Quantifying temporal patterns of viral mortality in soil

Atmospheric CO2 levels are affected by sources (industrial processes, respiration) and sinks (carbon fixation, storage of organic carbon). Viruses that infect microorganisms may play an important role in soil carbon sink properties. In soil, carbon is stored, or sequestered, by binding to minerals.

Recent studies show that long-term mineral-associated matter is mostly microbial debris, and its fraction out of soil organic carbon varies by soil type.

Hence, viruses that break microbial host cells create microbial debris and contribute to carbon sequestration in soil, yet are missing from carbon flux models.

We lack fundamental information about the physiology of soil viruses to link viral lysis of microbes and soil carbon sequestration.

Recently, we provided the first estimate of microbial mortality by viruses during soil re-wetting after the dry season, a peak time of carbon flux. Our study estimated mortality of up to 46% of cells within a week.

The main knowledge gaps preventing better accuracy of mortality estimates are the average number of viruses produced per host cell and infection rates year-round. I propose to address these gaps using a combination of classic microbiology and cutting-edge bioinformatics.

This project will: (1) generate physiological constraints for soil viral mortality across soil types with different amounts of microbial debris, and (2) establish a multi-year monthly time-series of viral infection dynamics of soil microorganisms.

I will focus on soil viruses infecting Actinobacteriota, an abundant phylum of drought-resilient soil bacteria that experiences consistently high viral infection rates.

A time-series of viral activity will place the culture-based physiological parameters within the context of spatiotemporal dynamics of the entire viral community.

This project will create an unprecedented ability to estimate the potential contribution of viruses to soil carbon sequestration and incorporate it into global carbon flux models.