SUBSET-SELECTION IN PLANT-BREEDING PRACTICE

Selection decisions in variety testing should partly be based on statistical motives. Useful tools for this are subset selection procedures. These procedures can also be used when the performed experiment has an incomplete block design or when a series of regional trials is studied. Various selection rules serve different selection goals, but they all need so-called selection constants. Often, these constants have to be approximated by computer simulation. For this simulation, and also for the execution of the selection rules, software has been developed and is available. Some practical adjustments and modifications of subset selection in plant breeding practice are proposed. Finally, a case study of selection in sugar beet is presented.