Using variety testing data to select soybean varieties: Guidelines for practitioners

Abstract

Soybean [Glycine max (L.) Merr.] variety selection is a crucial decision that impacts farm profitability. Effective variety selection requires performance evaluation across diverse environments to determine whether differences are due to genetic or nongenetic factors. Producers should use different types of multiple-location variety trials to select high-yielding varieties with resistance or tolerance to biotic and abiotic stressors prevalent in their region. In addition, information on other traits (e.g., plant height, lodging, green stem) that can be obtained should be considered when selecting a variety. Two different types of variety trials that are normally conducted by university extension programs and seed companies consist of replicated small-plot trials and on-farm large strip-plot demonstrations. The data from these trials are normally published in printed publications, downloadable online data sheets (e.g., PDFs, spreadsheets), and user-friendly online selection tools. The objectives of this management guide are to (a) describe replicated small-plot trials and unreplicated on-farm large strip-plot demonstrations, including attributes of each type; (b) demonstrate a data-driven approach to selecting high-yielding varieties with resistance to biotic and abiotic stressors; and (c) discuss additional agronomic and seed quality traits that aid in understanding the differences in varieties. Data and interpretation from multiple-location official variety trials and on-farm strip-plot demonstrations from the Louisiana State University Agricultural Center were used to demonstrate results commonly available from university extension programs. Evaluating yield and stress resistance data across multiple locations and years, including both replicated small-plot trials and on-farm strip-plot demonstrations, provides the most reliable basis for selecting varieties adapted to diverse and unpredictable environmental conditions.