Response of soil bacterial and fungal communities to summer drought and subsequent rainfall

Rewetting of dry Mediterranean grasslands triggers a flush of carbon substrates, fueling a large soil CO2 pulse, which constitutes an important component of the annual carbon cycle in these ecosystems. However, little is known about the dynamics of activity and resource allocation of the soil microbial community over the dry summer period, which likely sets the stage for the rapid response upon rewetting. In three California grasslands, soil prokaryotic and fungal communities were assessed (by DNA- and RNA-based sequencing) several times over a summer to track changes in the soil microbial community characteristics. In a companion greenhouse-based study, soil from a California grassland was subjected to three different Spring-summer dry-down treatments over four months: weekly water inputs, weekly water inputs for two months followed by drought, and no water input. In both experiments, the present (DNA-based) and potentially active (RNA-based) soil bacterial and fungal communities were assessed over time by sequencing, and the abundance of selected genes determined by qPCR analysis. At the end of summer, soil CO2 efflux rates were determined during a controlled wet-up and the soil microbial community was also analyzed post-wet-up. In soil samples from the field, we found an overall increase in bacterial 16S DNA and fungal 28S DNA gene copies (but not of rRNA) over the summer. At each site, the composition of the RNA-based bacterial community changed significantly as summer drought progressed, then returned to pre-drought composition within several hours of rewetting. Upon rewetting, bacterial mRNA transcript copies significantly increased at all sites, reflecting rapid resumption of activity. In the Spring dry-down experiment, we found significantly more bacterial 16S DNA and fungal 28S DNA gene copies in the dry treatment than in the weekly-watered soil treatment. Upon rewetting, bacterial mRNA transcript copies increased dramatically in both treatments that experienced drought, but not in the weekly-watered soil treatment. In the dry treatment, a larger soil CO2 wetup pulse was associated with a significant decrease of the number bacterial 16S DNA gene copies upon wetup. Our results highlight the resilience of soil bacterial communities in California grassland soils, and suggest that the extent of drying matters more than the temporal pattern of dry-down history for rainfall-stimulated bacterial RNA production, while the dry-down pattern is reflected in the soil CO2 pulse upon rewetting.