Warming effects on the structure of bacterial and fungal communities in diverse soils

Abstract Soil temperature affects the rate of C-cycle processes by influencing the activities of microbial communities but little is known about whether the effects on these communities are consistent in different soils. We studied bacterial and fungal communities in six different soils, originating from two different Canadian climatic regions and incubated for three years at a common field site under ambient or warming (ca. +4.6 °C for 653 d of warming) conditions. Similar to early responses (reported after 295 d of warming), soil respiration is consistently higher (up to 2.8-fold) for 653 d of warming. However, soil origin-related differences in the community structure of bacteria were smaller at the end of the experiment, suggesting that environmental factors and management practices are important in shaping the structure of communities, but that this is a cumulative effect over time. In contrast, warming-induced shifts were greater and consistent across soils after longer period of soil warming. These shifts coincided with significant increases in the relative proportions of some potentially copiotrophic bacteria (e.g., Thermoleophilia, Alphaproteobacteria, and Bacilli). Fungal responses to warming, determined at the end of incubation, were largely soil-specific. The relative proportions of some fungi (e.g., Nectriaceae) increased with warming while several other taxa (e.g., Mortierellaceae and Lasiosphaeriaceae) showed significant reductions with warming. Unlike community structure, differences in the DNA content, microbial biomass carbon (MBC), and abundance of bacteria or fungi between the control and warmed soils were minor. This might be related to depletion of readily available substrates because measurements were made a year after the annual addition of litter. Our observation that community shifts became more pronounced over time suggests that these differences may have been realized through physiological temperature optima and through resulting shifts in resource availability.