Development of an automated phenotyping platform for quantifying soybean dynamic responses to salinity stress in greenhouse environment

Abstract

Soybean is a salt sensitive crop and saline soil results in substantial reduction in crop productivity. The genetic resources for salt tolerance in soybean were observed but are not fully evaluated due to the difficulty in capturing and quantifying the dynamic responses to salinity stress for large, diverse germplasm collections. This study aimed to develop an automated plant phenotyping system in an established greenhouse and validate this system in capturing dynamic responses in soybean to salinity stress. A digital camera (red, green and blue) was used to take sequential images of soybean for salt tolerance in 21 diverse accessions with three replicates (63 plants in total). Data processing pipeline was built to extract plant morphological and color-based image features. Thirty-nine image features were extracted and used to estimate plant leaf chlorophyll content and rate the soybean tolerance for salinity stress. Statistical analysis shows that the extracted image features could be used to quantify the dynamic responses of plants to salinity stress. It is found that both leaf chlorophyll content and visual salt tolerance rate (STR) had high correlations with image features. This study demonstrated the usefulness of the developed system in measuring soybean dynamic response to salinity stress.