Changes in the interpretation of irrigation fertilizer experiments caused by blank alleys

Abstract

Field plot technique is of continued interest to and concern of scientists. End trimming to remove border effects is expensive and time consuming. To determine if there was a more appropriate method to analyze field plot results, a split-plot experiment with 4 replications of irrigated main plots and fertilizer subplots was utilized. Irrigation treatments consisted of irrigated and nonirrigated. Fertilizer treatments were ammonium nitrate and S coated urea at 0, 112, 224 and 336 kg N/ha. Corn (Zea mays L.) was used as the test crop. Plots consisted of 6 rows 81 cm apart. The center 2/3 of the 2 center rows of the 9.14 m long plots were harvested for plot yields. One remaining end of the center 2 rows (border) adjacent to a 3.05 m blank alley was selected at random within each replication and irrigation level to harvest for border yields. The relation of the plot yield to the border yield varied with irrigation level and growing season. Inclusion of borders indiscriminately in the harvest area when irrigation and fertility were both variables would lead to biased treatment differences for comparing fertilizer treatments at different irrigation levels or under irrigation.