Vijaya Raj Joshi1, Jeffrey A. Coulter1 and Axel Garcia y Garcia2, (1)Agronomy and Plant Genetics, University of Minnesota, St. Paul, MN (2)Agronomy and Plant Genetics, University of Minnesota, Lamberton, MN
Crop simulation models can predict crop growth, development and yield as a function of cultivar, soil, weather, and agronomic management. A study was conducted during 2016 in southwestern Minnesota to simulate site-specific maize growth and yield using the CERES-Maize model. Two fertilizer treatments, 0 and 118 kg N ha-1 were used; the latter was sidedressed at the six leaf-collar stage in strips across a heterogeneous 7-ha field. Georeferenced soil samples were collected within a grid from the 0-30, 30-60, and 60-90 cm layers before planting for total N, P, K, organic matter, and soil texture. Plant data collection included phenology and above ground biomass at the eight-leaf collar and tasseling stages, and grain yield at harvest. Soil textures determined at the study site were clay loam (CL), sandy clay loam (SCL), and sandy clay (SC) soil. The soil and plant data collected at each geo-referenced site were used to run and evaluate the CERES-Maize model. The model simulated the above ground biomass slightly better at tasseling stage (25% nRMSE) than at eight-leaf collar stage (31% nRMSE). The simulated yield (15% nRMSE) well captured the effects of soil texture and N treatments but was underestimated in SC soil. The preliminary results show the CERES-Maize as a promising tool to simulate site-specific maize biomass and yield. This is valuable to understanding spatio-temporal variation in maize response to N fertilizer at a field level and to develop precision N management strategies in maize production.