Image-Derived Traits Related to Mid-Season Growth Performance of Maize Under Nitrogen and Water Stress

Phenotypic measurements under controlled cultivation conditions are essential to gain a mechanistic understanding of plant responses to environmental impacts and thus for knowledge-based improvement of their performance under natural field conditions. Twenty maize inbred lines (ILs) were phenotyped in response to two levels of water and nitrogen supply (control and stress) and combined nitrogen and water deficit. Over a course of 5 weeks (from about 4-leaf stage to the beginning of the reproductive stage), maize phenology and growth were monitored by using a high-throughput phenotyping platform for daily acquisition of images in different spectral ranges. The focus of the present study is on the measurements taken at the time of maximum water stress (for traits that reflect plant physiological properties) and at the end of the experiment (for traits that reflect plant architectural and biomass-related traits). Twenty-five phenotypic traits extracted from the digital image data that support biological interpretation of plant growth were selected for their predictive value for mid-season shoot biomass accumulation. Measured fresh and dry weights after harvest were used to calculate various indices (water-use efficiency, physiological nitrogen-use efficiency, specific plant weight) and to establish correlations with image-derived phenotypic features. Also, score indices based on dry weight were used to identify contrasting ILs in terms of productivity and tolerance to stress, and their means for image-derived and manually measured traits were compared. Color-related traits appear to be indicative of plant performance and photosystem II operating efficiency might be an importance physiological parameter of biomass accumulation, particularly under severe stress conditions. Also, genotypes showing greater leaf area may be better adapted to abiotic stress conditions.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.