Evaluation of cultivar-testing locations in sugarcane

Summary

Selection of test locations, representative of conditions and practices of an area can be a challenging process in a breeding program. Data from two groups of sugarcane (trispecies hybrids of Saccharum sp.) cultivar experiments in Florida were analyzed to determine if relative cultivar response at any two of six current locations was sufficiently similar so that at least one location could be replaced by a location with a different environment. The parameter analyzed was metric tons per ha of sugar (THS). To determine similarity between location pairs for all cultivars within each group of cultivars, an unbiased stability-variance parameter ($\hat \sigma _{ i}^2$) developed by Shukla was used. After $\hat \sigma _{ i}^2$identified similar location pairs, single degree of freedom interactions were calculated for important cultivars to determine which of the location pairs identified by $\hat \sigma _{ i}^2$contained the two most similar locations. Use of the above procedure can assist in making optimum location assignments in a breeding program.