No description available.

The Matsqui area of the Lower Fraser Valley exhibits extreme soil heterogeneity, as the alluvial soils in the area have been deposited by the Fraser River as a series of coarse-textured ridges and finer-textured depressional areas. This variability poses some obvious problems with respect to agricultural management. The main aim of this study was to evaluate soil spatial variability in four fields, and to relate this soil variability to corn production and quality. Site conditions, topography, and soil chemical and physical variables were related to corn biomass and nutrient concentrations using conventional correlation/regression analyses, and more spatially representative techniques such as those provided by remote sensing and geographic information systems. Variations in such biophysical variables as soil moisture, elevation, and bulk density had consistent impacts on corn productivity, although these effects varied from field to field, being influenced by inherent soil properties and individual field management. Good relationships were found between pixel brightness values extracted from digitized colour infra-red photos and corn quality variables. In three out of four fields, near infra-red pixel values gave good estimates of total corn crude protein content. Significant relationships were also found between pixel brightness values and corn phosphorus and calcium contents in certain fields. The spatial variability of corn quality and biomass could be quantified using image analysis classification techniques. The resulting classified images indicate to the farm operator where high vs low quality corn is being produced, and thereby provide a tool for selectively managing and harvesting the fields. The relationships and quantification of corn productivity and quality in the fields can further be improved through incorporation of the image data with the biophysical data base using GIS techniques. A multiple regression equation showing a significant relationship between elevation and pixel brightness values, and total corn phosphorus concentration was incorporated within the GIS to produce a quantitative corn quality map for the field exhibiting this relationship. The GIS overlay capability facilitates the classification of several corn variables, and allows the results to be displayed in a spatial manner. For example, corn biomass and quality maps were overlain using GIS techniques, to produce a combination map which then reflected both the quality and quantity of corn found in the field. Through integration of remote sensing and GIS techniques, soil and crop variability can be displayed in a spatial manner. The output from such procedures can aid farm operators in making selective field management and harvesting decisions. / Land and Food Systems, Faculty of / Graduate