Skip to main content
ARS Home » Pacific West Area » Pullman, Washington » Grain Legume Genetics Physiology Research » Research » Publications at this Location » Publication #380122

Research Project: Improving Genetic Resources and Disease Management for Cool Season Food Legumes

Location: Grain Legume Genetics Physiology Research

Title: Performance evaluation of pea and chickpea breeding lines across seasons and locations using phenomics data

item ZHANG, CHONGYUAN - Washington State University
item McGee, Rebecca
item Vandemark, George
item SANKARAN, SINDUJA - Washington State University

Submitted to: Frontiers in Plant Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/11/2021
Publication Date: 2/25/2021
Citation: Zhang, C., Mcgee, R.J., Vandemark, G.J., Sankaran, S. 2021. Performance evaluation of pea and chickpea breeding lines across seasons and locations using phenomics data. Frontiers in Plant Science. 12:640259.

Interpretive Summary: Pulse crops including peas, lentils, and chickpeas have been historically important crops for thousands of years and are commercially produced throughout the USA Pacific Northwest and Northern Plains. Developing improved pulse varieties through plant breeding typically requires careful evaluation of multiple traits across several years and locations. Traits must be evaluated over the entire life history of a crop, starting at seedling emergence and continuing through flowering, harvest, and at times even after harvest during storage. Specially trained employees are needed to effectively evaluate field traits such as early season vigor, mature plant height, days to flower, and days to harvest. There can be differences between employees in how they evaluate the same trait and an employee may inconsistently evaluate a trait during the course of a long day in the field. One way to overcome these difficulties associated with trait evaluation is to use 'remote sensing' technologies, which typically employ a remote-controlled unmanned aerial vehicle (UAV) that has a camera attached for taking pictures of plants in the field. However, in order for this approach to be effective there has to be a very strong association between the pictures, or images, and actual plant traits such as days to mature and most importantly, yield. The objective of this research was to determine the relationships between images of peas and chickpeas obtained with a UAV and important traits including days to flower and seed yield. Both crops were evaluated for three crop seasons (2017-2019) in Idaho and Washington. We determined that an image of a chickpea plot taken early in the growing season, approximately 30 days after planting, could be used to predict seed yield of the plot with about 80% accuracy. A similar picture of a pea plot taken about 30 days after planting could predict seed yield with about 60% accuracy. Differences between chickpeas and peas may be due to peas having modified leaves that are called tendrils, which are absent in chickpeas, and may have made it more difficult to interpret images of pea plots. The remote sensing approach was less effective at predicting other traits including days to flower and days to mature. However, this remote sensing approach apprears to have promise for evaluating yield, which is ultimtately the most important trait for most crops. Remote sensing can be used to identify potentially high yielding breeding lines early in the growing season and these lines can be the focus of more labor and time consuming evaluations for other traits such as early maturity.

Technical Abstract: The Pacific Northwest is an important pulse crop production and breeding region of the USA. Currently, breeders of pulses, including peas and chickpeas, rely on traditional phenotyping approaches to collect agronomic data to support decision-making. Traditional phenotyping poses constraints on data availability (e.g. number of locations and frequency of data acquisition) and throughput. In this study, phenomics technologies were applied to evaluate performance and agronomic traits in pulse breeding programs using data acquired over multiple seasons, locations, and crop types. Unmanned aircraft system-based remote sensing system was employed to acquire image data of chickpea and pea advanced yield trials from three locations during 2017 to 2019. Acquired images were analyzed semi-automatically with custom image processing algorithms and features extracted included canopy area and summary statistics associated with vegetation indices. The study demonstrated significant correlations (P < 0.05) between image-based features (e.g. canopy area and sum normalized difference vegetation index) and yield (r < 0.93 and r < 0.85 for chickpea and pea, respectively) and days to 50% flowering or physiological maturity. Using image-based features as predictors, seed yield was estimated using LASSO regression models, in which the coefficient of determination (R2) was as great as 0.91 and 0.80 during model testing for chickpea and pea, respectively. The study demonstrated it is feasible to monitor agronomic traits of chickpea and pea and estimate pulse yield in multiple locations and seasons using phenomics tools. Phenomics technologies can assist plant breeders to evaluate performance of breeding materials more efficiently and accelerate breeding programs.