Skip to main content
ARS Home » Plains Area » Manhattan, Kansas » Center for Grain and Animal Health Research » Stored Product Insect and Engineering Research » Research » Research Project #429013

Research Project: Developing Automated Solutions for Maize Haploid Classification

Location: Stored Product Insect and Engineering Research

Project Number: 3020-43440-009-05-S
Project Type: Non-Assistance Cooperative Agreement

Start Date: Sep 1, 2016
End Date: Sep 30, 2020

The goal is to evaluate the potential of single-kernel near infrared (skNIR) spectroscopy to sort haploid seeds in sweet corn and pre-breeding germplasm. In field corn, haploid kernels have lower oil content and lighter seed weight relative to diploid kernels. Other kernel composition traits predicted by skNIR may also be different in haploids.

The University of Florida completed preliminary studies that suggest skNIR will be valuable for sorting haploid from diploid kernels. The Iowa State Doubled Haploid Facility provided haploid and diploid kernels from 28 female donors crossed by a common inducer line. A general Linear Discriminant Analysis (LDA) model was developed to predict ploidy class from the kernel composition traits estimated by skNIR. The model was able to substantially enrich the haploid seed class so that haploids made up 78% of the selected kernel pool. This represents a significant enrichment of haploid kernels over the starting frequency of 15% in the test population. The LDA model placed the strongest weight on the relative oil content, but kernel density, volume, and starch also contributed to haploid identification. These compositional changes suggest a haploid “syndrome” with a multi-factor chemical signature was detected in the NIR profile. However, skNIR classification was not uniform across all induction crosses. The NIR-selected haploid class ranged from 100% to 23% haploid kernels, depending upon the cross suggesting some female by inducer combinations are more suitable for NIR sorting. We propose to expand the evaluation of NIR haploid classification through this joint-funded collaborative agreement.