Computational model linking soil microbiome composition with environmental and drainage properties
University of Georgia researchers are developing a computational model that maps spatial relationships between soil microbial communities and environmental factors. The model integrates GPS coordinates, soil physical properties, and metagenomic data to predict microbial distributions across heterogeneous agricultural landscapes and assess how tile drainage shapes subsurface microbiomes.
Overview
Soil microbial communities drive geochemical nutrient cycling, yet their composition and function vary spatially with environmental conditions such as soil texture, moisture, and subsurface drainage. Researchers at the University of Georgia are building a computational model that links identified microbial taxa and functional genes with soil physical properties and geographic coordinates. By overlaying microbial data with sample site locations and in-field tile drainage maps, the model can predict microbial community composition at unsampled points across a heterogeneous landscape. This approach enables more precise field management by accounting for the microbial dimension of nutrient cycling that traditional soil testing overlooks.
Technical specifications
The model is built on a spatially resolved sampling design covering two seasonal campaigns:
- Spring sampling: 50 randomly stratified sites across the field
- Fall sampling: 50 sites positioned around drainage tiles to capture drainage-driven microbial effects
- Depth profile: Each site sampled at three depths (0–5 cm, 45–50 cm, and 95–100 cm) to capture vertical community shifts
The methodology is grounded in a validated precedent: a legacy study successfully used functional gene and population data from sequenced groundwater microbiomes to predict whether contamination originated from uranium or nitrogen sources across sites with high geochemical diversity. The current project extends this framework to agricultural soils by integrating:
- GPS coordinates for every sample point
- Environmental and soil physical property data
- Tile drainage network geometry
- Metagenomic profiles from 16S rRNA or whole-metagenome shotgun sequencing
The resulting model relates microbial community composition and function to spatial environmental gradients, enabling prediction of microbial distributions between sampled locations and assessment of how drainage infrastructure reshapes subsurface communities.
Technology readiness level
The project is in an active data collection and partnership-building phase. Soil samples have already been collected from both seasonal campaigns. The research team is seeking a partner to provide metagenomics profiling through 16S rRNA or whole-metagenome shotgun sequencing. Once sequencing data is integrated with the existing GPS and environmental datasets, the team will analyze spatial patterns, compare drainage-adjacent and stratified sites, and build the predictive computational model. The approach is validated in principle by prior published work linking functional microbiomes to geochemical contamination sources, and the current effort represents its adaptation and extension to agricultural soil systems.
About University of Georgia
The University of Georgia is a comprehensive public land‑grant research university serving a large student body across multiple campuses, known for applied scholarship and community partnership. Industry partners access core facilities and pilot‑scale capabilities—including the Food Product Innovation and Commercialization Center—for prototyping, scale‑up, and product validation. A downtown Innovation District and a statewide Cooperative Extension network link campus expertise to companies across Georgia, while proximity to Atlanta’s corporate and logistics hubs lowers barriers to engagement. Research is supported by competitive federal funding from agencies such as NIH, NSF, USDA, and DOE. A dedicated technology transfer office provides IP services, licensing, startup support, and incubator space to accelerate commercialization.
delivered to your inbox.