2020 ESA Annual Meeting (August 3 - 6)

OOS 21 Abstract - Modeling soil microbiome responses to drought

Steven Allison, Ecology and Evolutionary Biology, University of California, Irvine, Irvine, CA; Earth System Science, University of California, Irvine, Irvine, CA and Bin Wang, Ecology and Evolutionary Biology, University of California Irvine, Irvine, CA
Background/Question/Methods

As climate changes, drought frequency and intensity are expected to increase across many areas of the planet. Yet the response of soil microbial communities, or microbiomes, to drought remains uncertain. Because soil microbiomes influence key ecosystem processes such as decomposition and nutrient cycling, ecosystem responses to drought are likewise uncertain. To address these uncertainties, we incorporated desiccation resistance mechanisms into DEMENT, a trait-based and spatially-explicit model of soil microbiomes. We aimed to test how drought affected modeled rates of decomposition, and whether changes in decomposition rates persisted following drought due to legacy effects. We also tested the sensitivity of model predictions to variation in dispersal rates and costs of desiccation resistance associated with osmolyte production. Microbiomes in the model assembled under ambient conditions in Southern California grassland for 10 years before being subjected to three years of simulated drought. Changes in community structure and litter decomposition rates were simulated during and after the drought period.

Results/Conclusions

Under ambient climate conditions, microbiome composition and litter decomposition rates stabilized after 3-4 years. Drought perturbation resulted in lower decomposition rates and re-assembly of the simulated microbiome with shifts toward taxa that invested more in osmolyte production. Following drought, we observed a legacy of reduced decomposition rates and altered microbiome composition that persisted for multiple years. Legacy effects were greater with more intense simulated drought and with greater costs of osmolyte production. However, legacy effects were much smaller and shorter-lived if dispersal rates were increased in the model. Higher rates of dispersal allowed extinct taxa to recolonize litter following drought perturbation. These results demonstrate the utility of DEMENT in simulating processes of environmental selection and dispersal that can occur in field experiments. The simulations also reveal a potential for ecosystem impacts to persist beyond the initial period of drought. Such legacy effects have been observed in the field and may affect soil carbon balance under global climate change.