Improving Efficiency of Sugarcane Genotype Selection in Florida

The Canal Point (CP) sugarcane (Saccharum spp.) breeding program has developed highyielding cultivars for the organic soils of South Florida, but cultivar development has not been as successful for sand soils in this region. Our objectives were to improve fi nal-stage (Stage 4) selection effi ciency of this program by assessing interactions among genotypes, soils, and locations and other sources of variation and by comparing the relationship between repeatability and number of replications as a measure of precision for testing sugarcane genotypes on sand and muck soils. Sources of variation were partitioned for plant-cane, fi rst-ratoon, and second-ratoon data, from 7 yr and fi ve CP series in Stage 4 tested at eight locations on muck soils and two locations on sand soils. Variation related to genotype × location (soil) was highly signifi cant for cane and sugar yields for all series and crops, but most genotype × soil interactions were nonsignifi cant. This suggested that planting more Stage-4 tests on sand may improve genotype selection for these soils. The six replications used routinely in Stage 4 provided elevated levels of precision for both soil types. All locations with muck and sand soils in Stage 4 have ample testing resources for discriminating among genotypes. Our results indicate that, to improve sugarcane genotype selection for sand soils, increasing the number of sand locations would be a more successful strategy than increasing replications within locations. USDA-ARS, Sugarcane Field Station, 12990 US Hwy. 441 N., Canal Point, FL 33438. Received 23 Sept. 2009. *Corresponding author (ali