Microbes are major drivers of carbon and nutrient fluxes in Earth's terrestrial ecosystems; however, Earth system models used for climate change adaptation and mitigation strategies typically exclude explicit representation of soil microorganisms.
A team of researchers from Lawrence Livermore National Laboratory (LLNL), Lawrence Berkeley National Laboratory (LBL) and University of California, Berkeley developed a model that uses the information encoded in soil microbes' DNA to predict how they function and use carbon; ultimately they hope this tool will advance the accuracy of climate models.
The team's new observations provide a basis for improving how interactions between plant roots and soil microbes (bacteria and fungi) are represented in models, and points out traits and tradeoffs that soil bacteria must contend with, notably for organisms that cannot be grown in the laboratory. This new knowledge will enhance scientists' ability to predict how microbial processes affect the global carbon cycle in climate models.
"This new insight has implications for retaining more plant root-derived carbon in soils, and highlights the power of data-driven, trait-based approaches for improving microbial representation in biogeochemical models," said LLNL scientist Jennifer Pett-Ridge, a co-author of the paper appearing in Nature Microbiology.
One of the team's key results is that inherently slower-growing microorganisms are specifically favored when certain types of root carbon are released during late stages of plant development; these organisms have higher carbon-use efficiency (yield) without sacrificing growth rate (power).
The new model incorporates genetic information from soil microbes in novel ways. This enabled the scientists to better understand how certain soil microbes efficiently store carbon supplied by plant roots, and could inform agricultural strategies to preserve carbon in the soil, support healthy plant growth, and climate change mitigation.
Seeing the unseen: microbial impact on soil health and carbon
Soil microbes help plants access soil nutrients and resist drought, disease and pests. Their impacts on the carbon cycle are particularly important to represent in climate models because microbes are key arbiters of whether organic matter becomes stored in soil versus decomposed and released to the atmosphere as carbon dioxide. By building their own bodies from carbon that is shared by plant roots, microbes can stabilize it in the soil, and influence how much and how long carbon remains stored belowground. The relevance of these functions to agriculture and climate are critical issues because of the excess of CO2 in the atmosphere, and the large amount of soil carbon that has been lost due to tillage and erosion.
However, one gram of soil containing up to 10 billion microorganisms and thousands of different species and most microbes have never been studied in the lab. Until recently, the data that scientists had to inform these models came from a tiny minority of lab-studied microbes, many of which are unrelated to soil microbes that researchers need to represent in models.
The world beneath our feet
To address this challenge, the team used genome information to build a model that can be specific to any ecosystem they need to study, from California's grasslands to thawing permafrost in the Arctic. In their test case published in a Nature Microbiology article, the team used genomes from a California rangeland to provide insights into how soil microbes function, and applied this approach to study plant-microbiome interactions. Rangelands are economically and ecologically important in California, making up more than 40% of the land area.
This research is particularly focused on the microbes living near plant roots - a region called the 'rhizosphere.' This root zone, despite being only 1-2% of Earth's soil volume, is estimated to store up to 30-40% of Earth's soil carbon; much of the carbon is released by roots as they grow.
While the team's model originally focused on microbes growing in the root environment, the approach is not limited to a particular ecosystem. Microbial genetic information corresponds to specific traits, just as in humans, and this new model focuses on these traits as a way to reduce the vast complexity of genomic information. Microbial traits also are generalizable enough to be transferable to microbes and ecosystems all over the world.
The team developed a new way to predict important traits of microbes affecting how quickly they use carbon and nutrients supplied by plant roots. Using the model, the researchers demonstrated that as plants grow and release carbon and nutrients, distinct microbial growth strategies emerge because of the interaction between root chemistry and microbial traits. In particular, they found that microbes with a slower growth rate were favored by types of carbon released during later stages of plant development and were surprisingly efficient in using carbon - allowing them to store more of this key element in the soil.
This work was funded by the U.S. Department of Energy (DOE), Office of Science, Office of Biological and Environmental Research.
--LBL's Julie Bobyock contributed to this report.
