Comparison of a Posteriori Statistical Procedures to Reduce Spatial Variation in NDVI Measured at First Bloom in Cotton.

Breeding efforts are increasingly being directed towards the development of high-throughput phenotyping technologies in order to keep pace with advancements in genotyping. As the number of genotypes breeders are able to phenotypically characterize increases, experimental error attributable to soil heterogeneity will also increase. Traditionally, experimental error is minimized through field design, but research also suggests that a posteriori statistical procedures in combination with appropriate experimental designs improve the prediction accuracy of genotypic performance. The objective of this study was to compare the efficiency of a posteriori statistical procedures, specifically nearest neighbor (NNA) and moving means (MVNG) adjustments, at reducing spatial variation in normalized difference vegetation index (NDVI) in cotton (Gossypium hirsutum). NDVI was measured at first bloom across multiple breeding trials at Florence, SC in 2014 and 2015. The efficiency of NNA and MVNG at reducing spatial variation for complete and incomplete block experimental designs will be discussed.