High-throughput phenotyping using plant pixel area to predict agronomic performance...
Humphreys G, Gahagan C, Kalikililo A, Sadeghi-Tehran P et al (2018) High-throughput phenotyping using plant pixel area to predict agronomic performance in Canadian winter wheat. Poster (P1085) and abstract presented at 2018 ASA/CSSA International Meetings Nov. 4-7, 2018 in Baltimore MD, USA.
High-throughput phenotyping using plant pixel area to predict agronomic performance in Canadian winter wheat.
Humphreys G, Gahagan C, Kalikililo A and Morrison M. Agriculture & Agri-Food Canada, Ottawa Research and Development Centre, 960, Carling Ave., Ottawa, Ontario, Canada K1A 0C6
Effective high-throughput phenotyping of wheat lines is desirable to improve selection efficiency in breeding programs and to facilitate the use of whole genome based selection methods. Green pixel area (GPA), expressed as a proportion of the total pixel number in photographs of yield plots, has been used previously as a useful measure of plant establishment, growth and vigor. In this study, three phenomics factors were used: (1) plant pixel area (PPA) which is a measure of green pixel area in a yield plot filtered to remove weed plants, (2) active canopy coverage (ACC) which is defined as mean growing degree days above 50% canopy, (3) linear senescence rate (LSR) which is the rate of loss of PPA after maximum canopy coverage is reached. The purpose of this research was to investigate the relationships between PPA, AAC and LSR with important agronomic traits including grain yield, heading date, plant height, test weight and seed mass. In 2017, mean phenomics parameters were determined from weekly green pixel area estimates for all entries in two advanced Canadian winter wheat yield trials (EA and MT). In both trials, ACC and LSR was significantly (P<0.05) correlated with grain yield. LSR was also significantly correlated with test weight and seed mass. AAC was significantly correlated with test weight and seed mass for the MT trial. PPA was not significantly correlated with grain yield in either trial. PPA was significantly correlated with heading date, plant height in the MT trial but not in EA. In conclusion, LSR which is a estimate of “stay green” potential was the most promising predictor of grain yield, test weight and seed mass for entries in this study.