Global climate change and ecosystem carbon storage: An experimental investigation of ecologically-mediated feedbacks to climate in montane meadows

Climatic change can alter rates of terrestrial carbon assimilation and release by plants and soils, resulting in feedback which could either enhance or retard the rapid anthropogenic buildup of atmospheric CO2. The carbon-cycle response of terrestrial systems to climatic changes, however, is poorly understood, and the literature lacks studies distinguishing transient and long-term feedback responses and which also include biogeochemical effects of such responses.; My dissertation presents results from a 7-year meadow warming experiment, as well as from a companion study along a landscape-scale natural climate gradient. Together, these results elucidate both short-term and long-term controls on ecosystem carbon storage. Measurements of soil carbon pool sizes in near-surface soil revealed a substantial ecosystem carbon loss (300--600 g C m-2) from soils in an experimentally warmed meadow. Net ecosystem and soil respiration CO2 exchange measurements (including wintertime fluxes though the snowpack), litter harvests, vegetation censusing, and controlled laboratory soil incubations indicate that reduced litter inputs (due to a warming-induced shift in vegetation community composition)---not increased soil respiration---caused most of the carbon loss. The plausibility of the observed rapid loss of large amounts of carbon from soil is confirmed by a simple model of soil carbon dynamics constrained by observed levels of soil carbon in young and old fractions (as determined by carbon-14 measurements).; Species-specific differences in litter chemistry, together with analysis of patterns of plant community composition and soil carbon along a landscape-scale natural climate gradient imply, however, that soil carbon will eventually recover because lower quality litter inputs will reduce litter and soil decomposition rates.; Shifts in montane plant species composition with warming are therefore expected to exert strong short-term positive carbon-cycle feedbacks that will be ameliorated in the long term. Although this ecological mechanism will not, even on a global scale, affect expected equilibrium levels of anthropogenically-induced warming, it could substantially accelerate the near-term rate of warming. Because many of the anticipated negative consequences of climate change for human populations are dependent on the rate of change---not just its absolute amoun---this finding has important implications for the debate about policies to adopt for ameliorating anthropogenically-induced climate change.