Developing scalable breeding tools for nixtamalization end-product quality

Abstract

Nixtamalization is a high-heat, high-pH cooking process that loosens pericarp and softens the endosperm of maize grain and is used to create a dough called masa. Several globally significant foods such as tortillas and tortilla chips are made of masa and make up significant portions of consumed calories and nutrients around the world. Relatively few acres of maize are grown for masa-based products in the United States compared to other uses of maize, leading to reduced resources for germplasm improvement in traits that impact masa-product quality. Two traits that have notable downstream impacts on masa quality include moisture absorption and pericarp retention, which can impact the taste, texture, appearance, and machinability of the final product. This thesis addresses the resource gap by developing scalable tools and biological knowledge to improve these traits. Specifically, machine learning models were created to predict nixtamalization moisture content in both inbred and hybrid maize based on the near-infrared spectra of raw maize kernels. These models were used to assess the relationship that nixtamalization moisture content has with kernel composition, the genetic architecture of nixtamalization moisture content, and develop breeding strategies to improve nixtamalization moisture content. This thesis also investigated the compositional and morphological characteristics of maize kernels that impact nixtamalization pericarp retention, providing key foundational knowledge for improving pericarp retention. Together, these studies provide a framework for improving masa-based product quality, reducing waste, and enhancing the sustainability of food-grade maize production in breeding, sourcing, and manufacturing systems.