Spatial Variability of Plant Nutrients in Two Illinois Fields

197 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1993.The development of global positioning technology and variable rate fertilizer application equipment at the same time in which environmental and economic considerations guide many grower decisions concerning fertilizer application have raised old questions concerning the spatial variability of soil supplied plant nutrients in fields. Two Illinois fields which have been soil sampled in a 25 meter grid for several previous years and having a known history were selected to determine the variability of soil pH, and available P and K in a field environment.Soil and plant samples were taken from each field in a 25 meter grid. The maps constructed using geostatistical software showed important nutrient level features. A single fertilizer application rate over a field increased the variability of the field, while a variable rate application of a nutrient based on sound soil test information decreased the variability within the field. The rate of change of soil nutrient levels over time was not constant, but the variability of the change was as complex as the variation in soil fertility levels.Commercial variable rate applicators rely on mapping using a linear variogram model to develop the kriging equations. The linear variogram was not always the best fit model, but was most often a good fit. The use of a good fit over the best fit model did not appreciably change the resulting map which might be used to guide variable rate application.Analysis of several grid densities suggested underlying variability both man-made and natural. Samples taken in a 100 meter grid, which is currently used in some commercial variable rate programs, did not identify many important soil nutrient features. The study suggests that a 66 meter grid would be a minimum grid for directing variable rate application.Plant analysis showed that major features of nutrient availability could be mapped with plant analysis as well as soil test results. A rapid method for plant analysis, and two methods of soil nutrient analysis were also evaluated which may make densely sampled fields more practical.U of I OnlyRestricted to the U of I community indefinitely during batch ingest of legacy ETD