Soils store around three times as much carbon as the atmosphere. Carbon added to soils during decomposition, and retained there, moderates global climate change. The balance between soil carbon storage and mineralization is mediated by microorganisms that transform plant litter and exudates. A relationship between carbon fate in soils and microbial community composition has been frequently postulated, but disentangling the impacts of community composition from co-varying abiotic factors and litter quality is challenging. Furthermore, community traits driving changes in carbon fate are currently unknown. We aim to identify microbial traits that consistently predict patterns of carbon flow (storage versus mineralization) from decomposition of different litter types. The same natural communities extracted from one hundred soil samples collected from the western United States were placed on pine, oak and grass litter in microcosms. The microcosms were incubated in homogeneous conditions. After 45 days, we compared cumulative carbon dioxide and dissolved organic carbon (DOC) outputs. High throughput sequencing enabled the comparison of community traits.
Results/Conclusions
We found carbon flux patterns vary substantially when decomposition is undertaken by different naturally occurring microbial communities, irrespective of substrate, and despite consistent conditions within the microcosms. Although grass is generally considered more labile than pine or oak, grass decomposition produced less CO2 on average than the other substrates, while DOC quantities were similar across substrates. Litter type impacts community performance, with the same communities generating different carbon quantities on the three substrates. Bacteria, not fungi, were the strongest drivers of divergent functional outcomes on pine and grass with both composition and diversity being important. Determination of community traits driving carbon fate across the different litter types is ongoing, but already our work demonstrates that microbial composition drives differences in carbon flow. Furthermore, communities generating highly divergent carbon flux patterns in microcosms came from geographically intermingled soil samples, suggesting wide functional distributions relating to the carbon cycle may occur within single ecosystems in nature. Identifying universal microbial community traits across variable litter types and environments will enable incorporation of microbe-driven processes into ecosystem models. This work also suggests that soil microbial communities could be an important management tool for controlling climate change.