Abstract

Modern agriculture’s reliance on chemical fertilizers has greatly increased crop yields but has also contributed to soil degradation, environmental pollution, and unsustainable resource use. Plant growth–promoting rhizobacteria offer a sustainable alternative, enhancing plant growth through mechanisms such as nitrogen fixation, phosphate solubilization, siderophore production, and phytohormone modulation, while also protecting plants against pathogens and abiotic stresses. However, soil microbial communities are highly diverse and structurally complex, making the isolation of individual beneficial strains challenging. Traditional approaches rely on selective media and labor-intensive phenotypic screening, which often capture only a small fraction of soil diversity. To overcome these limitations, we combined computational predictions with targeted microbial isolation to identify bacteria associated with sorghum growth under low- and full-nitrogen conditions. Using a large-scale field dataset, we integrated 16S rRNA amplicon sequencing, soil properties, and plant phenotype data to perform change-point analysis, revealing bacterial strains correlated with plant traits in the rhizosphere and endosphere. Targeted limiting-dilution culturing of selected field samples yielded multiple candidates, including two Pseudomonas strains, AOC87 and AOC36, predicted to have positive and negative associations with plant performance, respectively. Genome sequencing and functional assays confirmed these predictions: the “positive” strain possessed a complete indole-3-acetic acid (IAA) degradation pathway, while AOC36 carried a type III secretion system. These results demonstrate that computational prediction pipelines can refine the search for functionally relevant microbes, bridging culture-based methods with high-throughput microbiome analysis and enabling mechanistic investigation. Building on these findings, Chapter Four explores the complexities of studying and applying candidate growth-promoting strains, specifically Variovorax, to sorghum. This chapter shows that Variovorax strains exhibit both conserved and variable traits: IAA degradation is broadly conserved, supported by syntenic IAA-degradation operons, yet the kinetics of degradation differ quantitatively across strains. Despite this conserved metabolic capacity, effects on sorghum root elongation were highly variable and did not directly correlate with IAA degradation rates. Instead, variability likely arises from strain-specific factors such as root colonization efficiency, persistence, and interactions with host hormonal pathways, including auxin–ethylene crosstalk. Collectively, these results highlight that while IAA degradation is a core feature of Variovorax, its influence on plant growth is context-dependent and shaped by ecological and physiological interactions with the host. Overall, this thesis establishes a data-driven framework that integrates microbial community analysis, spatial modeling, and strain-level characterization, enabling the discovery and functional testing of field-relevant microbes that impact crop performance.

Committee Chair

Rebecca Bart

Committee Members

Barbara Kunkel; Gautam Dantas; Ivan Baxter; Joshua Van Dyke-Blodgett

Degree

Doctor of Philosophy (PhD)

Author's Department

Biology & Biomedical Sciences (Plant & Microbial Biosciences)

Author's School

Graduate School of Arts and Sciences

Document Type

Dissertation

Date of Award

12-10-2025

Language

English (en)

Author's ORCID

0000-0001-7948-5925

Included in

Biology Commons

Share

COinS