Yields of adjacent rows of sorghums in variety and spacing tests

Two years' data are presented on the comparative yields of individual rows of five-row plats of sorghums in variety tests in an endeavor to find to what degree yields of adjacent varieties are influenced by competition and to find whether or not border rows are essential to the accuracy of variety tests under conditions prevailing at the Oklahoma Agricultural Experiment Station. The yields of individual rows did not differ significantly from the average yields of the inside rows protected from intervarietal competition or from the average yields of all five rows of the plats. Neither the yields of single rows nor those of combinations of rows showed significant differences. The outside rows of plats in the variety tests exhibited a somewhat higher degree of variability than did inside rows. Combinations of rows show a slightly lower degree of variability than single rows, as would be expected. The differences in the values of the standard deviations of the respective rows or combination of rows were not consistent. In the greater number of cases, they were not of sufficient magnitude to be significant. There was a tendency towards a higher degree of variability in the yields of respective rows in the earliest and latest dates of planting. Such greater variations may be expected as varieties are planted at times removed from their respective optimum dates of planting. The individual yields of the two outside rows of plats in variety tests and the yields of an inside rows show a high degree of correlation. The values of the coefficients of correlation in no case differed significantly for the same set of outside rows when correlated with the yields of the respective inside rows of plats. This shows that the two respective outside rows of the plats were influenced either in like manner or not at all by rows of adjacent plats. A marked degree of competition was found in spacing tests. The yields of rows of dense stands profit at the expense of yields of adjacent rows of thinner stands. The degree of competition is affected by environmental conditions. All data presented lead to the conclusion that single-row plats replicated frequently enough will give as reliable results as will plats with a larger number of rows repeated less frequently. In the utilization of single-row plats care must be exercised in the selection of varieties to grow next to each other. Extreme differences in growing habits must be avoided. Results of the spacing tests serve to emphasize the importance of uniformity of stands to the accuracy of variety tests.