Prediction of Soybean Root Response in the Field Using Nondestructive Seedling Three-Dimensional Root Features

Abstract

Core Ideas: A three-dimensional root reconstruction approach is accurate in predicting root response under field conditions. We investigated whether a laboratory-based root screening method could identify germplasm with better field performance. Prospects are good for implementing root-trait based selection in a crop breeding program. The root angle and fibrous root score determines soybean [Glycine max (L.) Merr.] productivity in water-limited environments. Studying root traits in the field is technically difficult to do nondestructively, and varying agroclimatic and edaphic factors create challenges, which could be circumvented using laboratory-based root studies with higher precision. In this study, we developed and validated a nondestructive root imaging platform and derived three-dimensional (3D) root features using algorithms that would predict the root response under field conditions. Nine soybean genotypes (five landrace accessions and four cultivars) were evaluated at seedling and reproductive stages in controlled environments and the field. Irrespective of the phenotyping platforms, soybean landraces showed higher phenotypic variation for different root architectural traits than cultivars. The plant introductions 200471 and 567731 had higher phenotypic values for root and shoot traits than cultivars Dunbar, Magellan, and Maverick. The seedling 3D root features—volume and surface area/volume ratio—were found to be the best predictors of field root response evaluated. This positive association is a promising step in using a nondestructive imaging system to screen soybean with a desired root system that matches different target environments. Thus, the selection based on accurate 3D root features will hasten the inclusion of root traits in soybean breeding programs specifically to improve productivity in harsh environments.