|Birrell, Stuart - UNIV OF MO|
|Kitchen, Newell - UNIV OF MO|
Submitted to: Information Agriculture Conference
Publication Type: Abstract Only
Publication Acceptance Date: June 30, 1995
Publication Date: N/A
Technical Abstract: Site-specific nutrient management depends on soil maps to develop fertilizer recommendations and yield maps to monitor the effects of the management. The ability to generate variable rate application maps using established soil fertility recommendations and to map the resulting yield is a major step forward. However, full benefits will not be realized until lthe critical factors that cause variations in yield can either be related to a measured crop growth parameter or identified as being caused by unknown parameters. The present conventional nutrient recommendation procedures generally assume that only factors that are part of the recommendation process are limiting. If this is not the case, these recommendation procedures may be invalid. In the claypan soils of Central Missouri, the dominant factors determining crop yields are the climatic conditions and moisture status of the soils. However, even without the effect of climate, conventional correlation statistics do not adequately identify critical factors. When a field consists of several areas with different limiting factors, the correlation between any single nutrient and yield on a field basis will be low, even if the correlation with yield for the limiting factor within an area is high. When the possible effects of pests, weeds, and other factors are included, the situation is even more complex. At this point the ability to develop mathematical procedures to automatically identify the critical parameter(s) within different areas of the field appears remote. However, mapping of soil nutrients and other important factors can form the basis from which a knowledgeable person can identify likely limiting factors and then apply the appropriate mathematical procedure(s) to test the hypothesis.