Multilocation, multiyear testing
Multilocation trials are an essential part of plant breeding because of the ubiquity of GxE interactions.
The number of locations is very dependent on seed/planting material availability and the level of resources available.
Locations need to be representative of the environments where a new variety will hopefully be grown.
Multiyear is necessary to assess stability of performance over seasons. In later generations, the size of each plot tends to increase (this improves the reliability of the trial), because planting material availability is increased and the number of entries to be tested falls.