Soils are the largest repository of organic carbon in the terrestrial biosphere, and climate warming could result in a substantial loss of this carbon to the atmosphere eliciting a positive feedback to climate. The magnitude of this loss depends, in part, on the temperature sensitivity of the soil microbial community to elevated temperatures. We have been examining soil microbial and nutrient cycling responses to the effects of chronic soil warming for over two decades at the Harvard Forest Long-Term Ecological Research (LTER) site in Petersham, MA, USA where soils have been continuously warmed to 5°C above ambient. Parameters measured include soil respiration, soil C and N stocks, N mineralization, extracellular enzyme activities, and microbial biomass, substrate utilization, growth rates, efficiency, and community structure. This talk will synthesize our current understanding of how the microbial community responds to an altered temperature regime, influencing soil C dynamics.
Long-term multifactor global change experiments are needed because short-term studies do not always anticipate longer term change (i.e., early results are not the story). As an example from our warming experiments, early results (<5 yr) demonstrated that soil respiration is significantly stimulated by warming; however, this effect diminishes with longer term (8-10 yr) warming, with elevated soil respiration in chronically warmed soils returning to ambient levels within a few years. Following 25+ years, we now see the surprising result that soil warming elicits a four-phase pattern of soil organic matter decay and carbon dioxide fluxes to the atmosphere and a substantial loss of soil C. We have also found that results are dependent on the measurement period. Many warming studies focus on the growing season; however, by measuring soil respiration year round over multiple years, we observed that use of growing season data overestimated daily respiration rates by as much as 10-40%; whereas season-specific predictions were generally within 2% of observed values. Thus failure to use season-specific data to model changes in temperature dependencies may over- or under-estimate soil C losses. Long-term warming has also demonstrated that treatment effects are often overshadowed by seasonal and interannual variability, particularly over the short-term (5-10 yr), requiring long-term measurements to discern treatment response trends. Finally, long-term measurements have demonstrated that global change factors mediate biotic interactions that feedback to influence ecosystem responses and have allowed us to evaluate press-pulse interactions and the effects of random, unexpected events.