Effect of Harvest Area and Replication on Detection of Treatment Differences in Sugar Beet Field Experiments

Abstract

Yield and sucrose determinations from field plot trials are a vital part of sugar beet research programs concerned with breeding, varie­ tal evaluation, soil fertility, disease, pest control, or other aspects of sugar beet production. Various plot sizes and harvest sampling tech­ niques are used, most of which seem to be based on general experience and convenience rather than a systematic evaluation of the errors encountered in such experimentation. Surprisingly , the authors have been unable to locate any recent publications concerning optimum harvest areas, plot size, or the magnitude of differences one might expect to detect using various experimental techniques. In 1942, the ASSBT Standardization Committee recommended that sample areas two rows wide and 30 feet in length be harvested for demonstration strip trials, and for varietal trials that plots two to eight rows wide and 30 to 75 feet in length be used with the entire Flot being harvested if possible (1)3. No justification for these recommendations was given nor were any references made to supporting experimental data. The most recent publications found specifically evaluating the effect of numbers of samples or sample areas on experimental error are those of Immer (3,4,5) . This work represents one of the earliest applications of the analysis of variance to sugar beet research in this country. These papers essentially describe the results of uniformity trials , and while the results are still relevant, it would seem appropriate to evaluate the errors involved under present conditions.