Masters Theses

Date of Award

8-1999

Degree Type

Thesis

Degree Name

Master of Science

Major

Plant, Soil and Environmental Sciences

Major Professor

Donald D. Tyler

Committee Members

H. Paul Denton, John B. Wilkerson, Michael D. Mullen

Abstract

Crop yields have been shown to vary both between and within fields. Current technology allows for an accurate measurement of yield variability using Global Positioning System (GPS) and yield monitoring equipment. However, determination of the source of this variability is complicated by spatial differences in soil fertility, soil series, slope, and past management practices.

This statewide study was designed to test the effectiveness of conventional soil survey maps against an intensive soil map created for various sites in Tennessee. Field Specific soil maps were developed using intensive soil sampling, incorporated with GPS and Geographic Information System (GIS) software. Relationships between soybean yield and soil mapping units were then statistically compared using spatial correlation models and S AS proc mixed procedures. Intensive soil maps better explained soybean yield variation (α = 0.05) although neither mapping technique was strongly related to yield. The site-specific maps were also better at distinguishing distinct yield groups by individual soil mapping unit.

Specific properties of the soil and crop landscape were also investigated to determine their affect on soybean yield. Properties that had a significant affect on yield included subsoil texture, slope, and pH. Slope and subsoil texture interactions and drainage and effective rooting depth (ERD) interactions also showed yield differences. Soil properties that did not affect yield included soil drainage class, ERD, available phosphorous, and available potassium. Interactions of ERD and subsoil texture, ERD and slope, drainage and subsoil texture, and drainage and slope also showed no yield differences.

Results of this study indicate that conventional mapping methods may not provide the necessary detail for use in today's precision farming applications. When investigating specific properties within soil units, a limit is set in explaining yield variation by the soil unit boundary and further variability related to the soil unit cannot be explained. Although site specific maps are better than conventional mapping methods at predicting yields, an investigation into specific soil properties within a field may be necessary in providing a useful tool for producers implementing precision farming crop management.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS