Biotic and abiotic controls on soil organic carbon quality along a paired pine and hardwood climosequence

Despite recent advances in our understanding of the specific mechanisms leading to soil organic carbon (SOC) formation and stabilization, the heterogeneity of SOC (from easily decomposable to recalcitrant) limits our capacity to make predictions on the responses of SOC to changing temperature. Differences in decomposition rate and permanence in soils (mean residence time - MRT) are indicators of the quality of SOC, where SOC quality decreases with decreasing decomposition rates and increasing MRT. This dissertation investigates SOC quality across a 22°C mean annual temperature (MAT) gradient in temperate forests in North America to address the following questions: (1) Is SOC quality in temperate forests related to MAT? (2) Does tree species composition (conifer versus hardwood) result in different SOC quality? (3) Is recalcitrant SOC less sensitive to temperature than labile SOC? (4) Can we extend the application of current SOC models to forested mineral wetlands at temperate ecosystems?; A combination of long-term (525-d) incubations, SOC fractionation, acid-hydrolysis, and radiocarbon analysis showed that both quality and quantity of SOC decrease with increasing MAT. When data were fit to a 3-pool kinetics model, labile (high quality) SOC was on average 2.1 +/- 0.2% (Mean +/- S.E.) of total SOC and both labile SOC concentration and its MRT (33 +/- 6 d) were negatively related to MAT. Conversely, both concentration (27 +/- 1%) and MRT (3485 +/- 879 yr) of recalcitrant (low quality) SOC were not related to MAT. Tree species composition affected SOC quality across sites: more total and recalcitrant SOC accumulates in hardwood than pine. Forested mineral soil wetlands accumulated significantly more SOC than upland forests from similar locations, but low temperature responses suggest that this stock of SOC may not be rapidly depleted when wetlands are exposed to warmer or dryer conditions. Temperature sensitivity (Q10) was similar between forested mineral soil wetlands and upland forests when soils were incubated under aerobic conditions and Q10 stabilized at an average value of 1.5, lower than currently modeled. The high temperature responses of SOC commonly hypothesized for wetlands worldwide may overestimate the possible C flux from forested mineral soil wetlands, when oxygen is not limiting. This study suggests that, in a scenario of climate change, greater SOC losses will occur from soils at cold environments (both upland and forested mineral wetland soils) due to greater accumulation of labile SOC that is highly sensitive to temperature. Relatively greater proportion of recalcitrant SOC accumulates at warm sites due to overall greater C inputs to soil. The effects of increasing global temperature on temperate forests in North America will be lower than currently hypothesized and SOC stabilization may be enhanced in upland sites where hardwoods replace pine.