Shape and size of plots in field experiments and spatial variability.

摘要

Soil variability in field experiments causes experimental error, which is accounted for by randomization and replication. When soil variability has a spatial component, it causes correlated errors within a block, which inflates the experimental error in field experiments and masks the true treatment effects. Shape and size of plots may affect the spatial patterns in field experiment. Uniformity field trial data on cotton yield conducted elsewhere were used during 2004 for studying the effect of shape and size of plots on spatial variability in yields. Various statistical procedures such as frequency plot analysis and semivariogram analysis were used to study the nature and magnitude of variability in the yield data obtained from different plot sites and shapes. Results showed that there was a considerable variation in yield data from different plot shapes and sizes. As the plot size increases, variability decreases. Spatial patterns in yields from different plot sizes were observed except plot sizes of 2x4 and 8x2 units, which showed random variability. So plot sizes of 2x4 or 8x2 units are optimum plot sizes for field experiments on cotton, which can be used to minimize the effect of spatial variability oil true treatment effects. As regards shape, square shape seems better for large plot sizes..