Optimizing hydroponic salinity screening: a novel approach using image-based phenotyping for double haploid rice selection

Abstract

The development of effective selection criteria and models under hydroponic salinity screening can be used image-based phenotyping (IBP) and statistical analysis to detect double-haploid rice with high adaptability to saline environments. Therefore, this study aimed to identify the IBP selection criteria and develop a model for potential tolerance in double-haploid rice under hydroponic salinity screening. The experimental design was a split-plot randomized complete block design. The main plot was NaCl concentration (0 mM and 120 mM), and the subplots contained ten genotypes with three replications. Selection criteria and model development were identified systematically and comprehensively through the best linear unbiased estimation, stress tolerance index, principal component analysis, factor analysis, and selection index. In addition, validation was also carried out based on conventional morphological characteristics, physiology, Na + and K + contents, and yield in saline land. The results showed that there are two tolerance index models: a morphometric (geometric) index represented by the total area and green area, and a colorimetric index defined by the green area percentage, CIVE, and GLI. The interaction of these indices effectively mapped the double-haploid rice genotypes based on their tolerance levels and adaptability to salinity stress. The colorimetric index was a reliable indicator of the potential adaptability of double-haploid rice lines in saline fields. This study provides a novel approach for developing effective selection criteria and models for rice tolerance, especially double-haploid line, under hydroponic salinity screening, which can accelerate the identification of genotypes with high adaptability to saline environments.

Code and data availability

The article describes image-based phenotyping of double-haploid rice under hydroponic salinity screening, but no paper-specific public dataset, image collection, or analysis code with an explicit authors' public URL is provided. The data availability statement only says datasets are in online repositories without names

No evidence-backed public reproduction asset is currently recorded.