Carbon Cycling in a Glacier Ice Meltwater Impacted Freshwater System of the Canadian High Arctic: Biogeochemical Processes of the Lake Hazen Watershed

Accelerated glacier ice mass loss and intensified meltwater fluxes are among principal vectors of change affecting northern latitudes in a warming climate. Aquatic systems impacted by glacier ice meltwater are vulnerable to hydrologic shifts in flow conditions that may render consequences to sources, and biogeochemical processes, controlling carbon cycling. Whereas the Canadian high Arctic is the third most extensively glacier covered region on Earth, it is relatively understudied with respect to carbon chemistry in glaciers, and meltwaters traversing proglacial freshwater systems (Chapter 2). To fill this knowledge gap, intensive multi-year (2015-19) sampling expeditions of the Lake Hazen watershed were completed, spanning various meltwater flow conditions. Lake Hazen is Canada’s most northern large lake, located within the Quttinirpaaq National Park, on Northern Ellesmere Island, Nunavut, and is hydrologically dominated by seasonal pulses of glacier ice meltwaters that rapidly navigate the proglacial environment along glacial rivers. Overarching research objectives were to assess carbon chemistry in glacial runoff, whether proglacial river continua were simple pipelines for organic and inorganic carbon pools, and the sources and sinks of carbon in Lake Hazen. \n \nDissolved organic carbon (DOC) has a heterogeneous chemical composition in glacial headwaters that is a confluence of supraglacial meltwater sources (Chapter 3). However, abundant polycondensed aromatic “black carbon” molecules identified in a glacier snow sample were only sparingly detected in glacial headwaters, suggesting that this organic material is removed in the supraglacial environment either via adsorption to sinking particles (i.e., cryoconite), or long-term processing. DOC in glacial headwaters is 14C-ancient, often pre-dating the most recent deglaciation (~ 5 ka BP), which provided compelling evidence that older sources of organic carbon must also be present. In fact, the δ13C and Δ14C of extremely low concentration DOC in glacial headwaters suggests that geogenic sedimentary rock organic carbon is an unprecedentedly important component of DOC. Downriver additions of DOC are increasingly humic-like, and shift towards more negative δ13C and Δ14C, which are in agreement with an OC source derived from terrigenous glacial till of the Wisconsin glaciation (< 75 ka BP). Downriver increases in 3H concentrations suggest that mixing of meltwaters with loose sediment prompt the exfiltration of 14C-ancient DOC from the riverbed. In fact, higher 3H and more negative Δ14C-DOC during high flow conditions support that the lateral expansion of the hyporheic zone releases DOC in ground ice thaw water to glacial rivers that is concentrated in 3H from the warm post-bomb period (1960s). \n \nBy contrast, dissolved inorganic carbon (DIC) is controlled principally by carbonation weathering reactions of carbonate minerals along glacial river continua, but also the extent of atmospheric exchange (Chapter 4). Increased suspended sediment (TSS, PC), undersaturated CO2, and isotopically negative δ13C-DIC and Δ14C-DIC are correlated across the dataset, and are typically associated with higher flow conditions. Greater availability of fine-grain particle surfaces for contact with meltwaters (i.e., suspended in solution, or in the riverbed/hyporheic zone) result in an increased prevalence of carbonate mineral weathering and CO2 drawdown, and liberation of isotopically (δ13C, Δ14C) negative DIC to glacial rivers. Despite additions of DIC along glacial rivers, fast-flowing meltwaters during high flow conditions do not fully equilibrate with atmospheric CO2, and remain isotopically negative along river transects. In fact, during higher flow conditions, the measured δ13C-DIC was much more negative than theoretical δ13C-DIC calculated assuming equilibrium conditions. The δ13C-PIC of suspended sediment is found to be more negative than conventional carbonates, suggesting that secondary carbonates are present, and/or that isotopic exchange mineral weathering reactions affects the δ13C-PIC. Whereas there is isotopic evidence of SO42- derived from the weathering of pyrite based on sulphate isotopes (δ34S-SO4, δ18O-SO4), sulphide oxidation coupled to carbonate dissolution (SOC-CD) is only a minor source of DIC. Indeed, undersaturation in CO2 (and fully saturated O2), downriver increases in pH and decreases in sulphate mass fraction, low SO42-, and alkalinity: DIC < 1 all pointed towards carbonation of carbonate minerals being the primary mineral weathering reaction and source of 14C-ancient DIC to glacial rivers. \n \nLegacy effects of 14C-ancient DOC and DIC in glacial rivers are pervasive in Lake Hazen. Flow-weighted calculations of the seven major glacial river discharges reveal similar quantity and isotopic (Δ14C) character of DOC to that in Lake Hazen. Slightly less 14C-ancient DOC in Lake Hazen is likely a result of in-lake mixing over many years (i.e., long lake residence time), but perhaps also small non-glacial inflows of comparatively 14C-modern DOC, mineralization processes, autochthonous production, and/or adsorption of 14C-ancient OC to sinking particles. By contrast, in-lake DIC has higher concentrations, and less negative isotopes (δ13C, Δ14C) compared to flow-weighted calculations of glacial river discharges. Carbonation weathering reactions of carbonate minerals persist over long periods of time in Lake Hazen, and an ingress of 14C-modern atmospheric CO2 only partially compensates for weathering induced CO2 drawdown. The legacy of 14C-ancient DOC and DIC will consequently be reflected in the Lake Hazen discharge (Ruggles River), and exported to coastal margins and marine environments. \n \nOne of the techniques used to assess dissolved organic matter (DOM) composition in this thesis was fluorescence spectroscopy and excitation emission matrix spectra (EEMS). Fluorescence of DOM using EEMS has been popularized over the last several decades as a fast, inexpensive, and non-destructive technique to characterize DOM in aquatic systems. Peak picking, derivation of fluorescence indices, and parallel factor analysis (PARAFAC) modelling are commonly applied to EEMS data to describe DOM fluorescence composition. However, iron (Fe) and pH influence DOM fluorescence and pose risks for misinterpretation of important DOM sources and biogeochemical processes when comparing freshwaters with different chemistries. As part of an ongoing collaborative project (Mead et al. unpublished), Fe (II), Fe (III), and pH titration experiments were performed on freshwater samples from diverse hydrological settings with inherent variability in DOM composition. In Chapter 5, the effects of Fe (II), Fe (III), and pH were assessed specifically based on their effects on DOM fluorescence EEMS, peaks, indices, and PARAFAC components. Non-uniform changes in DOM fluorescence intensity and spectral features were found that could not have been predicted a priori from concentrations of Fe or pH within the same sample or between samples. In particular, propagation of static fluorescence quenching, inner-filtering effects (IFEs), and co-precipitation of Fe-DOM complicated interpretations of changes of DOM fluorescence characteristics during Fe titration experiments. Effects on DOM fluorescence in freshwaters with Fe <0.5 mg L-1 and pH ~5 to 9 were found to be minimal, implying that comparison of DOM fluorescence data in aquatic systems with water chemistries within these ranges should yield robust characterization of fluorescing DOM.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.