Phinizy Swamp Floodplain

Crocker Lab

Uncovering simplicity in the microbiome

Phinizy Swamp Floodplain

Microbial communities are critically important biological systems, with impacts ranging from human health to the global climate. The rapid improvement of genetic sequencing technology over the past two decades was expected to give rise to a predictive microbiome science to facilitate design and control of these vital systems. However, building a predictive approach robust to the complexity of microbial communities in the wild remains a challenge. Our lab seeks to address this challenge by discovering underlying structure in natural microbial communities. We do this by employing a combination of statistics, machine learning, wet-lab experimentation, and mathematical modeling to bypass species-level complexity and develop a predictive understanding of these vital ecosystems.

NEWS

  • We are recruiting! Please reach out to Kyle if you are interested in joining the lab.
  • August 2026: Kyle is starting at Augusta University in the Physics and Biophysics Department.

Research

Ecological structure in environmental microbiomes

.js-id-feast-famine

Microbial ecology in the gut-brain axis

Human microbiome sequencing studies repeatedly link neurodegenerative disease to gut microbiome composition. Identifying the underlying ecological and metabolic mechanisms is difficult, however, given the complexity of the human microbiome and the near-impossibility of controlled intervention in situ. We therefore partner with the Mor Lab to work in Caenorhabditis elegans, performing experiments on a defined worm microbiome and using them to build statistical and ecological models that (mechanistically!) link microbiome properties to neurodegeneration.

Succession in a leaf

Successional dynamics are ubiquitous in microbial communities but hard to study in host-associated systems, where experimental access is limited. Pitcher plants offer a way in: their modified leaves collect rainwater, prey, and microbes, assembling a characteristic microbiome whose succession is shaped by prey input and host physiology. We therefore partner with the Bittleston Lab to probe fundamental successional processes in the pitcher plant Sarracenia purpurea.

Guild competition in the cow rumen

In situ studies link the ruminal microbiome to an efficiency gradient in dairy cows: higher-yielding animals emit less methane per unit milk. With the Mizrahi Lab, we hypothesize that this reflects competition for hydrogen between propionate-producing and methanogenic guilds. We combine experimental characterization with coarse-grained consumer-resource models to test this and to identify interventions that improve feed conversion efficiency.

Ecology in the latent space

Modern machine learning approaches excel at identifying low-dimensional patterns in high-dimensional datasets. A key challenge, however, is extracting generalizable scientific insight from these patterns. To address this challenge, we build physics-constrained ecological models in a latent space learned from natural microbiome data. We expect this approach to identify simple descriptions of microbial community dynamics that enable prediction across host-associated and environmental systems.

Nutrient-microbiome feedback in wetland flows

Waterways transport nutrients through wetlands, stimulating microbial activity that in turn alters nutrient availability. We partner with the AU Electronics Lab (led by Profs. Genevieve Reeves, Andy Hauger, and David McCall) to autonomously measure environmental and nutrient dynamics in local waterways. We couple these measurements with sequencing and experimental characterization of co-located microbes, allowing us to understand and predict the nutrient-microbiome feedbacks that play a key role in ecosystem health.

Predicting rewetting response

Soil rewetting stimulates nutrient release and triggers rapid metabolism in the soil microbiome, driving a large fraction of global CO2 emissions. Intriguingly, stereotypical successional patterns are observed in microbial activity following rewetting events. In this project, we ask 1) how these low-dimensional patterns emerge from ecological and evolutionary self-organization of microbial communities and 2) whether we can leverage successional structure to predict nutrient flux through natural microbiomes.

Metabolic function in host-associated microbiomes

.js-id-host-microbiome

Microbial ecology in the gut-brain axis

Human microbiome sequencing studies repeatedly link neurodegenerative disease to gut microbiome composition. Identifying the underlying ecological and metabolic mechanisms is difficult, however, given the complexity of the human microbiome and the near-impossibility of controlled intervention in situ. We therefore partner with the Mor Lab to work in Caenorhabditis elegans, performing experiments on a defined worm microbiome and using them to build statistical and ecological models that (mechanistically!) link microbiome properties to neurodegeneration.

Succession in a leaf

Successional dynamics are ubiquitous in microbial communities but hard to study in host-associated systems, where experimental access is limited. Pitcher plants offer a way in: their modified leaves collect rainwater, prey, and microbes, assembling a characteristic microbiome whose succession is shaped by prey input and host physiology. We therefore partner with the Bittleston Lab to probe fundamental successional processes in the pitcher plant Sarracenia purpurea.

Guild competition in the cow rumen

In situ studies link the ruminal microbiome to an efficiency gradient in dairy cows: higher-yielding animals emit less methane per unit milk. With the Mizrahi Lab, we hypothesize that this reflects competition for hydrogen between propionate-producing and methanogenic guilds. We combine experimental characterization with coarse-grained consumer-resource models to test this and to identify interventions that improve feed conversion efficiency.

Ecology in the latent space

Modern machine learning approaches excel at identifying low-dimensional patterns in high-dimensional datasets. A key challenge, however, is extracting generalizable scientific insight from these patterns. To address this challenge, we build physics-constrained ecological models in a latent space learned from natural microbiome data. We expect this approach to identify simple descriptions of microbial community dynamics that enable prediction across host-associated and environmental systems.

Nutrient-microbiome feedback in wetland flows

Waterways transport nutrients through wetlands, stimulating microbial activity that in turn alters nutrient availability. We partner with the AU Electronics Lab (led by Profs. Genevieve Reeves, Andy Hauger, and David McCall) to autonomously measure environmental and nutrient dynamics in local waterways. We couple these measurements with sequencing and experimental characterization of co-located microbes, allowing us to understand and predict the nutrient-microbiome feedbacks that play a key role in ecosystem health.

Predicting rewetting response

Soil rewetting stimulates nutrient release and triggers rapid metabolism in the soil microbiome, driving a large fraction of global CO2 emissions. Intriguingly, stereotypical successional patterns are observed in microbial activity following rewetting events. In this project, we ask 1) how these low-dimensional patterns emerge from ecological and evolutionary self-organization of microbial communities and 2) whether we can leverage successional structure to predict nutrient flux through natural microbiomes.

People

Principal Investigator

Kyle Crocker

Kyle Crocker

Assistant Professor

Department of Physics and Biophysics

College of Science & Mathematics

Augusta University

kycrocker@augusta.edu

Recent Publications

Quickly discover relevant content by filtering publications.

Functional regimes define soil microbiome response to environmental change

DOI

Inferring resource competition in microbial communities from time series

DOI

Timescale of environmental change modulates metabolic guild cohesion in microbial communities

DOI

Environmentally dependent interactions shape patterns in gene content across natural microbiomes

DOI

High-force application by a nanoscale DNA force spectrometer

See all publications