Application of Statistical Design Theory to Crop Estimation with Special Reference to Legumes and Mixtures of Cultivars¹

Abstract

Abstract Forage crop researchers often need to determine the proportion of legume, weed, and grass contents of hay. This is often done by sampling and hand‐separation or by visual estimates of the relative proportions. The sampling and hand‐separation method is costly and time‐consuming, and both procedures are subject to errors of separation and vagaries of sampling. As alternatives, procedures employing mixtures of cultivars and spring balance weighing design and balanced, incomplete block design theory are presented. Errors and biases associated with sampling and hand‐separation and visual‐estimate procedures are eliminated. Also, a procedure is presented to measure the general competitive effects of cultivars and to determine how much more (or less) effective a cultivar is in a blend than when grown alone.