http://hdl.handle.net/10012/18663

Kay, Mitchell (2022) Development and application of paleolimnological analyses to disentangle the roles of natural processes and anthropogenic activities on contaminant deposition and hydrological conditions across a northern delta, UWSpace - Theses, http://hdl.handle.net/10012/18663

Freshwater ecosystems across northern Canada provide important habitat for wildlife and have long supported the traditional lifestyles of Indigenous communities. Multiple potential stressors threaten the security of water supply to northern landscapes, which fosters need for information spanning broad spatial and temporal scales to inform adaptive and mitigative strategies. At the Peace-Athabasca Delta (PAD; northern Alberta), the world's largest boreal freshwater delta, existing data records have been too short and too sparse to resolve many concerns over the roles of major energy projects (hydroelectric regulation of river flow, oil sands development) and climate change on decline of flood frequency and magnitude and drawdown of shallow aquatic basins, and on supply of substances of concern. Intensive paleolimnological research during the past two decades at the PAD has evaluated past changes in contaminant deposition and hydroecological conditions to discern effects attributable to oil sands development along the Lower Athabasca River and to regulation of Peace River flow by the W.A.C. Bennett Dam. This thesis builds substantially on these previous studies to address knowledge gaps by applying conventional paleolimnological methods at new locations to improve understanding of temporal changes in contaminant deposition and hydrological change and developing an innovative paleolimnological approach for discerning variation in sediment sources over space and time to lakes within the PAD. Concerns of pollution in the PAD stem from potential for dispersal of contaminants released by bitumen mining and processing activities within the Alberta Oil Sands Region (AOSR), which straddle the Lower Athabasca River. Unfortunately, systematic monitoring began thirty years after onset of oil sands development and sampling locations have changed over time, which has hampered the ability to accurately track temporal trends or attribute sources of pollution at the AOSR and downstream locations. Previous paleolimnological studies in the PAD have provided critically missing baseline information by employing lake sediment deposited before oil sands development to evaluate lake or river-bottom sediment deposited after oil sands development for evidence of pollution. Results show no enrichment via fluvial or atmospheric pathways, however, analyses to date were limited to a sediment core from one upland lake and samples of river-bottom sediment collected from a few sites within the Lower Athabasca River and its distributaries within the PAD. In Chapters 2 and 3, contiguous measurements of trace elements (beryllium, chromium, lead, mercury, nickel, vanadium, and zinc) in lake sediment from floodplain and upland lakes were employed to develop knowledge of pre-disturbance concentrations, and to quantify the extent of enrichment in lakes at the PAD since onset of oil sands mining and processing activities via fluvial and atmospheric pathways, respectively. Results demonstrate no enrichment since onset of oil sands development via fluvial pathways. Also, no enrichment via atmospheric pathways was detected for vanadium, nickel and total mercury (THg) at upland lakes coincident with the onset of oil sands activities. Total mercury enrichment was detected at the start of the 20th century in sediment cores from two upland lakes, which is congruent with stratigraphic patterns observed in many other lake sediment records in response to long-range anthropogenic emissions across the northern hemisphere. Site-specific paleolimnological studies from four regions spanning large-scale bitumen mining on the Lower Athabasca River to gold mining in central Northwest Territories (including AOSR, PAD, Slave River Delta (SRD), Yellowknife region of central NWT) have provided a wealth of information about temporal patterns of deposition of substances of concern. Differences in laboratory methods and data analysis, however, have challenged ability to compare and contrast site-specific studies among regions. Opportunity to coalesce the current state of knowledge was capitalized on in Chapter 4 via systematic re-analysis of concentrations of key pollution-indicator trace elements in sediment cores from 51 lakes spanning the four key regions. Lake sediment records from lakes within the mining regions (AOSR and central Northwest Territories) illustrate enrichment of pollution-indicators since onset of mining and processing activities via atmospheric pathways, while no enrichment was detected at the PAD or SRD via fluvial pathways since onset of mining activities. The knowledge generated from Chapters 2-4 can be employed by multiple stakeholder groups to assess risks associated with contaminant dispersal across a vast region of northwestern Canada. Long-term perspectives provided by paleohydrological studies at the PAD have demonstrated that decline of flood frequency and magnitude and lake-level drawdown began decades before onset of Peace River flow regulation by the W.A.C. Bennett Dam. Many of these studies have been concentrated in the northern Peace Delta but concerns also exist about declines in river discharge, flood frequency and lake levels in the southern Athabasca Delta. Chapter 5 tests the hypothesis that the Embarras Breakthrough, a natural geomorphic change in distributary flow of the Athabasca River, is the main driver of recent hydrological change in the Athabasca Delta. Stratigraphic variations in the mineral matter content of sediment cores from nine floodplain lakes, including at sites within the Athabasca River terminus region, demonstrate that flood influence increased after 1982 at lakes along a distributary to the north of the Embarras Breakthrough and declined at lakes east of the Embarras Breakthrough. The timing of this bi-directional change confirms that the Embarras Breakthrough has caused the largest shift in hydrological conditions within the Athabasca Delta during the past ~120 years. Paleohydrological reconstructions employing conventional analyses have provided valuable insight into the hydrological evolution of the PAD but integration of the results across sites has remained a challenge due to marked differences in sediment composition across the spectrum of hydrological processes influencing lake water balances. At flood-prone lakes, physical methods (e.g., grain size, magnetic susceptibility) provide high information content, whereas biological or bio-geochemical methods (e.g., diatoms, plant macrofossils, cellulose oxygen isotope composition) provide high information content at perched basins. Elemental concentrations, however, can be determined accurately along the full gradient of mineral-rich to organic-rich sediment of flood-prone and perched basins, respectively, and can be used to delineate the three major sources of sediment supplied to lakes (Athabasca River, Peace River, and local catchment), which is a key advancement over previous paleolimnological studies. In Chapter 6, a mixing model framework was developed and evaluated via application to sediment cores from two adjacent lakes in the Athabasca Delta. Output from the mixing model aligns remarkably well with conventional loss-on-ignition analysis and paleohydrological interpretations from the same two lakes, which further illustrate the profound effect of the 1982 Embarras Breakthrough on hydrological conditions of lakes in Athabasca Delta. Interestingly, model results indicated that ~60% of the sediment originated from the Peace River during the largest ice-jam flood event in the hydrometric record (1974). Due to the success of this model, opportunity exists to apply the model to a network of lakes in the PAD, where elemental concentrations have been analyzed, to reconstruct spatial and temporal variation of pathways of sediment sources and infer changes in hydrological processes. The methods developed and applied in this thesis are anticipated to be broadly applicable to other freshwater landscapes where monitoring records remain too short and too sparse to discern effects of multiple stressors.

https://uwspace.uwaterloo.ca/bitstreams/0899400f-10c4-4aa9-91c8-ebae95ead41f/download