Yield Variability in Proso Millet Due to Plot Size 1

Abstract Field experimentation with a new crop requires determination of optimum plot size to minimize variability and obtain an unbiased yield estimate. Plot size influences the planting and harvesting equipment, and the amount of land needed to conduct a research trial. Variability in proso millet ( Panicum miliaceum L.) yields due to number of rows planted and harvested in field experimentation was compared using four cultivars for 4 years on Typic argiustoll soils. Plot size combinations selected were: plant one row, harvest one row; plant two rows, harvest two rows; plant three rows, harvest one row; plant four rows, harvest two rows; plant six rows, harvest two rows; plant six rows harvest four rows. The results of this experiment indicated: 1) three harvest rows were optimum to minimize variance based on the best quadratic fit, 2) the number of rows harvested had a greater effect on variance than the number of rows planted, 3) plots having one border row of their own genotype on each side produced higher yields than plots with either zero or two border rows on each side, and 4) adding border rows of the same cultivar on each side of the harvested row did not decrease variance. The optimum plot size combinations for testing proso cultivars in field experiments appears to be planting five rows and harvesting the center three rows.