M.Sc. & Ph.D. Theses

Gold mining in northern Canada, while profitable, has left a legacy of arsenic (As) contamination, especially in lakes. Arcellinida, lacustrine benthic protozoans, have great As-biomonitoring potential that has yet to be exploited by lake restoration initiatives. This research aims to further develop Arcellinida as a robust tool for monitoring As contamination and remediation efficacy in impacted lakes. In Chapter 2 a novel Arcellinida sample processing protocol, using chemical deflocculation, was developed to determine optimal approaches for reducing residual organic matter in lake sediments with minimal Arcellinida test degradation. The results revealed that treatment with 5% potassium hydroxide led to the highest reduction in observed organic content and rate of test damage. In Chapter 3 results of a study where Arcellinida were examined from 22 near-surface sediment samples (top 1 cm) collected from similar environments in two eastern Canadian lakes was carried out to evaluate faunal assemblages homogeneity in these targeted environments, and to assess the sensitivity of Arcellinida assemblages to interannual environmental variability. The results of this study confirmed the faunal homogeneity of assemblages in limnologically similar environments and demonstrated the sensitivity of Arcellinida assemblages to environmental changes at inter-annual time scales. In Chapter 4 an inter-lake survey of 91 lakes in the Yellowknife area, Northwest Territories (NT), Canada, was conducted to quantify the response of arcellinidan to gold mining-derived As contamination and to statistically determine As-tolerance limits of different Arcellinida species. The results demonstrate the applicability of using Arcellinida to delineate the spatial extent of As contamination and the reliability of using As tolerance limits based Arcellinida groups to infer As contamination levels in lake sediments. In Chapter 5 the arcellinidan response to intra-lake changes in As contamination, derived from two former gold mines, was investigated in 40 near-surface sediment samples
from two lakes in the low Arctic, central NT (Tundra Mine) and in a boreal setting, southern NT (Giant Mine). It was demonstrated that Arcellinida assemblages responded differently to the impact of As, likely due to different As mineralization in each lake, which was in turn a by-product of employing different ore-processing techniques at the former gold mines.

Metal(loid)s are expected to respond to 21st century warming as their stability in lacustrine systems is indirectly influenced by regional temperature and precipitation. In the Northwest Territories (NT), arsenic (As) is a metal(loid) of environmental concern due to elevated concentrations in bedrock and widespread mining-related contamination. To characterize the response of As to long- and short-term climate variability, we developed techniques that enable high temporal resolution analyses of sediment freeze cores, and applied them to a sediment cores (CON01, CON02) recovered from Control Lake, NT. New equipment was designed to enable Itrax X-ray fluorescence (Itrax-XRF) analysis of discrete sediment samples (Chapter 2). This equipment was used to test calibration methods that convert semi-quantitative Itrax-XRF results to near-total geochemical concentrations (Chapter 3). We found that the multi-variate log-ratio calibration provided the best approximation of actual geochemical concentrations. Subsequently, equipment was designed to permit scanning of freeze-cores using Itrax-XRF (Chapter 4). To characterize millennial-scale shifts in sedimentary As concentration in response to climate change, Arcellinida and ICP-MS analysis were conducted on core CON01 that recorded 4000 yr of sedimentation (Chapter 5). Arsenic concentrations the in core were related to shifts in the proportion of organic matter and shifts in minerogenic content (Rb, K). Comparison to regional records suggests changes in temperature impacted autochtonous productivity, which is hypothesized to have influenced sedimentary As concentrations. To characterize the response of As to quasi-periodic climate oscillations, CON02 was analyzed using Itrax-XRF. Itrax-XRF data were calibrated to paired ICP-MS data, and geochemical proxies for particle size (log(Zr/Ti)), in-lake productivity (log(Ca/Ti)), and As preservation (log(As/Ti)) were examined for cyclities using spectral and wavelet analyses. Periods of 8-15, 30-60, 90-130, and 180-300 yr were observed in all proxies. These periods are temporally related to the North Atlantic Oscillation/El niño/Schwabe sunspot cycles, the Pacific Decadal Oscillation, the Gleissberg cycle, and the Suess cycle, respectively. Cross-wavelet analysis of the paleo-proxies vs. published total solar irradiance reconstructions demonstrate significant relationships, suggesting a solar influence on climate and lake sediment geochemistry in the NT. These results suggest that As sequestration is impacted by short-period climate perturbations.

Harvey Lake, located within the Atlantic region of Canada, endures many storms, including Hurricane Arthur in 2014, and likely archives lake sediment records including past storm signals. Lakes are often characterized by a single central site, rather than a suite of sites, due to cost and time constraints. This research sought to develop a novel geomatics protocol to optimize coring site selection using a multi-site (n = 96) characterization. Areas prone to resuspension were modeled using lake morphology and historical wind speed records (1953-2015). Modeled resuspension areas agreed with the spatial distribution of sedimentary proxies (i.e., grain size and Itrax-XRF). End member mixing analysis identified a very coarse grain end member that likely reflects the deposition of resuspended sediments. In addition to the central basin, our approach highlighted Herbert’s Cove as a suitable coring target as it was in closer proximity to a source of allochthonous sedimentation.

This thesis is composed of manuscripts investigating the impact of climate events throughout the past ~2800 years on conditions within the lacustrine environment of Pocket Lake as reflected in changes of diatom assemblage. The first manuscript presented in chapter 2 is composed of a broad, paleoclimatic reconstruction at a resolution of ~20-30 years per sample based on changes in diatom assemblage integrated with geochemical and particle size datasets from the same core. In the second chapter a high-resolution investigation into the impact of the deposition of airfall tephra into Pocket Lake is presented. Together, these studies provide insight into the impact of broader decadal scale climate cycles and instantaneous, episodic events allowing for a more comprehensive understanding of their impact on lakes in sensitive, northern locations. These reconstructions will help inform predications regarding the potential impact of 21st century climate change.

Arcellinida are an established group of bioindicators in lake studies, but conventional labor-intensive microscopic analysis can limit the number of samples analysed. In this study, the FlowCam with VisualSpreadsheet (FCVS), a flow cytometer and microscope with machine learning software, was assessed as an instrument for rapid Arcellinida analysis. In a 2016 study, conventional Arcellinida analysis was performed on 46 samples collected from Wightman Cove, Oromocto Lake, New Brunswick, Canada. The samples were reanalyzed by FCVS and the two datasets were compared at the morphotype level using cluster and Bray-Curtis dissimilarity analyses. The methods produced highly similar arcellinidan assemblages that corresponded to specific lake habitats. FCVS was found to reduce analytical time by approximately 45%. FCVS shows potential as a reliable method for more rapid analysis of lacustrine Arcellinida; however, assemblage results can only be obtained at the morphotype level. Microscopic methods should still be used if species-level results are desired.

Frame Lake, a small (88.4 ha), shallow (< 6.5m), high-latitude lake found within the city limits of Yellowknife, Northwest Territories, Canada was selected to test the feasibility of using seismic sub-bottom profiling to estimate total volumes of heavy metal contaminated sediments in lacustrine environments. To ground-truth the sub-bottom profiling results, physical and ICP-MS analyses were carried out on freeze cores collected from Frame Lake’s southern basin, and sedimentological marker beds and 14C dating was used to chronologically constrain the lake depositional history. ICP-MS results showed high levels of arsenic contamination (up to 1538 μg g-1) in late 20th- century lake sediments, which contrasts sharply with measured Holocene values that averaged only 16 μg g-1 (n=41, ± 5.4 SD). The high arsenic content in lakebed sediments, which tends to be concentrated within specific horizons, results in distinct seismic reflectors within the acquired Sonar data. Stratigraphic horizons where arsenic was concentrated do not necessarily correlate with actual depositional events as changes in lake hydrology and redox conditions have resulted in remobilization and migration of arsenic in lake sediments. Direct GIS software comparison of core data against the sub- bottom profiler transect results permitted an interpolated lateral and vertical reconstruction of the distribution of variously contaminated sediments throughout the entire lake basin. Based on our analysis, a minimum of ~230,000 m3 of contaminated sediments would need to be dredged from Frame Lake to achieve a minimum residual sediment arsenic concentration of < 150 μg g-1.

This project analyzed morphological diversity in the modern gull (Aves: Charadriiformes) genus Larus. Measurements of the post-cranial skeletons were made on specimens from six species of Larus and analyzed using Principal Component Analysis (PCA). Similar analyses were conducted on a series of theropod taxa. Landmark-based geometric morphometric analysis was conducted on skulls of gulls and analyzed using PCA. Results show that the gull species are very similar for the morphological parameters analyzed, only differing in overall body size and cannot be easily distinguished based on their skeletal morphology. These results were used as a modern analogue for inferring species diversity in the theropod fossil record. The implications of this study suggest that skeletal morphology may not be able to distinguish closely related species of non-avian theropods. This suggests that the species diversity of extinct non-avian theropods was likely higher than can be of determined from their skeletal remains.

The central Northwest Territories of the Canadian Subarctic, and the myriad of lakes found therein, has through previous research been shown to archive records of past Holocene climate variability. Due to the low temporal resolution of previous paleoclimatic reconstructions questions remain as to past rates of natural climate change, a deficiency of concern to policy makers and modelers responsible with addressing human induced climate change. Two paleolimnological proxies of climate variability, end-member mixing analysis (EMMA) of lake grain-size distributions and the Arcellinida were further developed to increase the temporal resolution of paleoclimate reconstructions and provide a more complete picture of environmental response to past climate variability. A mm-interval subsampling strategy made possible through use of a custom-designed sledge microtome combined with EMMA of lake sediment cores significantly increased the temporal resolution of paleoclimate reconstructions. EMMA of lake sediment cores was used to describe changes in lake catchment hydrology and track the rate of climate deterioration and subsequent amelioration associated with known Holocene events such as the Holocene Climate Optimum (HCO) and the onset of the Neoglacial. Arcellinida as demonstrated here, are sensitive to treeline dynamics. Due to their well-preserved dissolution-resistant tests, rapid generation time and trophic position within lacustrine environments they offer unique insight into climate variability and change. Arcellinida are also characterized by ecophenotypes, which respond markedly to changing environmental conditions making them valuable tools for reconstructing environmental change. Three ecological transitions were seen in the observed Arcellinida communities through the past ca. 6600 cal yBP. Zone 1 spanned ca. 6600-2800 cal yBP and characterized the hydroecology of the lake during the HCO. Zone 2 spanned ca. 2800-140 cal yBP, which represents the Neoglacial. Zone 3 spans ca. 140 cal yBP to present and archives the recovery of the region from the Little Ice Age. Further development of these proxies will help to constrain the rate of climate variability and increase our understanding of the many ways in which the environment responds to climate change. This has implications not only for the Canadian Subarctic but for the Taiga biome, the largest terrestrial biome in the world.

The Athabasca oil sands (AOS) represent one of Canada’s most economically important natural resources. Understanding the origin, fate and long-term impact of these industrial operations requires geoscientific knowledge of the potential environmental impacts. To quantify current impact an understanding of long-term environmental drivers is required. Here we utilize arcellaceans (testate lobose amoebae) to identify the peleoecologic history of the area, against which ecological stress associated with anthropogenic activities may be defined. Surface sediment samples were collected from 63 lakes forming a radius encompassing potential airborne and waterborne contaminate pathways. Total mercury (THg) in lakes close to the AOS were similar to values recorded in lakes farthest away and to values recorded in lakes surrounding coal fire plants in central Alberta. All lakes except one had THg concentrations lower than current CCME guidelines. The spatial extent of arcellacean ecosystem stress did not correlate with THg concentrations. To understand the long-term history of the AOS area a core (ALE) spanning 1,870 cal yr BP – present was used to assess how climate anomalies influence lake hydrology and ecology. Deconvolution of the grain size spectra indicates that populations are likely responding to overarching environmental controls. The coarser bedload material was interpreted to reflect high discharge events that occur during spring melts. A ratio of the coarse:fine fractions demonstrates abrupt oscillations of both high and low frequency and indicates that variation in spring melt volume and velocity was likely forced by climate change. Further investigation of the grain size data showed that the hydrologic character of ALE has responded to known ocean-atmosphere cycles, particularly the El Niño Southern Oscillation (ENSO) and that these cycles are underwent abrupt expression shifts every ca. 270-370 years. We attribute the expression shift to movement of the jet stream, which controls air mass distribution at this locality. An investigation of the ecological character of ALE from 1875 AD – present reveals that the arcellacean community was also influenced by climate anomalies and that decreased nutrient input, which is a function of precipitation during the El Niño phase of ENSO, applied stress to benthic ecology in these boreal lakes.

Arcellaceans (testate lobose amoebae) were examined for 61 sediment surface samples from lakes in the vicinity of the Giant Mine near Yellowknife, Northwest Territories to; (1) quantify the impact of the mine on the Baker Creek Watershed region, (2) determine the utility of arcellaceans as indicators of arsenic and heavy metal contamination and gauge the success of remediation efforts. Several statistical methods, including cluster analysis, Deterended Correspondence Analysis (DCA), and Redundancy Analysis (RDA), were used to quantify the impact of mining activity on the arcellacean assemblages. Cluster analysis revealed five arcellacean assemblages associated with a range of environmental conditions (e.g. polluted, transitional and remediated). Partial RDA results confirm that arsenic has the greatest influence on the arcellacean distribution, explaining 10.7% of the total variance. Stress-indicating species (e.g. Centropyxids) correlate with high arsenic concentrations, while species characteristic of more healthy lake conditions (e.g. Difflugids) dominate sites with significantly lower arsenic concentrations.

This thesis is comprised of two manuscripts that focus on diatom ecological change through a late Holocene record from a 116.2 cm freeze core obtained from Danny’s Lake, Northwest Territories, Canada. The diatom results indicate that climate in this region has been relatively stable through the past 3330 cal. yr BP, although three distinct diatom assemblages are recognized. Time-series analysis was also carried out on select diatom species from the Danny’s lake sediment core. We correlate the c. 89 and c. 145 year cycles with the 90 – 140 year Gleissberg cycle, while the c. 309-year cycle is attributed to the 300-year overtone of the 2115-year Hallstadt cycle. This research is part of a multi-proxy project mandated to determine late Holocene climate variability along the route of the economically important Tibbitt to Contwoyto Winter Road (TCWR), a seasonal ice road that stretches 600 km from Yellowknife to Nunavut.

We examined the spatial and temporal variability of Holocene lake sediment accumulation at 22 sites from 18 lakes transecting boreal forest, tree line, and tundra zones in the central Northwest Territories, Canada. Over 140 radiocarbon dates were obtained, and accumulation rates (AR) were calculated at 100-year intervals from age- depth models constructed using the age-depth modeling software Clam. Sites with the shortest mean AR of 25±10yr/cm (1ó) occur primarily in the boreal zone. Sites with moderate (70±22yr/cm) and long (160±56yr/cm) AR are north of the treeline and display higher variability, strongly influenced by bathymetry. Many age-depth models are characterized by fluctuations in ARs that coincide with paleogeographical changes associated with proglacial lake evolution during the early Holocene, and subsequent climate changes inferred from proxy data. The insights gained on the spatial and temporal trends in ARs across the region are valuable for developing higher resolution age-depth models using the Bayesian software Bacon.