A Robust Automated Image-Based Phenotyping Method for Rapid Vegetative Screening of Wheat Germplasm for Nitrogen Use Efficiency

Giao N Nguyen; Pankaj Maharjan; Lance Maphosa; Jignesh Vakani; Emily Thoday-Kennedy; Surya Kant

doi:10.3389/fpls.2019.01372

A Robust Automated Image-Based Phenotyping Method for Rapid Vegetative Screening of Wheat Germplasm for Nitrogen Use Efficiency

Front Plant Sci. 2019 Nov 5:10:1372. doi: 10.3389/fpls.2019.01372. eCollection 2019.

Authors

Giao N Nguyen¹, Pankaj Maharjan¹, Lance Maphosa¹, Jignesh Vakani¹, Emily Thoday-Kennedy¹, Surya Kant^{1

2}

Affiliations

¹ Agriculture Victoria, Grains Innovation Park, Horsham, VIC, Australia.
² Centre for Agricultural Innovation, The University of Melbourne, Melbourne, VIC, Australia.

Abstract

Nitrogen use efficiency (NUE) in crops is generally low, with more than 60% of applied nitrogen (N) being lost to the environment, which increases production costs and affects ecosystems and human habitats. To overcome these issues, the breeding of crop varieties with improved NUE is needed, requiring efficient phenotyping methods along with molecular and genetic approaches. To develop an effective phenotypic screening method, experiments on wheat varieties under various N levels were conducted in the automated phenotyping platform at Plant Phenomics Victoria, Horsham. The results from the initial experiment showed that two relative N levels-5 mM and 20 mM, designated as low and optimum N, respectively-were ideal to screen a diverse range of wheat germplasm for NUE on the automated imaging phenotyping platform. In the second experiment, estimated plant parameters such as shoot biomass and top-view area, derived from digital images, showed high correlations with phenotypic traits such as shoot biomass and leaf area seven weeks after sowing, indicating that they could be used as surrogate measures of the latter. Plant growth analysis confirmed that the estimated plant parameters from the vegetative linear growth phase determined by the "broken-stick" model could effectively differentiate the performance of wheat varieties for NUE. Based on this study, vegetative phenotypic screens should focus on selecting wheat varieties under low N conditions, which were highly correlated with biomass and grain yield at harvest. Analysis indicated a relationship between controlled and field conditions for the same varieties, suggesting that greenhouse screens could be used to prioritise a higher value germplasm for subsequent field studies. Overall, our results showed that this phenotypic screening method is highly applicable and can be applied for the identification of N-efficient wheat germplasm at the vegetative growth phase.

Keywords: broken-stick model; controlled environment; digital imaging; high-throughput phenotyping; plant growth analysis.