Thesis (Ph. D.)--University of Washington, 2003. / Vita. Includes bibliographical references (leaves 125-140).

Thesis (PhD)--Montana State University--Bozeman, 2008. / Typescript. Chairperson, Graduate Committee: Brian L. McGlynn. Includes bibliographical references.

Thesis (Ph. D.)--College of William and Mary. / Vita. Includes bibliographical references. Also available online.

Reliable prediction of climate change and its impact on and feedbacks from terrestrial carbon cycles requires realistic representation of physiological and ecological processes in coupled climate-carbon models. This is hampered by various deficiencies in model structures and parameters. The goal of my study is to improve model realism by incorporating latest advances of fundamental eco-physiological processes and further to use such improved models to investigate climate-carbon interactions at regional to global scales. I focus on the CO₂ diffusion within leaves (a key plant physiological process) and large-scale disturbances (a fundamental ecological process) as extremely important but not yet in current models. The CO₂ diffusion within plant leaves is characterized by mesophyll conductance (g[subscript m]), which strongly influences photosynthesis. I developed a g[subscript m] model by synthesizing new advances in plant-physiological studies and incorporated this model into the Community Land Model (CLM), a state-of-art climate-carbon model. I updated associated photosynthetic parameters based on a large dataset of leaf gas exchange measurements. Major findings are: (1) omission of g[subscript m] underestimates the maximum carboxylation rate and distorts its relationships with other parameters, leading to an incomplete understanding of leaf-level photosynthesis machinery; (2) proper representation of g[subscript m] is necessary for climate-carbon models to realistically predict carbon fluxes and their responsiveness to CO₂ fertilization; (3) fine tuning of parameters may compensate for model structural errors in contemporary simulations but introduce large biases in future predictions. Further, I have corrected a numerical deficiency of CLM in its calculation of carbon/water fluxes, which otherwise can bias model simulations. Large-scale disturbances of terrestrial ecosystems strongly affect their carbon sink strength. To provide insights for modeling these processes, I used satellite products to examine the temporal-spatial patterns of greenness after a massive ice storm. I found that the greenness of impacted vegetation recovered rapidly, especially in lightly and severely impacted regions. The slowest rebound occurred over moderately impacted areas. This nonlinear pattern was caused by an integrated effect of natural regrowth and human interventions. My results demonstrate mechanisms by which terrestrial carbon sinks could be significantly affected and help determine how these sinks will behave and so affect future climate. / text

Much research has been devoted to understanding the ocean carbon cycle because of its prominent role in controlling global climate. Coastal oceans remain a source of uncertainty in global ocean carbon budgets due to their individual characteristics and their high spatial and temporal variability. Recent attempts to establish general patterns suggest that temperate and high-latitude coastal oceans act as sinks for atmospheric carbon dioxide (CO2). In this thesis, carbon cycling in two Canadian coastal ocean regions is investigated, and the uptake of atmospheric CO2 is quantified. A combination of ship-board measurements and highly temporally resolved data from an autonomous mooring was used to quantify the seasonal to multi-annual variability in the inorganic carbon system in the Scotian Shelf region of the northwestern Atlantic for the first time. The Scotian Shelf, unlike other shelf seas at similar latitude, acts as a source of CO2 to the atmosphere, with fluxes varying over two orders of magnitude in space and time between 1999 and 2008. The first observations of the inorganic carbon system in the Amundsen Gulf region of the southern Beaufort Sea, covering the full annual cycle, are also presented. Air-sea CO2 fluxes are computed and a carbon budget is balanced. The Amundsen Gulf system acts as a moderate sink for atmospheric CO2; seasonal ice-cover limits winter CO2 uptake despite the continued undersaturation of the surface waters. Biological production precedes the ice break-up, and the growth of under-ice algae constitutes nearly 40% of the annual net community production. The Scotian Shelf may be described as an estuarine system with an outflow of surface water, and intrusion of carbon-rich subsurface water by a combination of wind-driven mixing, upwelling and convection, which fuels the CO2 release to the atmosphere. In contrast, Amundsen Gulf may be described as an anti-estuarine, or downwelling, system, with an inflow of surface waters and an outflow of subsurface waters. Wind-driven and convective mixing are inhibited by ice-cover and restrict the intrusion of carbon- and nutrient-rich waters from below, maintaining the CO2 uptake by the surface waters. / PhD Thesis

In an upland forest in the Experimental Lakes Area (ELA), Ontario, dissolved organic carbon (DOC) concentrations in precipitation increased with passage through a forest canopy, as throughfall and stemflow. A maximum mean concentration of 67 mg/l occurred in the forest floor, which decreased to 11 mg/l in the B horizon, as DOC was sequestered in the soil. High DOC concentrations were measured in an ephemeral stream draining the upland, as a result of saturated overland flow, and the leaching of litter and woody material. / Porewater DOC concentrations decreased with depth in two bogs. The porewater DOC was comprised primarily of acidic fractions (70 to 87%), with a predominance of hydrophobic acids. The bogs were significant sources of DOC, exporting between 17.2 and 29.4 g DOC/m$ sp2$ over the study period. In 1993, a bog was flooded simulating the creation of a hydroelectric reservoir, and resulting in significant increases in porewater DOC concentrations at the surface. A maximum surface concentration of 223 mg/l was measured in late September. The flooding of the bog also resulted in significant increases in DOC concentrations in the draining waters, with the outflow of the watershed increasing from a mean of 20.5 mg/l in 1992, to 14.7 mg/l in 1993.

In order to better quantify the continuous net ecosystem exchange (NEE) at the farm scale, a 24 m high transportable tower-based eddy covariance and profiling system was tested during three field campaigns during the summer and autumn of 2006. The profile measurement system, equipped with a low-cost closed-path infrared gas analyser (LI-840) was designed to measure the carbon dioxide (CO2) storage especially during weak mixing periods under stable conditions. A tethered blimp-based measurement system for the nocturnal boundary layer (NBL) budget technique was deployed on several calm nights to compare with the concentrations and fluxes determined with the profiling system. The profiling system was also installed to determine horizontal and vertical advection. The two-dimensional horizontal positioning of the profile inlets allowed an examination of the CO2 advective gradient in relation to the wind direction. A near-source advection analysis provided parameters for further adjustments of the NEE at the actual farmland.

A hydrological and biogeochemical study was undertaken at the Mer Bleue bog, Ottawa, Ontario, Canada from May 22, 1998 to May 21, 1999. Basin runoff was generated by groundwater discharge at the peatland margin, and groundwater discharge was controlled by hydraulic gradients and horizontal hydraulic conductivities (Kh). Flux of dissolved organic carbon (DOC) measured at the basin outflow was 8.3 g C m-2 yr-1 and compared to within 23% of DOC flux estimated using a Dupuit approximation of seepage during the ice-free season. Annual DOC flux was 11% of the annual carbon sink. / Flownet analysis showed that seasonal patterns of groundwater flow were controlled by boundary condition changes that resulted from precipitation and evapotranspiration events. A pattern of recharge was most common over the hydrological year, but a discharge pattern was observed during a 40 day groundwater flow reversal. Evaluation of the peatland recharge-discharge function using in situ sodium concentrations and a diffusion model revealed that the peatland is a long-term recharge system. It is hypothesized that peatland biogeochemical function is controlled by long-term recharge despite annual occurrence of groundwater flow reversals.

Iron is recognized as a key element regulating primary production in large regions of the ocean, but nothing is known of its direct effect on higher trophic levels. Growth and metabolism of two species of heterotrophic protozoans fed iron-rich and iron-poor prey were thus examined. Maximum growth rates of Paraphysomonas imperforata and P. butcheri were observed only when Fe quotas of bacterial prey were greater than 70 $ mu$mol Fe:mol C. At lower Fe:C ratios, but at constant prey biomass (C/ml), both species grew significantly slower. Minimum Fe quotas of the flagellates at these slow growth rates ($ sim$10 $ mu$mol Fe:mol C) were similar to those of iron-limited phytoplankton and bacteria. Growth rate reduction was the result of direct elemental limitation by Fe, judging from the protozoans' positive response to Fe additions and from their biochemical characteristics. Filtration and carbon ingestion rates increased under Fe-limitation, but carbon gross growth efficiency (CGGE) decreased when Paraphysomonas imperforata consumed iron-poor bacteria. Ammonium regeneration efficiency was also reduced. The decrease in CGGE was a consequence of reduced activity of the iron-dependent electron transport system, greater DOC excretion, and greater CO$ sb2$ evolution by Fe-limited flagellates. Paraphysomonas imperforata excreted Fe, even when limited by this element, and retained less of the ingested ration and thus had a higher Fe regeneration efficiency than when consuming Fe-rich bacteria. According to recent measurements of biogenic Fe:C in the subarctic Pacific, our results suggest that heterotrophic bacterivorous flagellates may experience iron-limitation in remote oceanic regions. Such limitation could profoundly affect C, N and Fe cycling in the sea.

This study examines carbon flow in the Dipper Harbour salt marsh, a macrotidal system located on the north-west coast of the Bay of Fundy, New Brunswick. The vegetated marsh surface is composed of three major zones; the Spartina alterniflora-dominated low marsh, the Plantago maritima-dominated Middle marsh, and the Spartina patens-dominated high marsh. The total net primary production (NPP) of these dominant macrophytes is 860, 300 and 650 g C m$ sp{-2}$ yr$ sp{-1}$ respectively. In all plant zones, 66% of the NPP occurs in the belowground fraction. / Empirical measurements of organic matter burial indicate that the marsh sediment acts as a carbon sink, accumulating between 75 and 105 g C m$ sp{-2}$ yr$ sp{-1}$. The tidal export of aboveground plant biomass in the form of particulate organic matter accounts for a net loss of carbon ranging from 65 to 170 g C m$ sp{-2}$ yr$ sp{-1}$. An experiment examining the exchange of dissolved organic carbon (DOC) suggest a net annual export of roughly 500 g C m$ sp{-2}$ yr$ sp{-1}$. The forementioned fluxes are combined with estimates of surface gas exchange and algal productivity in order to construct a carbon budget. The budget predicts a DOC export term of 365 g C m$ sp{-2}$ yr$ sp{-1}$, which is of the same order of magnitude as that obtained from the empirical DOC data. / The results of this study show that the Plantago zone plays a significant role in the circulation of carbon in the Dipper Harbour salt marsh. This suggests that the patterns of carbon circulation in northern marshes may differ considerably from those in marshes further south where no distinct Plantago zone has been reported.