Variations in stand as sources of experimental error in yield tests with corn

In order to determine the sources of error in harvesting plat trials with selfed lines and crosses of corn an experiment was conducted at University Farm, St. Paul, and at Waseca, Minnesota, for the two years, 2928 and 1929. Selfed lines of dent varieties and single and double crosses of both flint and dent varieties were used at St. Paul, while at Waseca only selfed lines and single crosses of selfed lines of Silver King dent corn were used. With six selfed lines in the University Farm test, the average yield per plant for 3-plant hills, 3 1/2 feet apart, was 91.7% of that for single plants spaced 1 foot apart. Single plants of selfed strains, when planted 3 feet apart, gave an average increase, at University Farm, of 35.7% over 1-foot spacing in the row, while 2-foot spacing's have a 17.3% increase. The data indicated that the less vigorous strains attained maximum growth at less than 3 feet apart in the row. At Waseca the selfed lines were planted in hills 3 1/2 feet apart each way. There was evidence of competition, one blank hill adjacent to a 3-plant hill resulting in a 6.0% increase in yield of that hill, while two adjacent blanks gave a corresponding 5.5% increase as an average for two years. Apparently, two blank hills resulted in no greater increase than a single blank hill. One-plant hills yielded only 39.9% as much as perfect-stand hills. With single crosses at University Farm in 1928, the average yield of a 3-plant hill when adjacent to two blanks hills, one blank hill, and one 2-plant hill with blanks on the corners was increased 12.6, 7.0, and 4.8%, respectively, over the yield of 3-plant hills which had competition on all sides. Similar comparisons were made with double crosses in 1929 for 3-plant hills adjacent to two blanks, one blank, and one 2-plant hill, with corresponding increases of 13.1, 16.2, and 3.8%. The six dent crosses gave practically no increase in 1928, while two flint crosses showed marked response to extra room for development. At University Farm in 1928, 2-plant hills with blank hills on two comers yielded 85% as much as 3-plant hills surrounded by a full stand. In 1929, 2-plant hills surrounded by a full stand yielded 86.7% as much as corresponding 3-plant hills. The results obtained for crosses at Waseca were similar for the two years. As an average of both years, failure to discard single 2-plant, 1-plant, or blank hills resulted in losses of 15, 40, and 67%, respectively, in yield from the space allotted to the hill with deficient stand over a 3-plant hill. Two-plant hills of F1 crosses yielded 24.6% less than 3-plant hills, at Waseca, as an average of both years. At University Farm in 1929, 2-plant hills yielded 20.6, 8.4, and 16.2% less than 3-plant hills for the dent, flint-dent, and flint crosses, respectively, or an average of 13.3%.