Conference on Applied Statistics in Agriculture

UNREPLICATED SPATIAL DESIGNS COMPARED USING OPTIMALITY CRITERIA

In field trials including large numbers of varieties, it is often impossible or impractical to replicate each variety. In these situations, the researcher may choose to use only one replicate of each test variety and to include a "check" variety every so often so that the spatial variability of the field may be determined. Five different check patterns were purposefully designed, each possessing distinct characteristics. The purpose of this study is to determine which spatial patterns for the check variety are better able to identify the spatial structure in a field and to rank the experimental varieties accurately. The problem was approached in two ways. First, the check patterns were compared using optimality criteria. Then, the patterns were applied to an actual field experiment, and the data collected was used to identify the spatial structure of variation in the field and to test for experimental variety differences. It is shown that the results from the optimality criteria were not necessarily comparable to what was actually observed in the field.