Drying-rewetting events induce enormous dynamics in soil biogeochemistry. Two different responses upon rewetting have been identified; bacteria either start growing immediately upon rewetting, coinciding with respiration rates that peak immediately and then decrease exponentially (“Type 1”), or bacterial growth starts increasing exponentially after a lag period, with a sustained period of elevated respiration, sometimes followed by a further increase in sync with bacterial growth (“Type 2”). Previous studies have suggested that the response upon rewetting could be related to the harshness of the disturbance as experienced by the microbes, with more “harsh” (i.e. longer or more severe) drying resulting in a Type 2 response. However, this has yet to be tested in intact ecosystems.
We investigated the legacy of drought on microbial responses to drying and rewetting using grassland soils from a natural precipitation gradient in Texas. Mean annual precipitation spanned a 500 mm range with constant mean annual temperature. Soil properties (pH, SOM) did not vary systematically across the gradient. Air dried soils from 18 sites were rewetted, with bacterial growth, fungal growth and respiration measured at high temporal resolution over 7 days. The moisture dependence of microbes was also measured in all soils.
Results/Conclusions
We predicted that soils with less previous exposure to drying-rewetting events (historically wetter soils) would have a Type 2 response. However, all soils exhibited a Type 1 response, with respiration, bacterial and fungal growth increasing immediately upon rewetting and typically stabilizing after c. 20 hours. This suggested that microbial communities in the historically drier soils did not perceive exposure to drying-rewetting as “harsh”, which could not be explained by a shift in the drought tolerance of microbes across the gradient. However, the amount of carbon used for growth compared to respiration during a drying-rewetting perturbation increased in historically dried soils, suggesting that an enhanced C-use efficiency had been selected for.
Given the lack of systematic variation in other soil properties with rainfall, we expected levels of microbial activity to be similar across the gradient. Instead, we observed marked differences in the magnitude of CO2 release and microbial growth among soils, whereby rewetting of historically wetter soils stimulated higher rates of microbial growth and a greater release of CO2, compared to rewetting of historically drier soils. This might be explained by differences in resource availability (i.e. plant input) across the gradient, and thus an indirect effect of drought on microbial communities.