We present a recently developed suite of R functions for the design and analysis of agronomic trials. The functions were developed with multi-environment plant breeding trials in mind, but are widely applicable to field-based trials in general. Users can generate classical RCB designs or more flexible un-replicated designs augmented with replicated checks. Users also have the option of using a design created elsewhere (i.e. the Agricolae package in R), which can be manipulated using available functions to work with the interface of the analysis functions.

The analysis function has been designed to be as flexible as desired. Users can input the design matrices for the fixed and (multiple) random effects, covariance structures for the variance parameters, as well as the structure of the residual error term. Alternatively, users can provide the proper model and the model terms will be automatically built. If not defined in the covariance or residual error structure, users can opt for the function to fit a moving average covariate and/or a correlation structure accounting for spatially-related error, such as an AR1 x AR1 correlation structure.