommendations is feasible.