SUGARCANE GENOTYPE SELECTION ON MUCK AND SAND SOILS IN FLORIDA - A CASE FOR DEDICATED ENVIRONMENTS

Traditionally, the cooperative sugarcane breeding program at Canal Point, Florida, has selected genotypes exclusively on organic Histosols, also known as muck soils, in the early to middle stages of the program, resulting in the possibility that many genotypes adapted to sand soils are discarded. The objective of this study was to determine the presence of genotype × soil interaction on muck and sand soils, amongst genotypes in the second clonal selection stage. A significant genotype × soil interaction for yield traits would warrant the selection of genotypes on sand soils at earlier stages. Two series (groups) of genotypes were planted on a Torry muck soil at Canal Point, and on a Margate sand soil near Clewiston. Limited seed cane necessitated planting each series in different years at different sites, resulting in a confounding of soil with year. Genotypes were not replicated aside from check cultivars and a small subset of genotypes. All genotypes were analyzed for cane and sugar yield traits. Genotype × soil (G × S) and genotype × year (G × Y) interactions were modeled using the check cultivars. The results from the checks suggested that significant genotype × environment (G × E) interactions amongst the new genotypes are due more to G × S than G × Y. Significant G × E interactions were detected amongst the replicated new genotypes for six traits in the 2006 series and four traits in the 2007 series. For all new genotypes, the Spearman (rank) correlation was calculated for each trait across locations. All correlations were statistically significant, but the values were low (0.18 to 0.40). The exception was stalk weight, which was moderately correlated (0.49) between locations in both series. In practice, only one clone selected at both locations, CP 06-2042, advanced to the final stage of selection. Our results suggest the breeding program should begin genotype selection on sand soils at earlier stages in order to increase the likelihood of identifying superior cultivars for sand soils.