ABSTRACT

Development of spatial technologies, in particular yield-monitoring systems, has greatly enhanced our appreciation for the extent of spatial divergence in cotton production fields. Given the importance of fiber quality in determining the value of the cotton, there is considerable interest in understanding the spatial variability of cotton fiber properties and its dependence on underlying field variability. Previous work has relied on hand harvesting of the cotton at intervals throughout the field, and measuring the fiber properties of the harvested lint. While this approach may be suitable for research plots or small-scale production fields, it is not amenable to full-scale production research. Moreover, the hand-harvested cotton has different fiber properties than cotton from the same area that is machine harvested. Both the harvesting and ginning methods alter the cotton fiber, and introduce deviations from the fiber qualities that would be measured by producers. We have developed a system of fiber sampling that removes a portion of the cotton during machine picking. These cotton samples are then ginned in a small research gin, and classed at the USDA-AMS Classing Office. Spatially registered maps are built of the various fiber qualities. Discounts based on the fiber properties and total lint discount are determined according to the Spot Cotton Quotations. Spatially registered maps of fiber discounts are also developed. Fiber value maps are combined with cotton yield and management inputs to determine spatially registered cotton profitability. Management scenarios that optimize profitability are then built for each of the discrete management zones as identified by the response of cotton growth, yield, fiber properties, and profitability.