Semester

Fall

Date of Graduation

2025

Document Type

Thesis

Degree Type

MS

College

Davis College of Agriculture, Natural Resources and Design

Department

Division of Plant and Soil Sciences

Committee Chair

Louis McDonald

Committee Member

James Thompson

Committee Member

Lewis Jett

Committee Member

Ember Morrissey

Committee Member

Daniel Panaccione

Abstract

Quantification of macronutrients in food requires a large investment in facilities and instrumentation as well as ongoing expenses of labor, reagents, and hazardous waste disposal. These costs pose a significant barrier to determining macronutrient quantities in the thousands of uncharacterized landrace and heirloom crops and discourage traditional crop breeding projects from ensuring that their selected crops maintain desirable macronutrient profiles in addition to target agronomic traits. Our objective was to test whether a hand-held Raman spectrometer—a technology which has found use in plant disease diagnostics but remains largely overlooked in nutrient quantification-- could provide a feasible means for quantifying the macronutrients of protein, starch, and fiber in some of the 480 Appalachian heirloom beans. Eight heirloom beans were submitted for laboratory analysis for protein, starch, and neutral detergent fiber (NDFom) by standard methods. These eight samples were expanded into a sample set of 99 by spiking them with varying quantities of starch, cellulose, and hemicellulose or combining them in 1:1 and 2:1 ratios before mixing in a ball mill. Raman spectra were collected and data underwent iterative dimensionality reduction to remove irrelevant and noisy data. Partial least squares regression (PLSR) with external validation was performed on the data after dimensionality reduction. R2pred for protein and starch exceed 0.95, and R2pred exceeded 0.9 for NDFom. Relative RMSEpred for protein, starch, and NDFom were 0.73%, 1.9%, and 7.5% respectively. Our approach produced results equivalent to other techniques and has promise for significantly reducing labor, instrumentation, reagents, facilities, and time required to determine macronutrient content. Furthermore, it potentially exceeds the accuracy of NIR—water does not interfere with Raman scattering. By enabling rapid, low-cost macronutrient quantification in the field, Raman spectroscopy allows researchers to efficiently evaluate macronutrient content heirloom and landrace legume varieties.

Download

Share

COinS