Automated phenotyping for early vigour of field pea seedlings in controlled environment by colour imaging technology

What is the problem?

Early vigour of field pea can be improved by established breeding practices but the lack of a high-throughput phenotyping tool is an obstacle of developing such genetic improvement programs.

Australian researchers from Agriculture Victoria in Horsham carried out trials in the automated plant phenotyping platform of Plant Phenomics Victoria to investigate how effectively an automated phenotyping platform could estimate various early vigour traits seen in pea seedlings.

What did the research involve?

• Field pea seeds were selected to ensure that seeds had a similar level of germination and then grown for use in two different experimental conditions.
• Australian and New Zealand variety pea plants (44 genotypes) were grown in a greenhouse and loaded onto the fully automated plant phenotyping system after harvest. The growth conditions in the greenhouse were controlled with a 12 h photoperiod at 24ËšC during the day and 18ËšC during the night.
• All 44 pea genotypes were trialled during the 2016 cropping season in the open-air field with the experimental site consisting of Vertosol heavy clay soil characteristics in average annual rainfall of 450 mm and a temperate climate.
• Traits measured included shoot biomass, top-view characteristics, water use efficiency, growth rate, height and leaf area.

What were the key findings?

The actual biomass measurements and the values estimated by the automated plant phenotyping system were closely related.

While there was significant variation within the estimations, the estimated biomass, estimated water use efficiency, top-view area and, to a lesser extent, the relative growth rate and estimated height were the most consistent trait variables.

Final comment

Nguyen et al. (2018) have developed a high-throughput, automated, digital image plant phenotyping method.

The estimated biomass found using the automated phenotyping technology showed the strongest correlation to the field trial biomass observations (Pearson’s correlation coefficient = 0.98) and estimations based on top-view area appeared to also show some correlation to observed values (Pearson’s correlation coefficient = 0.70).

Details of all 44 genotypes tested in this study can be found in the study’s supporting information (S1 Table).