Often in field experiments, the plots are arranged in a rectangular grid pattern of rows and columns, such that a linear or non-linear function of row or column position
can be fit as a covariate in the model to account for some of the field variability.