Modeling local and global spatial correlation in field-scale experiments

Precision agriculture has renewed the interest of farmers and researchers to conduct on-farm planned comparisons and researchers with respect to field-scale research. Cotton (Gossypium hirsutum L.) yield monitor data collected on-the-go from planned field-scale on-farm experiments can be used to make improved decisions if analyzed appropriately. When farmers and researchers compare treatments implemented at larger block designs, treatment edge effects and spatial externalities need to be considered so that results are not biased. Spatial analysis methods are compared for field-scale research using site-specific data, paying due attention to local and global patterns of spatial correlation. Local spatial spillovers are explicitly modeled by spatial statistical techniques that led to improved farm management decisions in combination with the limited replication strip trial data farmers currently collect.