|Mitchell, J - UNIVERSITY OF CALIFORNIA|
Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: November 4, 1999
Publication Date: N/A
Technical Abstract: Managers need decision tools for sustainable management. Indices of soil quality may fill this need. We tested two methods for choosing a minimum data set (MDS) and two methods of index formation. Ten years of chemical, biological, and physical data from vegetable rotations of the Sustainable Agriculture Farming Systems project near Davis, California were used. The MDS components were chosen using principal components analysis (PCA) or a hierarchical decision tree. Indices were created using additive or a decision support system. The efficacy of each minimum data set was tested by multiple regression of MDS variables against management variables such as yield, % weed cover, and water use. For the MDS chosen by PCA for tomato, r(^2) values were 0.64 against yield, 0.97 against weeds, and 0.99 against water use. The corn MDS regressed against yield, weeds, and water use had r(^2) values of 0.81, 0.90, and 0.90, respectively. Results are reported for organic, low-input, and conventional treatments.