Links between soil microbial communities, functioning, and plant nutrition under altered rainfall in Australian grassland

The size, frequency, and timing of precipitation events are predicted to become more variable worldwide. Despite these predictions, the importance of changes in precipitation in driving multiple above- and belowground ecosystem attributes simultaneously remains largely underexplored. Here, we carried out 3 yr of rainfall manipulations at the DRI-Grass facility, located in a mesic grassland in eastern Australia. Treatments were implemented through automated water reapplication and included +50% and -50% amount, reduced frequency of events, and an extreme summer drought. We evaluated the spatiotemporal responses of multiple ecosystem attributes including microbial biomass, community composition and activity, soil nutrient content and availability, and plant nutritional status to altered rainfall regimes. We found that changing precipitation patterns resulted in multiple direct and indirect changes in microbial communities and soil and plant nutrient content. Main results included greater availability of soil macronutrients and reduced availability of micronutrients under drought, and taxon-specific changes in the composition of soil microbial communities in response to altered rainfall. Moreover, using structural equation modeling, we showed that, in summer 2015, plant macronutrient contents, a widely used ecological indicator of pasture quality, were simultaneously explained by greater soil nutrient availability and the structure of soil microbial communities, and significantly reduced by lower rainfall. Plant micronutrients were also reduced by lower rainfall and explained by changes in microbial attributes. Despite treatment effects on many of the soil, microbial, and plant variables analyzed across the 3 yr of study, many of these ecosystem attributes varied greatly across sampling events. This resulted in many significant interactions between the rainfall treatments and experimental duration, suggesting complex system-level responses to changing rainfall in our grassland, and a high natural buffering capacity of the ecosystem to varying rainfall conditions. Some interactions manifested as changes in the coefficient of variation of ecosystem attributes, particularly in response to changes in the timing of precipitation events and the extreme summer drought. Finally, we posit that a detailed understanding of plant-soil-microbial interactions, and the role of climate in modifying these linkages, will be key for adapting the sustainability of grasslands to a future that will be shaped by climate change.