Date of Award

12-2025

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Plant and Environmental Science

Committee Chair/Advisor

Dr. Christopher Saski

Committee Member

Dr. Jeffery Adelberg

Committee Member

Dr. Hudson Smith

Abstract

Fusarium oxysporum race 4 (FOV4) is a fungal pathogen that infects cotton (Gossypium spp.) plants, resulting in fusarium wilt (FW) disease that causes significant wilting, yield loss, and eventually death. Since cotton is such an important crop globally, with tens of millions of bales of fiber produced and sold worldwide each year, it is critical to develop effective strategies to combat this pathogen. This research aims to achieve that goal through two projects: (i) develop an efficient co-culture system to test fungal treatments in vitro, and (ii) develop a computer vision pipeline to more accurately assess plant resistance through quantifying fusarium wilt induced stem staining.

The goal of these projects is to provide breeders and researchers with tools that enable faster, easier development of cultivars and treatments to combat Fusarium wilt and protect global cotton production. These efforts will significantly advance FOV4 management in cotton and lay a foundation for future work in this area. Ultimately, however, the co-culture system did not distinguish between resistant and susceptible genotypes, as both produced similarly diseased seedlings. This is likely because the liquid culture conditions were overly favorable for FOV4, generating such high fungal pressure that even the resistant genotype was overwhelmed. Despite this limitation, the system remains a valuable in vitro platform for other applications, such as salinity or nutrient response assays. In contrast, our computer vision (CV) model performed exceptionally well. It achieved an accuracy of 0.998, a mean validation Dice coefficient of 0.82, and an F1 score of 0.96, with a mean absolute error (MAE) of 1.56 and a total runtime of one hour for 274 test images. These metrics indicate that the model is highly accurate, precise, and high throughput compared with manual, visual phenotyping.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.