Blocking Direction in N.C. Wheat Performance Trials.

Most agronomists block their yield trials horizontally without consideration for variation. The direction of field variation may not always follow the slope of the field and is not normally known to the researcher. Even knowing the makeup of soil types in the field may not help since variability is not always associated with soils. Several have advocated blocking trials for operational purposes, i.e. in the direction of planting and harvesting, since blocking should be used for local control of variation and often times is ineffective. North Carolina Official Wheat Variety Trials were compared blocked horizontally and vertically with three years of data each. Data were analyzed using Proc Trend , a polynomial regression spatial analysis routinely used by the NC OVT Program, and the standard ANOVA. Vertical blocking resulted in a smaller, not significantly, standard error (163 vs. 175 kg/ha) compared to horizontal blocking when using spatial analysis but slightly larger, not significantly, standard error when using horizontal blocking (214 vs. 200kg/ha). Yields were higher and entry numbers were larger during the three years of vertical blocking compared to the years of horizontal blocking which should have resulted in larger s.e. During the three years of vertical blocking there were three incidents where vertical blocking allowed the trials to be planted and/or harvested while retaining integrity of the data where horizontal blocking would not have permitted without jeopardizing the data. I conclude that vertical blocking is feasible with spatial analysis, although it is not recommended without using spatial analysis.