Mapping variability of soybean leaf potassium concentrations to develop a sampling protocol

• Published in: Agrosystems, geosciences & environment geosciences & environment, 2023-12, Vol.6 (4), p.n/a
• Main Authors: Ortel, Carrie C., Roberts, Trenton L., Hoegenauer, Kyle A., Poncet, Aurelie M., Slaton, Nathan A., Ross, W. Jeremy
• Format: Article
• Language: English
• Subjects: Agricultural production; Crop production; Data collection; Error analysis; Fertilizers; Hunger; Interpolation; Leaves; Loam soils; Methods; Nutrition; Potassium; Potential fields; Profits; Sampling; Soil testing; Soybeans; Spatial dependencies
• ISSN: 2639-6696; 2639-6696
• DOI: 10.1002/agg2.20439

The spatial variability of soybean [Glycine max (L.) Merr.] leaf‐K (potassium) concentrations must be considered when collecting samples to monitor crop K status. Five commercial soybean fields were sampled at a 0.4‐ha grid resolution at two reproductive growth stages to quantify the trifoliolate tissue‐K concentration. The objectives of this study were to identify the potential field variability of soybean leaf‐K concentrations in typical Mid‐south US soybean production fields, evaluate interpolation methods, and develop a sampling protocol for in‐season soybean tissue monitoring. No consistent spatial dependencies were found in the leaf‐K concentrations across the fields and sample times, indicating that a soybean tissue‐K grid sampling protocol cannot be generalized to a specific area size. Inverse distance weighted (IDW) and rasterization interpolation methods were considered to predict leaf‐K concentrations between the sampled grid points at grid resolutions ranging from 0.4 to 4 ha. The IDW method consistently predicted leaf‐K concentrations between the known values with less error than rasterization. Rather than grid sampling, composite leaf samples should be collected based on management zones to provide a simplified sampling protocol. Within each management zone, a composite sample must consist of uppermost fully expanded trifoliolate leaves collected from at least 18 locations to ensure that the sample measures within the 95% confidence interval of the area average leaf‐K concentration. The developed sampling protocol coupled with the dynamic critical tissue‐K concentration curve will provide producers with the ability to effectively monitor soybean for potential hidden hunger and verify K deficiency symptoms in season. Core Ideas Soybean leaf‐K (potassium) concentration was not consistently spatially autocorrelated at 0.4‐ha grid resolutions. Inverse distance weighted interpolation consistently predicted leaf‐K values with less error than rasterization. One composite sample should include leaves collected from at least 18 points within the area of interest. Soybean leaf samples should be collected based on management zones to monitor leaf‐K concentrations.