Description
The Arctic is rapidly changing as a result of climate forcing induced by rising greenhouse gas concentrations. Given the large soil organic carbon pools of this biome, understanding how potential changes to arctic ecosystems will affect the Arctic's net carbon balance is imperative for improving predictions of future global climate. However, complicating this understanding is the large heterogeneity of arctic landscapes. There is currently the need for more wholeecosystem studies not only to improve regional flux estimates but also to determine how smallscale variability integrates to form the whole-ecosystem response to climate-induced changes in tundra conditions. This dissertation addresses this research need with a combination of regionalscale observations and whole-ecosystem moisture experimentation in the arctic coastal tundra near Barrow, Alaska. Regional-scale variability in cumulative growing season net CO_ exchange (NEE) was very large and was strongly tied to declining productivity associated with ecosystem development across the dominant landscape unit: thaw lakes and an age sequence of drained thaw lake basins. Contrary to many previous small-scale studies, moisture (aside from lakes) was not a dominant factor controlling regional-scale variability in NEE. However, this result was supported by a study of whole-ecosystem NEE and ecosystem respiration (ER) in a large-scale moisture manipulation experiment. This study confirmed what the few previous large-scale experiments have found: that increased wetness does not necessarily reduce ER and increase carbon storage. Furthermore, the release to the atmosphere of respired CO_ in moist and wet conditions was strongly enhanced by increased wind speed. This effect was shown to be largely missed by small-scale chamber measurements and is currently inadequately considered in commonly used models to partition ER from NEE determined by eddy covariance. Landscapescale variability in CH_ emissions was also large, but was mostly controlled by ecosystem moisture status and had very little relation to ecosystem development or productivity, identifying contrasting patterns and controls on fluxes of CO_ and CH_. The large control of CH_ flux variability by soil moisture was confirmed by the large-scale moisture manipulation experiment in which an experimentally raised water table resulted in higher CH_ emission. This experiment identified further control of moisture on autumn CH_ emissions, linking the decline in autumn CH_ emissions to the decline in liquid moisture during soil freezing. A higher water table slowed the soil freezing process, prolonging higher CH_ emissions later into the autumn and early winter. Combined, these results indicate that the variability in CO_ and CH_ emissions is large but can be explained and predicted in order to improve and validate regional flux models. Taken together, these results suggest that increased soil moisture in arctic areas may increase both CO_ and CH_ emissions, while increased lake drainage could turn strong CO_ source areas into large CO_ sinks as vegetation develops and soil organic matter accumulates
