Determining Relationships and Patterns in the Nutrisolutions® Plant Tissue Database Using Multivariate Analytics.

WinField™ is a data-driven company that delivers unmatched data and insights to help growers reach maximum yield potential. The NutriSolutions® 360 program offered by WinField™ provides plant tissue, soil and resin capsule analysis to optimize in-season plant nutrition and product applications. Since 2009, WinField™ has collected more than 300,000 tissue samples through the NutriSolutions® 360 program in 37 different crops. Despite the success of the NutriSolutions® 360 program, interpreting nutrient deficiencies and making recommendations based on tissue testing is still complicated by crop growth stage, nutrient interactions, and localized field conditions. The NutriSolutions® 360 tissue database provides an opportunity to use big data for deeper understanding and discovery of plant nutrition. Using multivariate approaches, 35,000 corn tissue samples collected from 2010 to 2014 in the state of Minnesota were analyzed. On average, 60% or more of the samples were less than adequate for nitrogen, sulfur, potassium, manganese, and zinc concentrations during vegetative growth stages from 2010 to 2014. Correlation coefficients between nutrients indicated relationships between nitrogen and sulfur, phosphorus, copper and between sulfur and copper. A relationship between nitrogen and potassium was not detected. Multivariate analysis confirmed univariate nutrient relationships while indicating synergy between phosphorus and zinc and antagonism between potassium and magnesium during vegetative growth stages. Deficiency of nitrogen, potassium, sulfur, manganese and zinc was the most common pattern followed by deficiency of zinc alone. Correcting nutrient deficiencies may require consideration of multiple nutrients simultaneously and micronutrients that interact with nitrogen, phosphorus, or potassium. Using big data and multivariate analytical approaches can reveal new insights to better guide nutrient application decisions.