Date of Award

Summer 8-15-2017

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Evolution, Ecology & Population Biology)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Microbes are bountiful and associated with every animal and plant kingdom. Furthermore, microbes can alter host phenotype, development, health and functioning. However, this is not a one-way interaction, hosts can structure microbial communities by changing the environment to be suitable for certain microbial species. Several studies have characterized microbial communities associated with hosts to answer two2 main questions in ecology: who’s there, and what are they doing? However, two questions from the field of community ecology are often ignored (1) what forces are structuring the microbial communities (how was the community formed) and (2) how stable are these communities. Vellend synthesized that all communities are governed by four main processes: drift, selection, speciation and dispersal. These processes can be grouped into 2 components of assembly, either deterministic (selection, speciation, dispersal) or stochastic (drift, dispersal limitation). The goal of my thesis was to (1) understand the relative contribution strengths of these processes on microbial communities and (2) how stable is the assemblage of microbial community over time and during an infection.

In order to determine if microbial communities are structured deterministically or stochastically, I studied the root endophytic microbiome, which has been shown to directly impact plant physiology. By analyzing 252 root endophytic bacterial (REB) communities, which had been perturbed using antibiotics and sterilization, I show the communities are assembled deterministically. The strongest selective force structuring the REB communities was plant identity even in a perturbed state. I demonstrate show the interplay between REB communities and plant phenotype by linking the variation in the reduction of biomass in autoclaved soils to changes in the abundance of bacterial species. This suggests hosts can selectively increase or decrease the abundance of bacterial species that will increase the plant’s fitness. Consequently, this allows plants to co-exist by specializing on different bacterial species.

To determine the stability of microbial community structure, I studied the urine microbiome of individuals who are do not have urinary symptoms and those who are suspected to have a healthy and who have a suspected Urinary Tract Infection (UTI). By analyzing the urine microbiome of 220 urine samples, microbiome, I show that the urine microbiome is in an altered state during an infection and is stable over time in asymptomatic women for healthy patients. Healthy individuals are enriched with Lactobacillus crispatus and L. iners while individuals with suspected UTIs are enriched with Ruminococcus torques, Propionibacterium acnes and Escherichia coli. There is a plethora of putative pathogens uncovered only with non-conventional culturing methods. Roughly 21% of individuals with suspected UTIs did not have the putative cultured pathogen at high relative abundance but a different known UTI pathogen . This suggests that UTIs could be caused by a dysbiosis of the urine microbiome rather than direct inoculation of an organism from the gastrointestinal tract. an invasion from a different body site.

Collectively these studies show that microbial communities can be structured by the host and host state, and are deterministically assembled. Further work to investigate how the host can structure the microbial community possibly through changing environmental conditions or through immune response is warranted.


English (en)

Chair and Committee

Gautam Dantas

Committee Members

Carey-Ann Burnham, Justin Fay, Andrew Kau, Barbara Kunkel,


Permanent URL: