EXPERIMENTAL DESIGNS FOR PREDICTING CROP PRODUCTIVITY WITH ENVIRONMENTAL AND ECONOMIC INPUTS FOR AGROTECHNOLOGY TRANSFER

The papers in this compilation were presented at the Workshop on Experimental Designs for Predicting Crop Productivity with Environmental and Economic Inputs, held May 20-24, 1974 at the University of Hawaii and are arranged in three categories. The first section explains BSP"s objectives and its use of the U.S. Soil Taxonomy System to classify the soils tested in BSP experiments. This section also details the methodology of BSP experiments describing site and crop selection procedures and the stable and variable factors affecting soil and climate. The main objectives of the experiments were to prove the transferability of crops raised in similar soils and to develop a universal equation describing yield response for any one soil given appropriate variables. Papers in the second section describe approaches to, and results from, BSP field experiments. It was suggested that those proposing BSP fertility experiments should both determine: the country soil situation and the extent of tests appropriate for soil variables (nitrogen, phosphorus, potassium, stand, irrigation, etc.) and specify the variables and statistical and economic models used in applying the test results. Cornell University"s nitrogen and phosphorus soil experiments proving the high production potential of well-drained acid soils in the tropics are reviewed. The third section includes descriptions of experimental designs suitable for soil fertility experiments. Prediction equations and other mathematical models for estimating optimal inputs and outputs are discussed and recommendations are made for improving their utility. Topics include the successful extension in Japan of a fractional factorial design for a fertilizer experiment to 20 crop characteristics such as culm length, number of panicles per ear, etc., and economic aspects of production functions for soils such as the derivation of yield and profit functions for sugarcane. Bibliographies are appended to each paper and a list of workshop participants is provided.