Development of rapid tests to predict germination and vigour and their potential for automation using image analysis

Project: Other

See relations at Aarhus University


There are demands for rapid methods of predicting germination and vigour. Published research has revealed that early radicle emergence (RE) counts and measurements of electrical conductivity (EC) predict normal germination (NG) and vigour of two Brassica species. This project will 1) assess the potential of both tests to predict germination and vigour of commercial seed lots of five other Brassica species, and 2) examine the application of two systems of image analysis, RGB digital imaging (GEVES system) and multispectral analysis (Videometer), to assess radicle emergence. Seed lots covering a range of NG, above and below the commercially acceptable level, will be used to identify 1) lots with unacceptable NG 2) differences in germination between lots with acceptable NG. Work on vigour will focus on lots with high acceptable NG; initial vigour assessments will be made using the Controlled Deterioration test and the rate of, and final, emergence in field or glasshouse trials. Automated germination progress curves, produced using the GEVES digital imaging system will select times for RE counts that predict NG. RE counts using multispectral imaging will reveal if the two image analysis systems give reproducible data and predict NG. The potential for measurements of the EC of seed soak water, taken at different times during 24 hours soaking to predict NG will also be examined. Both the EC and the selected RE counts will be compared to assessments of vigour to determine if they can be used as vigour tests. Members of the Germination and Vigour Committees will take part in comparative tests on two of the species using the selected RE and EC assessments. These will use both manual assessments and image analysis systems to test their ability to assess NG and vigour and to provide data on repeatability and reproducibility for test validation.
Effective start/end date01/09/201901/03/2021

ID: 194774301