Agricultural landscapes are known to increase phosphorus (P) losses to waterways, contributing to the eutrophication of freshwater surface water bodies. In cold agricultural regions, the nongrowing season drives annual P transport and discharge. Previous research has focused on discharge from fields and watersheds to understand P dynamics in response to hydroclimatic events such as snowmelt and rain storms. Although P supply in soils has been considered a dominant mechanism driving P runoff, the dynamic nature of this supply on an annual basis in response to climate drivers is poorly understood. The goal of this thesis is to determine climatic (seasonal, moisture and temperature) controls on the supply of soluble P in agricultural soils. Two experiments were set up: one in a field setting and the other in a lab setting. The field study involved a snow-manipulation experiment in an agricultural field, in which soil P pools and net transformation rates were quantified under snow and limited-snow conditions. The lab experiment explored the impacts of frost severity, frost duration, frost cycle number and moisture addition on soil concentrations of water-extractable soluble reactive P (SRP), total dissolved P (TDP) and Olsen P, microbial biomass P, aggregate stability, and concentrations of SRP, TDP and TP (total phosphorus) in leachate draining from cores. In both studies, frost magnitude did not significantly impact soil P fractions or supply. Although soil water extractable P (WEP) was greater during the non-growing season than summer, this was not impacted by increased frost due to the removal of snow cover. The lab study also showed that frost magnitude did not impact P supply; however, both frost duration and moisture additions appeared to affect P supply. Water extractable P was positively related to moisture content in both experiments. An improved understanding of climate drivers on P cycling is needed in light of climate change. This thesis suggests that the supply of P may be impacted by a changing climate, but more due to moisture shifts rather than temperature.

Small inland waters (SIWs) – waterbodies smaller than 100 km2 – are the predominant form of lakes globally, yet they are highly subject to water quality degradation, especially due to harmful algae blooms (HABs). Space-borne remote sensing has proven its capability to detect and map HABs in coastal waters as well as large waterbodies mostly through estimating chlorophyll-a (Chla). However, remote retrieval of near-surface Chla concentration in SIWs is challenging due to adjacency effects in remotely sensed signals and substantial in situ optical interferences of various water constituents. Although various algorithms have been developed or adapted to estimate Chla from moderate-resolution terrestrial missions (~ 10 – 60 m), there remains a need for robust algorithms to retrieve Chla in SIWs. Here, we introduce and evaluate new approaches to retrieve Chla in small lakes in a large lake catchment using Sentinel-2 and Landsat-8 imagery. In situ Chla data used in this study originate from various sources with contrasting measurement methods, ranging from field fluorometry to high-performance liquid chromatography (HPLC). Our analysis revealed that in vivo Chla measurements are not consistent with in vitro measurements, especially in high Chla amounts, and should be calibrated before being fed into retrieval models. Calibrated models based on phycocyanin (PC) fluorescence and environmental factors, such as turbidity, significantly decreased Chla retrieval error and increased the range of reconstructed Chla values. The proposed calibration models were then employed to build a consistent dataset of in situ Chla for Buffalo Pound Lake (BPL) – 30 km length and 1 km width – in the Qu’Appelle River drainage basin, Saskatchewan, Canada. Using this dataset for training and test, support vector regression (SVR) models were developed and reliably retrieved Chla in BPL. SVR models outperformed well-known commonly used retrieval models, namely ocean color (OC3), 2band, 3band, normalized difference chlorophyll index (NDCI), and mixture density networks (MDN) when applied on ~200 matchups extracted from atmospherically-corrected Sentinel-2 data. SVR models also performed well when applied to Landsat-8 data and data processed through various atmospheric correction (AC) processors. The proposed models also suggested good transferability over two optical water types (OWTs) found in BPL. Based on prior evaluations of the models’ transferability over OWTs in BPL, locally trained machine-learning (ML) models were extrapolated for regional retrieval of Chla in the Qu’Appelle River drainage basin. The regional approach was trained on in situ Chla data from BPL and retrieved Chla in other six lakes in the drainage basin. The proposed regional approach outperformed a recently developed global approach (MDN) in terms of accuracy, and showed more applicability than local models given the scarcity of in situ data in most lakes. In addition, ML models, e.g., SVR, performed consistently better than other models when employed in the regional approach. A rare phenomenon of marked blue discoloration of ice and water in winter 2021 in Pasqua Lake, a small lake in Qu’Appelle Watershed, provided an opportunity to assess the regional approaches in estimating chlorophyll-a for waterbodies where enough training data is not available. Therefore, using a developed model based on data from BPL, we produced Chla maps and could successfully relate the discoloration event to a late fall bloom in Pasqua Lake. We included the details of that study in Appendix A. Altogether, the models and approaches introduced in this thesis can serve as first steps toward developing a remote-sensing-based early warning system for monitoring HABs in small inland waters. Results showed that the development of an early warning system for SIWs based on Chla monitoring is currently possible, thanks to advancements in medium-resolution satellite sensors, in situ data collection methods, and machine learning algorithms. However, further steps need to be taken to improve the accuracy and reliability of systems: (a) in situ data need to be consistent for being fed into remote sensing models, (b) retrieval models and AC processors should be improved to provide better estimations of Chla, and (c) regional approaches might be developed as alternatives for local and global approaches in the absence of accurate AC processors and scarcity of in situ Chla data.

Freshwater systems and fish communities face many anthropogenic threats such as climate change, pollution, and invasive species, causing a rapid loss of biodiversity. To protect freshwater fish basic knowledge of their numbers, distribution, and habitat is required, but can be difficult to obtain. Remote freshwater systems can experience human impacts but are even less understood due to lack of access and potential hazards. Patagonia (the southern tip of South America) is sparsely populated but contains many freshwater systems that can be indirectly impacted by human activities. The introduction and subsequent naturalization of several invasive salmonid species in rivers and lakes. Despite the potential for adverse effects on these ecosystems, they are generally under researched and uncharacterized. One such freshwater environment is the rivers draining from the stratovolcano Melimoyu in northern Chilean Patagonia, where limited access, high flows and considerable river debris makes traditional sampling from boats or wading impractical or dangerous. New biomonitoring techniques such as environmental DNA (eDNA) detection can be used to sensitively and non-invasively obtain data of species presence or even community structure through analysis of water samples and may provide an avenue for obtaining data about freshwater communities in remote systems such as Melimoyu. Applying eDNA barcoding or metabarcoding techniques in remote systems may allow researchers to gain knowledge about the biota in these environments where traditional sampling may be limited or impossible. Environmental DNA barcoding was used for the detection of three different fish species: invasive brown trout (Salmo trutta) and Atlantic salmon (Salmo salar), and native puye (Galaxias maculatus) in the rivers draining Volcán Melimoyu. Brown trout eDNA was detected at seven sites across four rivers, Atlantic salmon eDNA was not detected at any sample sites, and puye eDNA was detected in one river with high certainty. At several sites eDNA detection techniques were accompanied by backpack electrofishing. The detection of brown trout eDNA was potentially influenced by differing environmental conditions (e.g., flow) between sampling events. Puye was not always detected by eDNA despite being collected during electrofishing. eDNA can be a powerful biomonitoring tool for detection of fish in remote systems, especially if applied with consideration of controls for contamination, transport, and storage of samples. Samples for eDNA barcoding can be collected under less-than-ideal conditions, and with an appropriate sampling regime, applied to remote systems to obtain valuable data on distributions of individual fish species. However, in future studies eDNA metabarcoding (i.e., simultaneous detection of all fish eDNA present using universal primers) may be a more powerful tool for use in remote freshwater environments to gain an understanding of entire communities. Collections of supplementary data could be used to inform occupancy models for a better understanding of the eDNA and presence of fish in these freshwater systems. This study demonstrates the potential application of eDNA for informing resource managers about fisheries resources, even in areas where traditional fisheries techniques are difficult or impractical to complete.

Contemporary Inuit diets are comprised of both country and store-bought foods, which each confer benefits and risks to Inuit physical, mental, cultural, spiritual and socio-economic health. Inuit residing in Inuit Nunangat (the Canadian traditional homelands of the Inuit) disproportionately experience food insecurity and impacts of climate change, threatening the quality and safety of foods consumed. Elevated concentrations of certain environmental contaminants in Inuit Nunangat represent a concerning source of Inuit dietary exposure to contaminants through country food consumption. Further, Inuit are experiencing disconcertingly high rates of chronic diseases, are consuming less nutritious and culturally significant country foods, and are consuming more unhealthy, non-nutrient dense store-bought foods. It is therefore imperative that Inuit communities have access to evidence-informed and culturally relevant information promoting healthy and safe diets to support their nutritional and cultural well-being. Dietary messages addressing the health risks and benefits of country and store-bought food choices and activities in the Inuvialuit Settlement Region (ISR) of the Northwest Territories (NWT) aim to reduce harm and improve health among Inuvialuit (Inuit from the Western Arctic). However, an understanding of how dietary messages are developed and disseminated in the ISR remains unknown and best practices for collaborative approaches to nutrition communication grounded in Inuvialuit culture and knowledge is understudied. This project aims to fill these gaps in knowledge, extending our understanding of dietary message communication strategies in Inuit communities. The purpose of this thesis is to (1) Characterize current public health dietary messages in the ISR (Study 1); (2) Identify how territorial, regional and local dietary message disseminators, local country food knowledge holders, and the public in Tuktoyaktuk can co-develop culture-centered dietary messages to more effectively promote healthy, safe and culturally appropriate diets in the community (Study 2); and (3) Provide recommendations to territorial, regional and local dietary message stakeholders to further improve dietary messaging in the ISR and NWT (Studies 1 and 2). This study utilized an Indigenous research paradigm and community-based participatory and decolonizing research approaches. An in-person interview (n=1) (February 2020) and telephone interviews (n=13) (May-June 2020) were conducted with key informants (health professionals, government employees and community nutrition program coordinators) in Inuvik, Tuktoyaktuk, Paulatuk and Yellowknife (Study 1). An Inuvialuk community researcher conducted storytelling interviews with country food knowledge holders (n=7) and community members (n=3), and a talking circle with local public health dietary message disseminators (n=2) between June-July 2021 in Tuktoyaktuk (Study 2). Follow-up key informant telephone and videoconference interviews with territorial and regional dietary message disseminators (n=5) were completed in June 2021 (Study 2). Interviews were analyzed using thematic analysis. The findings indicated that dietary messages disseminated to the public in the ISR are developed at all scales and communicated by territorial and regional (allied) health professionals, territorial and regional health department representatives, regional and local food program coordinators, academic researchers, country food knowledge holders and local leadership through a variety of in- person, written, audio and online methods. Country food knowledge holders communicate their own messaging through the sharing of Inuvialuit knowledge while harvesting and preparing country food in their communities. Public health dietary messages focus predominantly on a) healthy store-bought food choices, b) nutritional advice about store-bought and country foods and c) safety risks of consuming country foods. Federal and territorial messaging is seldom tailored to the ISR, lacking representation of the Inuvialuit food system and consideration of local food realities. Key barriers to regionally tailored, culture-centered dietary message development and dissemination in the ISR included a lack of collaboration between stakeholders involved in communications and limited resources required to develop trusting, respectful and collaborative relationships between dietary message stakeholders. Participants at all levels support increased inclusion of cultural and community perspectives about food to develop regionally and locally tailored dietary messaging, especially about country food harvesting and preparation knowledge and skills. Although most dietary message stakeholders wish to be involved in co- development processes, some country food knowledge holders desire leading traditional communications about country foods in Tuktoyaktuk. This project has made an important contribution to the literature on health and risk communication about country and store-bought foods in northern Indigenous communities by characterizing dietary messages disseminated in, for and within the ISR, examining residents’ awareness of messages, and identifying best practices for co-developing regionally and locally-tailored, culture- centered dietary messages in the ISR. Findings from this project have informed the creation of the Inuvialuit Food Messages Survey to evaluate the effectiveness of dietary messages as part of the ongoing Country Foods for Good Health project. Findings have also informed recommendations to NWT and ISR dietary message stakeholders to more effectively promote healthy, safe and culturally appropriate diets in Tuktoyaktuk and the ISR through the (co-) development and dissemination of culture-centered dietary messaging that supports Inuvialuit food sovereignty. Additionally, the process of conducting this thesis during the COVID-19 pandemic has led to methodological innovations for working remotely with community researchers and is able to provide key recommendations for researchers that can be used post- pandemic. These findings and recommendations have practical applications for other Inuit Nunangat regions and Canadian northern Indigenous communities interested in understanding and improving dietary messaging communication strategies.

Hydrological monitoring in complex, dynamic northern floodplain landscapes is challenging but increasingly important as these ecosystems come under threat from multiple stressors, including climate-driven decline in freshwater supplied by rivers draining the hydrographic apex of western North America. Sustainable approaches capable of tracking status, trends and drivers of lake water balances in complex, remote landscapes are needed to inform ecosystem stewardship and water-security actions. The Peace‐Athabasca Delta (PAD) in northern Alberta, Canada, is a Ramsar Wetland of International Importance reliant on episodic river ice‐jam flood events to recharge abundant perched lakes and wetlands. However, the frequency of these floods has been in decline for decades over much of its area. Compounding concerns about water-level drawdown have prompted the need to improve knowledge of lake water balances and establish a lake monitoring program. Yet, the delta’s remoteness and dynamic nature present challenges to these goals. In this thesis, I address hydrological knowledge gaps essential to understanding spatial and temporal patterns of hydrological processes and their influence on lakes in the PAD. First, we assess the legacy influence of a large-scale ice-jam flood in 2014 on hydrological and limnological status of lakes in the PAD by integrating spatial and temporal data. Analysis of water isotope compositions and water chemistry measured at numerous lakes across the delta shows that hydro-limnological effects of the flood event of 2014 failed to persist beyond the early ice-free season of 2015. Isotope-inferred paleohydrological records from five hydrologically representative lakes in the PAD indicate that periodic desiccation during the Little Ice Age occurred at the most elevated basin in response to locally arid climatic conditions and reduced flood frequency, yet other lower elevation sites were influenced by high water level on Lake Athabasca owing to increased snowmelt- and glacier-derived river discharge. In contrast, water isotope data during the past 15 years at all five lakes consistently document the strong role of evaporation, a trend which began in the early to mid-20th century according to sediment records and is indicative of widespread aridity unprecedented during the past 400 yr. We suggest that integration of hydrological and limnological approaches over space and time is needed to inform assessment of contemporary lake conditions in large, complex floodplain landscapes. Next, we use water isotope compositions, supplemented by measurements of specific conductivity and field observations, from 68 lakes and 9 river sites in May 2018 to delineate the extent and magnitude of spring ice‐jam induced flooding along the Peace and Athabasca rivers. Lake‐specific estimates of input water isotope composition (δI) using a coupled-isotope tracer approach were modelled after accounting for the influence of evaporative isotopic enrichment. Then, using the distinct isotopic signature of input water sources, we develop a set of binary mixing models and estimate the proportion of input to flooded lakes attributable to river floodwater and precipitation (snow or rain). This approach allowed identification of areas and magnitude of flooding that were not captured by other methods, including direct observations from flyovers, and to demarcate flow pathways in the delta. We demonstrate water isotope tracers as an efficient and effective monitoring tool for delineating spatial extent and magnitude of an important hydrological process and elucidating connectivity in the PAD, an approach that can be readily adopted at other floodplain landscapes. Finally, we use over 1000 measurements of water isotope composition at ∼60 lakes and 9 river sites during the spring, summer and fall of five consecutive years (2015–2019) to identify patterns in lake water balance over time and space, the influential roles of evaporation and river floodwaters, and relations with meteorological conditions and river water levels. Calculation of evaporation-to-inflow ratios using a coupled-isotope tracer approach, displayed via generalized additive models and geospatial ‘isoscapes’, reveal strongly varying lake water balances. Results identify distinct areas vulnerable to lake-level drawdown, given the likelihood of continued decline in ice-jam flood frequency, longer ice-free season duration and reduced snowmelt runoff. Results also demarcate areas of the delta where lakes are more resilient to factors that cause drawdown. The former defines the Peace sector, which is influenced by floodwaters from the Peace River during episodic ice-jam flood events, whereas the latter describes portions of the active floodplain environment of the Athabasca sector which receives more frequent contributions of Athabasca River floodwaters during both spring ice-jam and open-water seasons. Efficiency of water isotope tracers to capture the marked temporal and spatial heterogeneity in lake water balances during this 5-year time span, and their diagnostic responses to key hydrological processes, serves as a foundation for ongoing lake monitoring, an approach readily transferable to other remote and dynamic lake-rich landscapes.

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.

Snow is an essential component in hydrology systems, and more speci cally, monitoring variations in snow cover can provide valuable information about water supply, wildlife habitats, and climate changes. In recent years, the potential of wide band snow radar (e.g. 2-8 GHz) has been discovered with a series of campaigns in Operation IceBrdige (OIB). Due to data from OIB mainly focusing on sea ice, most of the algorithms were also developed for snow on sea ice. As a result, this thesis aimed to test the applicability of the interface-based pulse peakiness snow depth retrieval method to snow on land. In addition, due to the common usage of radar altimeter in sea ice classi cation, this thesis also explored the possibility of adapting some of the ideas in sea ice classi cation to develop another retrieval method. Both approaches were tested on the 6 major vegetation types (tree, tall shrub, riparian shrub, dwarf shrub, tussock, and lichen) in the study area. Snow depth derived from Airborne Laser Scanner (ALS) point clouds was used as the reference for snow depth retrieval. Running the recalibrated pulse peakiness algorithm yielded a Mean Absolute Error (MAE) of 12 cm, 27 cm, 29 cm, 13 cm, 9 cm, and 10 cm for tree, tall shrub, riparian shrub, dwarf shrub, tussock and lichen respectively. It was concluded that the principles behind the pulse peakiness approach is valid for snow on land. The presence of surface vegetation and the hummocky terrain of the study area did present some considerable challenges. In comparison, the classi cation approach using K-means while produced some accurate results under speci c situations, was not as robust as the existing pulse peakiness approach. Although, it was argued that the classi cation approach was in early stages and there were some potential for better results.

Excess phosphorus (P) from agricultural watersheds promotes eutrophication in downstream aquatic systems. Reservoirs retain P generated from farm fields and protect downstream waters. Reservoirs also act as hotspots for P transformation, as anoxic conditions can facilitate the release of stored P from the lake sediments. The role of inland reservoirs in P speciation at the watershed scale is relatively unexplored. This problem is growing in importance as approximately half of the global river volume is at least moderately impacted by damming, and is projected to reach 93% with all the planned or proposed dams (Grill et al. 2015). Here we use a decade of soluble reactive P (SRP) and total P (TP) concentration data at the inlet and outlet of two reservoirs, Belwood Reservoir and Conestogo Reservoir, in the Grand River Watershed, Canada. The annual SRP and TP percent retention varied at both reservoirs, showing that the reservoirs acted as a sink in some years and as a source in other years. The percent TP retention in Belwood Reservoir varies from -40% to 32%, while percent TP retention in Conestogo Reservoir is generally lower, between -72% to 25%. The SRP retention in Belwood Reservoir varied between -68% and 43%, while SRP retention in Conestogo Reservoir varied between -71% and 28%. Interestingly, the source-sink behaviour is visible for both SRP and TP and they are similar between years. That is, in years that Belwood Reservoir acts as a source of TP, the reservoir often acts as a source of SRP too. At the seasonal scale, we found that both reservoirs increase the proportion of bioavailable P (SRP:TP ratio) from inlet to outlet between April and October. We then built a process-based model to examine the P cycling and sediment-water interactions controlling this speciation of P in the Belwood Reservoir. The model was able to capture downstream SRP export with NSESRP = 0.57-0.86 and TP export with NSETP = 0.60-0.91. The model had difficulty capturing the SRP:TP magnification from inlet to outlet and sediment P accumulation, especially for the first few years of model simulation (2007 - 2012). Model results highlight the role of internal loading during the summer months. As dam construction is on the rise globally, it is critical to understand the impact of reservoirs on the relative reactivity of P in order to mitigate nuisance and potentially harmful algal blooms.

Baseflow with origins from groundwater is a critical component of streamflow sustaining it throughout the year especially during dry periods. To better understand the role of baseflow in streamflow, accurate estimates are needed. This study calculates baseflow through existing graphical and digital filter methods, using actual streamflow data from a gauging station at the Alder Creek Watershed (ACW) and synthetic streamflow data at ten study points within the same watershed simulated with HydroGeoSphere (HGS) (Aquanty Inc., 2018). There are four widely used graphical (Sloto and Crouse, 1996; Aksoy et al., 2008) and six digital filter (Lyne and Hollick, 1979; Chapman and Maxwell, 1996; Furey and Gupta, 2001; Eckhardt, 2005; Tularam and Ilahee, 2008; Aksoy et al., 2009) approaches for baseflow estimation being used and compared. To determine the most optimal estimation approach, baseflow estimates from real data are assessed based on the concept of hydrologic plausibility (Nathan and McMahon, 1990), while baseflow estimates obtained from the HGS streamflow record with graphical and digital filter methods are compared to those computed directly by HGS. Overall, results from this study indicate that baseflow hydrographs reveal a seasonal pattern. During wintertime, streamflow is composed almost entirely of baseflow, whereas during summertime, baseflow only consists approximately 20% to 60% of streamflow. After comparing the baseflow estimates with those computed by HGS, the most optimal approaches at ten points are assessed. Results show that the best approach for six points is the FUKIH (Aksoy et al., 2009) approach; for three points is the Chapman and Maxwell (1996) approach; and for one point is the Eckhardt (2005) approach. In conclusion, it is inferred that the most optimal approach within the ACW varies spatially.

The purpose of this study was to better understand drivers of water problems and their implications for water governance scholarship and practice. Drivers here are identified as social and environmental forces affecting a system, such as a Great Lakes basin. Eutrophication in the western Lake Erie basin provides the empirical setting to achieve the study purpose, using the following objectives: (i) identify drivers of eutrophication in the western Lake Erie basin, (ii) determine whether these drivers are taken into account in nutrient management efforts, and (iii) assess the relationships among drivers and water governance. Two parallel cases of nutrient management, in the Canadian and United States’ (US) portions of the western basin, are explored using policy Delphi surveys with practitioners and researchers, and content analysis of formal nutrient management documents such as strategies and agreements. Findings from the research revealed that nutrient management efforts are identifying many of the same drivers (e.g., agricultural operations) that emerged from the data. Overlaps between the US and Canadian case studies indicate a shared understanding of the causes of eutrophication in the western Lake Erie basin. Areas of overlap between the policy Delphi surveys and document analyses within each of the case studies indicate shared perspectives from experts and formal nutrient management documents. These overlaps were expected, since Lake Erie is a transboundary body of water shared between Canada and the US, and the two countries have worked together for over a century to address water quality and quantity concerns through binational agreements and agencies. However, there were also differences in the sets of identified drivers of eutrophication that raised questions about the strength of shared understandings both within and between the US and Canadian case studies. For example, some drivers identified in the Canadian policy Delphi were not identified in the Canadian document analysis or in the US case study. These gaps indicate areas where nutrient management in the western Lake Erie basin could be improved through coordinated efforts to develop shared problem framings of eutrophication. The study findings empirically demonstrate the importance of identifying drivers when addressing water problems as well as the challenges that can occur when problem framings are mismatched within and across jurisdictions. When determining whether these drivers of eutrophication are taken into account by nutrient management efforts in the western Lake Erie basin, a similar pattern emerged. There was agreement among both the US and Canadian policy Delphi surveys and document analyses that several major drivers of eutrophication are taken into account in nutrient management efforts by multiple mechanisms ranging from research to regulation, including efforts involving agricultural operations. However, gaps were identified both within and between the US and Canadian case studies on whether some drivers of eutrophication are taken into account by nutrient management efforts. The concept of sufficiency emerged from Canadian and US policy Delphi participants. In this way, a driver may be taken into account to some extent, but these efforts were identified as insufficient to mitigate the driver as a cause of eutrophication. Together, these findings raise questions about the effectiveness of nutrient management efforts in the western Lake Erie basin. Specifically, the results suggest that the persistence of eutrophication may be due at least in part to a failure to take drivers into account. From the policy Delphi survey and document analysis data, evidence also emerged regarding the relationships between drivers and water governance. Existing framings of eutrophication do not explicitly identify water governance as a driver of eutrophication. The results demonstrated that water governance could be a driver of eutrophication, for example by shaping nutrient management actions. The results also showed that there are dynamic relationships among drivers of eutrophication and the water governance system, with influences moving bidirectionally across levels and scales. This empirical demonstration contributes to the currently understudied subject of driver directionality in the social-ecological systems and water governance literatures. This research makes several contributions to the water governance and SES literatures. The dissertation contributes to emerging discussions in the water governance literature on the importance of identifying and accounting for drivers, including their relationships with water governance, when addressing water problems. There are currently knowledge gaps on characterizing drivers within specific contexts and understanding the relationships between driver perception and management. I address these gaps by identifying the drivers of eutrophication in the western Lake Erie basin and determining whether those drivers are taken into account by nutrient management efforts. The use of the document analyses in combination with the policy Delphi surveys was a methodological approach that has not been widely used. In this research, the two methods provided nuanced and novel perspectives on drivers of eutrophication. In the eutrophication literature broadly, and in discussions of Lake Erie specifically, drivers of eutrophication are often framed as biophysical or socioeconomic factors. The identification of the water governance system and nutrient management efforts, including actors, policies, and program effectiveness, as drivers of eutrophication in the western Lake Erie basin is a contribution to the eutrophication literature that is unique to this research. This identification and characterization of water governance systems as a driver also adds value to other explorations of water problems and their associated water governance systems. Overall, the research contributes to understandings in the water governance literature on the relationships that exist between water governance and drivers by demonstrating that these relationships are bidirectional and exist across scales. The empirical findings of the research demonstrate the utility of regional case studies for filling identified knowledge gaps in the SES literature on understanding how drivers interact, how governance affects solutions to water problems, and how these relationships have influence across scales, as well as demonstrating the need for additional research. The findings of this research also have implications for the empirical practice of water governance in the western Lake Erie basin. By identifying gaps in identifying drivers and accounting for drivers through nutrient management efforts, the research identifies opportunities to improve nutrient management efforts and water quality outcomes in the western Lake Erie basin. Specifically, there are differences in how drivers of eutrophication are understood and taken into account by nutrient management efforts. These differences indicate that work is necessary both within the Canadian and US jurisdictions as well as binationally to deepen the shared understanding of eutrophication and how its causes are perceived. To improve water quality in the western Lake Erie basin, it is necessary to critically examine nutrient management effectiveness and incorporate the novel drivers of eutrophication identified by the research into framings of eutrophication as well as nutrient management solutions.

Nutrient losses from agricultural fields are the largest sources of phosphorus (P) entering the Great Lakes in North America. Research has suggested that multiple conservation practices (CPs) used together (stacked) are an effective way to reduce the amount of P losses individual fields; however, some P loss still occurs. Advancements in the chemical removal of P have provided landowners with an opportunity to capture P that has leaves fields in runoff before it enters local waterways and act as a final polishing agent. A commercially available phosphorus sorbing material (PSM) in the form of a geotextile filter was installed on two well managed fields in midwestern Ontario (ILD and LON) to determine its efficacy in removing dissolved reactive P (DRP), total P (TP) and total suspended solids (TSS) from surface runoff, thereby reducing edge of field P losses. Laboratory tests on unused and used filter material were also conducted to try to determine the sorption potential, amount of P stored in the filter and the mechanisms of P removal. During the two-year study period, the filter removed 0.018 kg/ha of DRP, 0.4 kg/ha of TP and 8.75 kg/ha of TSS at the ILD site. In contrast, the filter at LON released 0.22 kg/ha of DRP and 0.15 kg/ha TP, but removed 37 kg/ha of TSS. The filter most effectively removed P within 8 months of filter installation, suggesting that time was a critical factor impacting performance, among others. Laboratory tests on unused (new) and used (field) filter material indicated that the raw filter material had a large potential to adsorb DRP under controlled conditions, and that this potential was smaller in used material from LON but not ILD. The extraction of P from used filter material indicated that the filters retained approximately 200 mg/kg of DRP at each site (0.0027 kg/ha at ILD and 0.0022 kg/ha at LON) with the majority of this P held in more soluble form likely associated with the metal oxides/clay filter components. This suggests that previously retained P has the potential to be rereleased from the filter. Additionally, the amounts of P held in the filter material calculated via lab tests was considerably lower than the amount removed through the water samples calculations suggesting that some of the P removed by the filter did not stay inside the material. The results of this study demonstrate that P has the potential to be chemically removed at the edge-of-field, but the efficacy of filters as a CP differs in both space and time. This thesis has shown how an edge-of-field filter for surface runoff can be implemented in a field setting in midwestern Ontario, and has identified and which factors are most important to determining the efficacy of this practice.

Soil erosion remains a primary challenge in the 21st century threatening fresh water and cropland that supports more than 95% of global food production. It is of significance to plan for and prevent soil erosion in its initial stages rather than labor intensive repairing later. The Middle Thames River watershed has suffered from severe erosion issues for more than ten years with 21% highly erodible lands throughout the basin, where extensive soil conservation measures are highly encouraged. A series of practical measures that landowners can apply to enhance soil health and water quality while preserving or increasing agricultural production are termed farmland Best Management Practices (BMPs). Among these measures, grassed waterways, as broad and shallow channels to move concentrated surface runoff, are considered as one of the most effective measures to prevent ephemeral soil erosion. Therefore, identifying the site-specific opportunities for grassed waterways implementation in the Middle Thames River watershed can support targeted soil conservation and the watershed planning. This study aims to identify the potential locations for grassed waterways implementation in the Middle Thames River Watershed using four different techniques with high-resolution data (Compound Topographic Index model, Stream Power Index threshold model, weighted linear overlay, fuzzy logic analysis). The Compound Topographic Index model and Stream Power Index threshold model have been developed to predict the existing and potential grassed waterways at the field level. Then the Compound Topographic Index and Stream Power Index threshold models, the multi-criteria decision analysis (MCDA) has been conducted to map the priority areas for grassed waterways implementation at the watershed scale. The output maps of the Compound Topographic Index model and Stream Power Index threshold model display the location and length of predicted grassed waterways in each field. To better visualize the results of the Compound Topographic Index model and Stream Power Index threshold model, the density distribution maps of predicted grassed waterways throughout the studied watershed have been created based on the outputs from Compound Topographic Index and Stream Power Index threshold model. The performance of the Compound Topographic Index and Stream Power Index threshold model have been assessed by visual evaluation, occurrence evaluation and length evaluation. After developing Compound Topographic Index and Stream Power Index threshold models, the multi-criteria decision analysis (MCDA) has been conducted to map the priority areas for grassed waterways implementation at the watershed scale. Twelve factors were selected as criteria of MCDA based on literature review, data availability and geographic knowledge. Two methods including weighted linear combination and fuzzy logic analysis were employed in MCDA, which produced two outputs maps of priority areas for grassed waterways implementation. The results of these two maps have been validated using existing grassed waterways. The results of the Compound Topographic Index model and Stream Power Index threshold model display the existing and predicted grassed waterways in each field. The Compound Topographic Index model with the threshold of 600 has identified 30 existing grassed waterways, while the Stream Power Index threshold model with the threshold of 0.01 standard deviation identified 23 grassed waterways. Several discontinuities exist in predicted grassed waterways along the trajectories of digitized grassed waterways. The lengths of predicted grassed waterways by Compound Topographic Index model have a much better agreement with observation than that of Stream Power Index threshold model. The density distribution map of Compound Topographic Index and Compound Topographic Index model presented high-density areas of predicted grassed waterways which are mainly situated in the northern and central part of the study area, especially the areas along the upstream of Middle Thames River and Nissouri creek. The low-density areas for grassed waterways implementation are mostly located in the southwestern part of the study area. The results of weighted linear combination and fuzzy logic analysis displayed the high-priority areas mainly located in the northwestern part of the watershed, especially along the upstream of Nissouri creek. It is found that these upstream areas have relatively steeper slope gradient than other areas in the studied watershed, with dominant soil type of sandy loam and silty loam. There are more areas belonging to the lowest priority zone and lower areas falling into the most priority level in the fuzzy logic analysis output map, compared with the map of weighted linear combination. The fuzzy logic analysis required less prior knowledge of the relationship among criteria, which provide more flexibility and convenience to decision makers. The validation of both weighted linear combination and fuzzy logic analysis output maps displays relatively good performance, based on the criteria that a greater percentage of grassed waterways implementation must occur in the higher priority zones (Kanungo et al., 2009).

In recent times, with the increasing availability of large datasets, applications of machine learning techniques have grown at a rapid speed. However, due to the black-box nature of these tools, it can be hard for model builders to understand the detailed structure of the system that machine learning models simulate. Agent-based modelling (ABM) is a popular approach to studying complex systems., One of the challenges for this technique is to design the decision making processes of the agents in the model. As machine learning tools have a strong ability to transform the information from the raw data into a functional model as the decision making processes for agents in ABMs. Because an ABM can provide a detailed structure for the system that the machine learning model simulates, it is reasonable to combine the two kinds of models. However, although in previous studies, some researchers combine the two models, most of them use one of the two models as a validation tool for the other, rather than to integrate the machine learning model into the decision making processes of agents in ABMs. Therefore, this thesis focuses on integrating a machine learning model into the ABM, and contrast it with the ABMs with two traditional decision making models, including an optimal model and a stochastic model. To compare the three decision making models, we use farmers’ BMP adoption case in the Upper Medway subwatershed, and contrast the three models through three metrics, including the percentage of BMP adoption, size of agricultural land of BMP adoption, and the correlation between BMP adoption and landuse types. As a result, the ABM with the machine learning model presents a high level of accuracy compared with the other two traditional models, but its adaptability to other cases and the robustness to uncertainties still require a further study.

Fishes play crucial roles in the ecology of aquatic environments and contribute to the multi-billion-dollar fisheries industry. The integrity of their populations and health needs to be maintained for future generations and research on the biology and effects of stress on fish can contribute to this cause. While much is understood about the adrenergic response to stress which results in the secretion of catecholamines and cortisol, there is much to be understood about molecular mechanisms of stress, such as the role of microRNAs. MicroRNAs (miRNAs) regulate post-transcriptional molecular responses by binding to mRNA and labelling them for degradation or blocking translation, effectively decreasing target protein translation levels. The response of miRNA transcript levels to environmental stressors, such as increased water temperatures, have been measured in fish since 2009. However, there is still much that is poorly understood about the effects of fish stress on miRNA levels, such as how time sensitive the response is, whether the response is tissue specific, and whether it is possible to measure miRNAs in non-lethal or non-invasive samples, such as mucus or the water surrounding fish. Furthermore, there are many gaps in understanding of how miRNA levels are altered in non-model species, such as salmonids. Most studies of fish stressors often focus on single stressor studies to elucidate the molecular mechanisms however, fish are not exposed to stressors individually in the aquatic environment. Therefore, it is important to study the simultaneous effects of multiple, emerging anthropogenic stressors of concern (e.g., increased water temperature, decreased dissolved oxygen, and pharmaceutical contaminants), on fish and their miRNA levels. The overall goal of my thesis is to determine how transcript levels of miRNAs are regulated when fishes are exposed to stress. More specifically, I wanted to further characterize the miRNA response in different tissues and at different timepoints, in both model and non-model fish species, to determine the conservation or specificity of the miRNA response. I also aimed to determine if it was possible to sample miRNAs in non-lethally collected samples as a novel method of measuring stress in fish. Furthermore, I measured predicted downstream responses (mRNA transcript levels, protein abundances, and enzyme activities) to understand the functional implications of changes to miRNA transcript levels and to describe the molecular response to acute and anthropogenic stressors. In studying the effects of chronic exposure to anthropogenic stressors on zebrafish gonads, I found that the miRNA response was reversible and associated with adverse reproductive impacts (Chapter 2). In studying the effects of different lengths of exposure to anthropogenic stressors on zebrafish liver and muscle tissues, I determined that the miRNA response was specific to length of exposure, tissue type, as well as the sex of the fish, and that fish were activating cell stress, decontamination, metabolic, and reproductive responses (Chapter 3). In studying the effects of acute stress on rainbow and brook trout liver tissues, I found that miRNA transcript levels, mRNA transcript levels, and metabolic enzyme activities were altered in a time-dependent manner post-stress and that there was much intra-species and inter-species variability (Chapter 4). In studying the effects of acute stress on rainbow trout blood plasma, mucus, and the surrounding water, I found that miRNAs were able to be measured and transcript levels were altered following stress in all three non-lethal sampling locations (Chapter 5). Altogether, I have contributed further to identifying specific transcript levels of miRNA that respond to acute and anthropogenic stressors in multiple fish species. I have also characterized how the miRNA response is associated with the presence of the stressor, the length of exposure to the stressor, and the length of time following exposure to the stressor. These data are helpful in understanding the molecular regulation and response to stress and broadly contribute to understanding how miRNAs play a role in how organisms can adapt to transient or ongoing stressors. In addition, the downstream molecular responses associated with changes in miRNA transcript levels were also measured in response to these stressors and highlight the metabolic, reproductive, and cellular stress responses that the fish were activating when exposed to anthropogenic stressors, as well as filling in gaps of metabolic enzymes that are part of the acute stress response. My research also highlights the complex role that miRNAs play in finetuning the molecular response to stress, as there are still many gaps in understanding what the altered miRNA transcript levels are targeting and post transcriptionally regulating. In the future, instead of focusing on identifying miRNAs that are regulating a particular transcript or pathway of interest, priority can be given to identifying miRNAs that are crucial in driving the stress response or in allowing a particular individual or species to adapt to stress.

Unlike conventional salt or dye tracers, artificial/synthetic DNA hydrologic tracers are essentially non-toxic and several can be used simultaneously. These features have implications for DNA to be used to better understand environmental processes and map hydrological pathways in complex environments, such as watersheds. Synthetic DNA tracers also have the potential to help with a better understanding of the behaviour of environmental DNA (eDNA) and its application in biomonitoring. Some components of eDNA exist as free/naked DNA not bound to other substances or protected by cellular material. The goal of the current research was to assess the fate and transport of naked DNA (short single stranded DNA sequences) in a small stream as a potential environmental tracer. As a proof-of-concept, two unique DNA tracers were released into an upstream location in Washington Creek (southern Ontario). After releasing the tracers, water samples were collected 100 m and 350 m downstream and breakthrough curves of tracer concentration were plotted over time. Both tracers behaved similarly with a mass recovery of 71% (T11) and 80% (T22) at the 100 m downstream sampling location and about 70% for both tracers at 350 m downstream. The downstream tracer peak arrival times were 15 – 16 min and 30 – 31 min at the 100 m and 350 m sampling sites, respectively, demonstrating that naked DNA injected into the stream can quickly travel downstream. This suggests that eDNA, in the naked form, may survive considerable distances downstream from the source and has implications for biomonitoring strategies. Additional unique DNA tracers were designed and optimized for future experiments. DNA tracers create many opportunities for applications in environmental sciences, especially if they can be combined with other substances to alter their environmental properties and fate (i.e., nanoparticles). DNA tracers can be attached to nanoparticles to protect DNA degradation in harsh environments or influence their zeta potential.

High-latitude cold regions are warming more than twice as fast as the rest of the planet, with the greatest warming occurring during the winter. Warmer winters are associated with shorter periods of snow cover, resulting in more frequent and extensive soil freezing and thawing. Freeze-thaw cycles (FTC) influence soil chemical, biological, and physical properties and any changes to winter soil processes may impact carbon and nutrients export from affected soils, possibly altering soil health and nearby water quality. Changes to non-growing season climate affect soil biogeochemical processes and fluxes and understanding these changes is critical for predicting nutrient availability in cold region ecosystems and their impacts on downstream water quality. These impacts are relevant for agricultural soils and practices in cold regions as they are important in governing water flows and quality within agroecosystems. Agricultural systems are source areas for nutrient pollutants due to fertilizer use and have been the target of numerous management strategies. Sustainable agricultural practices have been increasingly employed to mitigate nutrient loss due to erosion, but nutrient export via surface runoff, subsurface leaching, and volatilization allows for continued high nutrient losses (Beach et al., 2018; King et al., 2017). Chapter 1 of thesis discusses the non-growing season climate changes altering winter soil processes and reviews the major nitrogen transformation processes leading to nitrogen losses in agricultural soils. In Chapter 2, I present a soil column experiment to assess the leaching of nutrients from fertilized agricultural soil during the non-growing season. Four soil columns were exposed to a non-growing season temperature and precipitation model and fertilizer amendments were made to two of the columns to determine the efficacy of fall-applied fertilizers and compared to other two unfertilized control columns. Leachates from the soil columns were collected and analyzed for cations and anions. The experiment results showed that a transition from a freeze period to a thaw period resulted in significant loss of chloride (Cl-), sulfate (SO42-) and nitrate (NO3-). Even with low NO3- concentrations in the applied artificial rainwater and fertilizer, high NO3- concentrations (~150 mg L-1) were observed in fertilized column leachates. Simple plug flow reactor model results indicate the high NO3- leachates are found to be due to active nitrification occurring in the upper oxidized portion of the soil columns mimicking overwinter NO3- losses via nitrification in agricultural fields. The low NO3- leachates in unfertilized columns suggest that FTC had little effect on N mineralization in soil. In Chapter 3, I provide a brief review of nitrification inhibitors and how soil properties impact nitrification inhibitor efficacy. There are only a few studies on the relationship between nitrification inhibitor efficacy and climatic factors, especially in regard to FTC. I conducted a sacrificial soil batch experiment to determine if and how nitrification inhibitors were impacted by FTC to further explore the results of Chapter 2. The batch experiment revealed the nitrification inhibitors were effective at mitigating NO3- production under freeze-thaw conditions but more effective at mitigating these losses under thaw conditions. The soils exposed to the FTC condition experienced significant N mineralization flushes in contrast to the lack of mineralization induced by FTCs in the experiment detailed in Chapter 2. In Chapter 4, I summarize the key findings of this thesis. The results showed fertilizer loss and nitrification inhibitor effectiveness are affected by freeze-thaw cycling in arable soil. The experimental and modeling results reported in this thesis could be used to bolster winter soil biogeochemical models by elucidating nutrient fluxes over changing winter conditions to refine best management practices for fertilizer application. Ultimately, these results and the conclusions drawn from them highlight several research pathways that could be undertaken to progress our understanding of the complex interactions between FTC and fertilizer dynamics.

With recent developments in oil and gas exploration technologies that have opened regions of Canada’s northern territories and the threat of climate change, uncertainties around how these factors may impact the environment in these areas are profuse. To reduce uncertainty and allow for mitigation planning, having effective baseline monitoring of environmental systems, such as groundwater, is critical. However, baseline monitoring studies of groundwater resources in these regions are complex and expensive to undertake as compared to studies in more southern regions. This is due mainly to the remoteness, lack of infrastructure and presence of discontinuous permafrost that complicates the use of traditional groundwater monitoring methods in northern regions. The work outlined in this thesis set out to improve baseline monitoring studies of groundwater in discontinuous permafrost areas. A suite of geochemical and isotopic tracers combined with physically based hydrologic measurements were tested in two summer field campaigns within the Bogg Creek Watershed, a small subcatchment of the Mackenzie River in the Northwest Territories (NWT). These data were acquired through strategic sampling utilizing portable and lightweight equipment, guided by previous remote sensing work and an aerial infrared survey. This field data was combined with a variety of other data sets acquired through public records and reports, as well as through collaboration with interested third parties. Physical data provided evidence for groundwater discharge in some areas, while the geochemical and isotopic evidence allowed for fingerprinting of these groundwater sources. In total, 5 groundwater source groups were identified in the study area. These included shallow seepage water and organic active layer porewater (both Ca-SO4¬), suprapermafrost groundwater originating from mineral soils (Ca-HCO3), subpermafrost groundwater from the Little Bear Formation aquifer (Na-HCO3), and subpermafrost groundwater from the Martin House Formation (Na-Cl/HCO3). Evidence of suprapermafrost and subpermafrost groundwater contributions were found in Bogg Creek and its headwater tributaries as well as in several springs. Suprapermafrost groundwater influence occurs throughout the watershed but is most dominant in upland tributaries. Evidence for Little Bear Formation groundwater influence was found in the geochemical and 87Sr/86Sr signature of one of the tributaries of Bogg Creek, downstream of several mapped icings. Evidence of Martin House Formation groundwater influence was noted in geochemistry from a spring complex, associated with mapped icings. This spring water appeared to be mixtures of both subpermafrost and suprapermafrost groundwaters, but this is not certain. There is also evidence for Martin House Formation water in the lower reaches of Bogg Creek. δ18O and δ2H data suggest these contributions are quite small within the creek and its tributaries but are more substantial in springs. This highlights the sensitivity inherent in each method, where geochemistry is quite responsive to minor contributions of subpermafrost groundwater due to a much higher degree of contrast between solute concentrations in suprapermafrost and subpermafrost groundwater endmembers. There is less contrast in δ18O and δ2H. 87Sr/86Sr appeared to confirm what was indicated in the geochemical results, while δ13C in CH4 and 3H were inconclusive. A conceptual model of the site hydrogeologic system was developed based on this information. Some uncertainty remains on exact groundwater origins due to the limitations of the geochemical and isotopic species used, but overall results indicate that these techniques could be applied in similar discontinuous permafrost environments. Portable sampling techniques used in the shallow subsurface might be able to effectively characterize suprapermafrost groundwaters in an area quickly and efficiently, limiting the need for extensive monitoring well networks. Large contrasts from this water found in surface water or springs may be indicative of water originating from another source, such as from a subpermafrost aquifer. However, a certain amount of subpermafrost groundwater sampling is needed, as many uncertainties arise with missing endmembers.

Terrestrial snow is an important freshwater reservoir with significant influence on the climate and energy balance. It exhibits natural spatiotemporal variability which has been enhanced by climate change, thus it is important to monitor on a large scale. Active microwave, or radar remote sensing has shown frequency-dependent promise in this regard, however, interpretation remains a challenge. The aim of this thesis was to develop constraints for radar based SWE retrievals which characterize and limit uncertainty with a focus on the underlying physical processes, snowpack stratigraphy, the influence of vegetation, and effects of background scattering. The University of Waterloo Scatterometer (UWScat) was used to make measurements at 9.6 and 17.2 GHz of snow and bare ground in a series of field-based campaigns in Maryhill and Englehart, ON, Grand Mesa, CO (NASA SnowEx campaign, year 1), and Trail Valley Creek, NT. Additional measurements from Tobermory, ON, and Churchill, MB (Canadian Snow and Ice Experiment) were included. The Microwave Emission Model for Layered Snowpacks, Version 3, adapted for backscattering (MEMLS3&a) was used to explore snowpack parameterization and SWE retrieval and the Freeman-Durden three component decomposition (FD3c) was used to leverage the polarimetric response. Physical processes in the snow accumulation environment demonstrated influence on regional snowpack parameterization and constraints in a SWE retrieval context with a single-layer snowpack parameterization for Maryhill, ON and a two-layer snowpack parameterization for Englehart, ON resulting in a retrieval RMSE of 21.9 mm SWE and 24.6 mm SWE, respectively. Use of in situ snow depths improved RMSE to 12.0 mm SWE and 10.9 mm SWE, while accounting for soil scattering effects further improved RMSE by up to 6.3 mm SWE. At sites with vegetation and ice lenses, RMSE improved from 60.4 mm SWE to 21.1 mm SWE when in situ snow depths were used. These results compare favorably with the common accuracy requirement of RMSE ≤ 30 mm and underscore the importance of understanding the driving physical processes in a snow accumulation environment and the utility of their regional manifestation in a SWE retrieval context. A relationship between wind slab thickness and the double-bounce component of the FD3c in a tundra snowpack was introduced for incidence angles ≥ 46° and wind slab thickness ≥ 19 cm. Estimates of wind slab thickness and SWE resulted in an RMSE of 6.0 cm and 5.5 mm, respectively. The increased double-bounce scattering was associated with path length increase within a growing wind slab layer. Signal attenuation in a sub-canopy SWE retrieval was also explored. The volume scattering component of the FD3c yielded similar performance to forest fraction in the retrieval with several distinct advantages including a real-time description of forest condition, accounting for canopy geometry without ancillary information, and providing coincident information on forest canopy in remote locations. Overall, this work demonstrated how physical processes can manifest regional outcomes, it quantified effects of natural inclusions and background scattering on SWE retrievals, it provided a means to constrain wind slab thickness in a tundra environment, and it improved characterization of coniferous forest in a sub-canopy SWE retrieval context. Future work should focus on identifying ice and vegetation conditions prior to SWE retrieval, testing the spatiotemporal validity of the methods developed herein, and finally, improving the integration of snowpack attenuation within retrieval efforts.

Freshwater quality issues are among the most pressing challenges of our time. Such issues are increasingly complex and tend to recur when we fail to acknowledge the interacting stressors that influence them. One example of a recurring issue is the prolific growth of Cladophora (a benthic nuisance alga) in the eastern basin of Lake Erie. Water managers thought they had corrected the issue by controlling nutrient loading from the 1970s to the1990s; however, the Cladophora issue returned in the mid-2000s and has persisted due to new factors changing the way the ecosystem works. The Grand River in Southern Ontario remains Lake Erie’s largest contributor of nutrients in Canada, and so is the focus of current management efforts. Problems like this, which are caused by several interacting factors in a given space over time, are known as cumulative effects. Much of the literature on cumulative effects and/or water quality monitoring in this dissertation reflects conventional practice focused on the perspectives of water scientists and managers; however, this dissertation does not replicate this approach. Instead, the social-ecological context surrounding freshwater quality monitoring in the study area is critically considered by incorporating diverse community perspectives alongside conventional perspectives. In the study area, Indigenous communities have treaty rights to participate in the governance of the watershed (which sits entirely within the Haldimand Tract), but these communities – like others – have not been engaged as partners in water quality monitoring or management. One reason for this is that community and Indigenous knowledges often come in different formats than conventional scientists are used to dealing with, and so these forms of community ‘data’ are not easily integrated with conventional data. As Canada moves towards a mandate for reconciliation with Indigenous communities, ignoring the challenge of bringing together different ‘ways of knowing’ is no longer acceptable. Inspired by the Cladophora challenge and the need to diversify monitoring practice, this research strives to answer the following question: How can cumulative effects water quality monitoring be enabled and involve diverse perspectives in the Grand River-Lake Erie interface? This research encourages the democratization of water quality monitoring to ensure more diverse persons can participate in the gathering of water quality information and that their diverse ways of knowing may supplement conventional science in management and decision-making. In other words, this dissertation explores approaches for diversifying perspectives that contribute to our understanding of freshwater quality in the study area. A multimethod approach to research was undertaken to explore what may be done differently. Methods used in this research include a systematic review of monitoring programs (Chapter 3), key informant interviews (Chapter 4), in-person and online workshops (Chapters 5, 7, and 8), and artistic research (Chapter 6) – a new approach in the context of water quality monitoring and management in the study area. First, the systematic review of monitoring programs highlighted aspects of current monitoring to maintain and improve upon. Then, key informant interviews raised 106 strengths, weaknesses, opportunities, and threats, as well as 51 recommendations. I also discuss a culture shift towards more holistic thinking and more collaborative water governance, which study participants deemed necessary to develop a strong and resilient cumulative effects monitoring program. To enable this culture shift, two examples of artistic research were implemented to demonstrate potential approaches for diversifying practice. Following, eight recommendations are provided for implementing cumulative effects monitoring in the study area. The multimethod approach results in a framework for collaboration (i.e., organizational structure and process framework) to enable more diverse and collaborative water quality monitoring in the study area that contributes to our ability to understand and address cumulative effects. The proposed framework is community-led (whether catalyzed by community members or invited by government) and incorporates equal weighting of Indigenous and western priorities and monitoring indicators – a unique and potentially transformative contribution to literature and practice. The use of artistic research as an equitable means of community involvement is also new in the study area. Finally, because involving diverse persons to contribute their perspectives demanded the development of different approaches than currently practiced, the research process and its process-related lessons and recommendations may contribute to raising the standard for future research and practice in water quality monitoring. This research also has implications that extend beyond strengthening the practice of water quality monitoring. The core outcomes of the later chapters – e.g., recommendations towards collaborative and community-based monitoring processes coupled with a culture shift regarding the creation and application of knowledge – would, if practiced, support at least three broader transformations in society: a formal sharing of responsibility over natural resources, increased collaboration that is mindful of diversity, and systemic changes in support of Canadian-Indigenous reconciliation. While many aspects of the future scenarios described in the concluding chapter are likely a generation away (or longer) and are far beyond the scope of any one thesis project, my hope is that possible actions catalyzed by this research and other efforts like it will collectively move society in a different, more equitable direction.

Microfluidics-based Lab-on-a-Chip platforms have drawn ever-increasing attention from both academy and industry due to their advantages for dealing with small volume of fluids and for integrating multiple processes into one platform. These advantages are the direct benefits of miniaturization which also brings challenges, especially in sensing and heating. The challenges are augmented in the context of droplet microfluidics because of their fast motion, curved interface and reduced volume (i.e. pico- to nano-liter). Droplet microfluidics utilizes water-in-oil or oil-in-water droplets that can be generated in microchannel networks at kHz rates as mobilized test tubes. It presents tremendous potential to serve as a tool for high throughput analysis that are in high demand in many areas such as material synthesis, life science research, pharmaceutical industry and environmental monitoring. Many applications require temperature control and both fundamental and applied research need droplet sensing to assist in understanding droplet motion and developing techniques for manipulating droplets. Microwave sensing offers unique advantages by differentiating materials based on their electrical properties at high speeds. Moreover, it enables simultaneous heating of individual droplets. Previous studies demonstrated the potential of microwave resonator for point of care (POC) applications and for simultaneous sensing and heating. However, neither of them has yet be fully realized. In addition to the technical challenges such as the use of bulky and expensive vector network analyzer (VNA) for sensing that limits the potential for POC applications, fundamental understanding of microwave heating and its coupling with droplet microfluidics is lacking. This thesis is designed to fill the gap with the ultimate goal of enhancing droplet microfluidics as an enabling tool for a wide range of applications by realizing the full potential of microwave sensing and heating. With the goal of maximizing the capacity of droplet microfluidics serving as an enabling tool for many applications, this thesis focuses on exploring microwave sensing and heating for droplet microfluidics. The thesis started with the investigation of the coupling between microwave heating and droplet motion to shine light on the mechanism of microwave heating induced droplet mixing. Followed the improved understanding of microwave heating, on-demand droplet generation via microwave heating was explored and demonstrated. To realize simultaneous sensing and heating which is powerful for droplet microfluidics, two resonators need to be considered and the primary concern for two resonators in a single microfluidic chip is the crosstalk between the two resonators. The third chapter was designed to investigate the fundamental challenges of integrating two resonators within a typical microfluidic device footprint. Finally, a POC application of microwave sensing is demonstrated for real time detecting lead in drinking water system which has been one of the crisis raised recently.

Background: Climate change in tandem with previous colonial policies have severely impacted local traditional food systems, resulting in a disruption to traditional knowledge sharing, a greater reliance on market foods and disproportionate rates of food insecurity. Objectives: The purpose of this thesis is to understand how an Indigenous self-government can leverage a mobile abattoir (also known as a trailer facility) to increase food access for community members. This pertains to understanding what is possible within the trailer facility as well as understanding the power of an Indigenous self-government in influencing policies that pertain to incorporating the trailer facility into the local food system. Methods: The methodology of this thesis is based upon the principles of participatory action research, evaluation, and settler colonial theory. Data was collected in interviews with trainees of the country food processing training course, surveys from community members and a title screening process of all the legislated acts and regulations in the Northwest Territories. Two different analyses were conducted. First, a reflexive thematic analysis of interview data from trainees of the course was used to inform a process evaluation on a country food processing training course held within the trailer facility. Second, an applied approach to Norman Fairclough’s (1989) critical discourse analysis was adapted to the policies identified within title screening process. The resulting data corpus was too large to incorporate into this applied thesis. Data reduction resulted in a data set of two contemporary treaty texts: The Sahtú Dene Métis Land Claim Agreement and the Délı̨ne Final Self-Government Agreement. Results: We found that the trailer facility was perceived to be a tool that could create employment opportunities, teach youth the importance of the traditional way and create traditional/country foods that cater to the food preferences of the community’s youth. The trailer facility was found to not be a viable place to hold the country food processing training course in due to its size and faulty equipment inside. The factors that influenced why the trailer was not viable for the training course were found to be compounded by policies from different levels of government. Through investigating how these policies implicate Délı̨ne, it was revealed that local food development is not within the jurisdiction of an Indigenous self-governance. Wildlife conservation discourses, and agricultural discourses were found within this thesis as affecting an Indigenous self-government from self-determination and food sovereignty. Conclusion: Policies at both the territorial and the regional level create barriers for Indigenous self-governments to implement a trailer facility to its local food system to increase food access. The findings of this thesis highlight how contemporary treaty texts such as Indigenous self-government agreements continue to legitimize the authority of settler governments and control Indigenous Peoples. This thesis provides empirical evidence of the mechanisms within contemporary treaty texts that settler colonialism uses to undermine Indigenous self-determination and Indigenous food sovereignty.

Mining operations at Giant and Con mines (Northwest Territories, Canada) resulted in the release of >20,000 tonnes of arsenic trioxide (As2O3) into the atmosphere, mainly during the 1950s, which were deposited on the surrounding landscape. Studies of arsenic concentrations in lake water and sediment have concluded that no potential ecosystem health effects exist beyond a 40-km radius of the mines. However, paleolimnological studies at distances well beyond 100-km have identified elevated arsenic concentrations aligning with the timing of peak emissions. To improve characterization of the legacy footprint of emissions, spatiotemporal patterns of metal deposition were reconstructed from the analysis of sediment cores at lakes located 10-40 km (near-field) and 50-80 km (far-field) along the prevailing northwesterly wind direction (NW) and 20-40 km to the northeast (NE). Results based on concentrations of mining-associated metal(loids) (arsenic, antimony, lead) in radiometrically-dated (210Pb, 137Cs) sediment cores, enrichment factors, and total excess inventories for arsenic and antimony assert that deposition of these pollutants was greatest closest to the mines and along the prevailing wind direction (NW). Enrichment is evident as far as 80-km to the NW (considerable for arsenic; severe for antimony) and 40-km to the NE (considerable for arsenic; severe for antimony) suggesting pollution from the mines likely travelled distances beyond those explored here. Additionally, the presence of elevated metal concentrations in uppermost sediment strata at near-field lakes suggest that deposition of anthropogenic-sourced metals from lake catchments remains ongoing. Differences in the degree of enrichment and stratigraphic profiles among lake groups are likely due to availability of catchment-sourced legacy metals and post-depositional mobilization from stores in lake sediment. Long-term sources of legacy metals in the near-field environment urge further research on metal mobilization linkages between terrestrial and aquatic ecosystems.

Potential for downstream delivery of contaminants via Athabasca River floodwaters to lakes of the Peace-Athabasca Delta (PAD), northeastern Alberta (58.6°N, 111.8°W), has raised local to international concern. Prior investigations have shown metals concentrations in sediment of lakes supplied by river floodwaters are not enriched above pre-industrial baselines. Additional real-time aquatic ecosystem monitoring approaches are needed to complement sediment-based techniques where time intervals captured are uncertain. Here, we quantify enrichment of eight metals (Be, Cd, Cr, Cu, Ni, Pb, V, Zn) at the base of aquatic food webs, relative to sediment-based pre-industrial baselines, via analysis of biofilm-sediment mixtures accrued on artificial substrate samplers deployed during summers of 2017 and 2018 in >40 lakes spanning hydrological gradients of the PAD. Widespread flooding in spring 2018 allows for assessment of metals enrichment by Athabasca River floodwaters. A main finding is that river floodwaters are not implicated as a pathway of metals enrichment to biofilm-sediment mixtures in PAD lakes from upstream sources. MANOVA tests revealed no significant difference in residual concentrations of all eight metals in lakes that did not flood versus lakes that flooded during one or both study years. Also, no enrichment was detected for concentrations of biologically inert metals (Be, Cr, Pb), and those related to oil-sands development (Ni, V). Enrichment of Cd, Cu, and Zn at non-flooded lakes, however, suggests uptake of biologically active metals complicates comparisons of organic-rich biofilm-sediment mixtures to sediment-derived baselines for these metals. Results lend confidence that this novel approach could be adopted for lake monitoring within the Wood Buffalo National Park Action Plan.

Global climate change has sparked various concerns over the future of the Arctic. One of the major concerns around the environmental and ecological health of the Arctic is directly related to the deterioration of the permafrost. Permafrost is described as frozen soil below 0oC for at least two consecutive years, and is recognized by the World Meteorological Organization (WMO) as an Essential Climate Variable (ECV). Geophysical methods have been used to detect and measure the extent of the permafrost in various cold regions of the Earth. Traditional methods such as electrical resistivity tomography (ERT), electro magnetic induction (EMI), and seismic have all been used to characterize permafrost in the subsurface. However, there are smaller scale features at the near surface requiring attention. In this work, we used of a permafrost probe, an electrical resistivity tomography system, and an electromagnetic induction system to measure the depth to the permafrost table from the ground surface. The study was performed in the Sahtu Region of the Northwest Territories, approximately 30 kilometers south of the Town of Norman Wells, Northwest Territories. Two sites were selected; one on a drill pad, one near a lake shore. The soils mainly consisted of homogeneous organic-rich till. The permafrost probe measure a depth to permafrost table of approximately 70 centimeters at the drill-pad site (MW04T) and approximately 30 centimeters at the lake shore site (Marg Lake). A Syscal Junior 48TM ERT system was installed perpendicular to the topological features, such as the lake shore, and the tree line. The electrode spacing was small due to the shallow nature of the permafrost, and the dipole-dipole method was selected to collect measurements. The ERT data was inverted using Res2DInvTM and the output data correlates well with the permafrost probe measurements. A ground conductivity meter (GCM) was used to assess the capability of using a non-ground-coupled geophysical methods in this terrain to detect permafrost discontinuity. We deployed the Geonics EM-31TM and EM-34TM systems. The electrical data was collected over the same permafrost probe and ERT survey lines and measurements were plotted using MATLABTM. The data suggest that the GCM systems were able to e ectively detect the change in permafrost table depth, correlative todirect measurement of permafrost depth that were used concurrently. This study serves as a baseline analysis of using small-scaled ground-based geophysical systems to detect permafrost discontinuities in this region, and informs the future development of aerial-based systems and methods to gather multiple strings of data to estimate permafrost table depth, and the integrity of the permafrost.

Water management involves monitoring, predicting, and stewarding the quality and quantity of groundwater recharge at the watershed scale. Recharge sustains baseflow to streams and replenishes water extracted by pumping at wells; it is frequently estimated using numerical models that couple or fully integrate surface water and groundwater domains and use water budgets to partition water into various components of the hydrological cycle. However, uncertainty associated with the input data for large components such as precipitation and evapotranspiration may hinder model accuracy, and preferential flow dynamics such as depression focused recharge (DFR) may not be represented at typical modelling scales (≥10s of sq. km) or with typical approaches. The present study addressed two themes related to groundwater sustainability and vulnerability: 1) the sensitivity of modelled recharge estimates to the spatial variability of rainfall, and 2) the vulnerability of public supply wells to DFR during large-magnitude rainfall or snowmelt events. The region investigated during this research was the Alder Creek watershed (78 sq. km), a typical southern Ontario setting overlying glacial moraine sediments with mostly agricultural land use, some urban and aggregate resource development, and whose recharge supplies multiple municipal well fields for the cities of Kitchener and Waterloo. Rainfall is often the largest component of the water budget and even a small uncertainty percentage may lead to challenges for accurately estimating groundwater recharge as a calculated residual within a water budget approach. However, rainfall monitoring networks typically have widely spaced gauges that are frequently outside the watershed of interest. Assessment of the influence of spatially variable rainfall on annual recharge rate estimates was performed by comparing transient simulations using input data from three different rain gauge networks within a coupled and fully-distributed numerical model. A local network of six weather stations with rain gauges was installed and operated in and around the study watershed for three years, and data from six regional stations (within 30 km of the watershed) and one national station (3 km from the watershed) were obtained from publicly available sources. Time series of distributed, daily rainfall were interpolated via the inverse distance squared method using data from each of the rain gauge networks for three calendar years. The temporal and spatial snowfall distribution was consistent among all scenarios, to maintain focus on differences caused by the rainfall input data. Results showed that annual average recharge rates could differ considerably between scenarios, with differences sometimes greater than the water-budget derived uncertainty for recharge. Differences in overall recharge between pairs of scenarios involving the local rain gauge network were largest, varying by up to 141 mm per year, or 44% of the steady state recharge estimated in a previous study. Streamflow estimates for the local rainfall simulations were closer to observations than those using regional or national rainfall. Because the three scenarios used the same set of underlying soil parameters, the results suggest that the availability of local rainfall measurements has the potential to improve the calibration of transient watershed hydrogeological models. The second theme of the present study was exemplified by the Walkerton tragedy in 2000, where pathogenic microbes were rapidly transported from ground surface to a public supply well during a heavy rainfall event. The vulnerability of such wells to surface-originating contaminants during major hydrological events remains poorly understood and is difficult to quantify. Such events may result in overland flow collecting in low topographic locations, leading to localized infiltration. If focused recharge occurs in the immediate vicinity of a public supply well, the threat to the water quality of that well may significantly increase temporarily. These conditions are frequently encountered within the glaciated landscape of southern Ontario. Conventional approaches for defining the threat of groundwater under the direct influence of surface water (GUDI) do not routinely account for this type of transient infiltration event and instead assume steady state flow fields without localized recharge. The present study combined the monitoring and modelling of a site in southern Ontario where DFR is routinely observed to occur within 50 m of a public supply well. Extensive site characterization and hydrologic monitoring were conducted at the site over a period of 3.5 years, specifically during large-magnitude hydrologic events including heavy rainfall and snowmelt. Integrated surface water – groundwater models employing HydroGeoSphere (HGS) were used to quantify the transport of potential contaminants infiltrating beneath a depression and a creek and the associated risk to the public supply well. Simulated relative concentrations at the well were below “detection” for typical median contaminant concentrations in surface water but > 1 cfu/100 mL with travel times between 118 and 142 days for creek and DFR solutes, respectively, based on maximum initial surface water concentrations. Results suggest that DFR and localized recharge could increase the threat to overburden wells under extreme conditions. Ponding reduced travel time by at least 58 days for the DFR solute. In order to extend the analysis of recharge estimate sensitivity to spatial rainfall variability to the longer term, and to incorporate the influence of actual evapotranspiration (AET) uncertainty, a method was developed to employ stochastic rainfall time series and AET estimates in a Monte Carlo framework to quantify the resulting variability in recharge estimates and three groundwater management metrics. Stochastic rainfall time series were generated via a parametric, mixed exponential method for three virtual stations within the Alder Creek watershed and constrained by field-derived spatial correlation coefficients. Observed snowfall data from one nearby national weather station were used to calculate total precipitation. Stochastic annual AET estimates were generated based on: 1) calculated annual potential evapotranspiration at the national weather station, 2) observed variation about the Budyko curve in 45 US MOPEX watersheds with PET/P ratios within ±0.05 of the average ratio calculated for the national weather station near the watershed, and 3) a correction factor to remove AET from the saturated zone. Recharge rates for the Alder Creek watershed were calculated via a 46-year vadose zone water budget for each of 16,778 realizations. The surface water fraction of streamflow was estimated using hydrograph separation results for the watershed. It was hypothesized that spatially variable precipitation would exert more influence on recharge than AET because it is a larger component of the local water budget. Groundwater recharge results were applied to three different metrics related to water quality, well vulnerability, and water quantity. Results suggest that estimates of non-point source contaminant loadings to the water table could differ by up to ±14% from the average. Worst case changes in capture zone area estimates for a public supply well could be up to ±15% different from the average. The ratio of maximum to minimum cumulative recharge over all realizations was 1.31, though contributions from spatial rainfall variability alone led to a ratio of 1.15. This suggests that AET uncertainty and spatial rainfall variability each contribute nearly the same amount of variability to recharge estimates. This latter ratio is less than the result (~2) from a previous study of a much larger watershed in Spain. The results highlight the importance of AET estimates for recharge rate estimation, and their potential impacts on land use planning and groundwater management. This method could be used to project impacts of climate change on recharge variability at the watershed scale. Overall, results suggest that the spatial variability of rainfall could impact recharge rate estimates in numerical models of small to medium sized watersheds (e.g., 78 sq. km), especially during short simulations. Annual recharge estimates could vary over a range equivalent to 44% of a previously estimated steady state value, though long-term (46-yr) estimates could vary over a range equivalent to 12% of this value due to averaging over time. Non-point source loadings and capture zone areas could vary up to ±7.0% and ±7.4% from the average, respectively, over the long term due to spatial rainfall variability, though uncertainties associated with AET could increase this to ±14% or ±15%, respectively. The hydrological event characterization and well vulnerability modelling of the second research theme suggest that localized recharge could lead to increased microbial risks for wells screened in overburden sediments during large hydrological events (≥ 40 mm rainfall over 4 days) through the phenomenon of temporary ponding. The method developed for the long-term stochastic recharge rate analysis could be applied in other settings as an alternative to, or to complement, large-scale, fully-distributed 3D numerical modelling.

Lakes are ecologically, economically, and culturally significant resources that are, at the same time, very fragile and sensitive to human disturbances. During the last decades, intensified urbanization and discharge of nutrients lead to the increase of lake eutrophication in many regions of the world. Moreover, biogeochemical cycles within the lakes are changing due to climate warming, which increases water temperature and affects physical and hydrological lake regimes. In this thesis, I investigate a vast scope of the natural and anthropogenic processes affecting the biogeochemical cycles in lakes at different scales. In particular, I examine the cascading effect of the climate, regional weather, human interventions, and microbial control on phosphorus dynamics in lakes. In Chapter 2, I demonstrate that on the lake scale, phosphorus cycle is driven by internal loading and iron recycling, while it is vulnerable to the reduction of ice cover. To achieve that, I expand the existing MyLake model by incorporating a sediment diagenesis module. Moreover, I develop the continuous reaction network that couples biogeochemical reactions taking place both in water column and sediment. In the modeling scenarios, I assess the importance of the sediment processes and the effects of the climatic and anthro- pogenic drivers on water quality in Lake Vansjø, Norway. I also highlight the importance of phosphorus accumulation within the lake that controls timing and magnitude of bio- geochemical lake responses to external forcing. This includes projected changes in the air temperature, absence of ice cover, and potential management practices. In Chapter 3, I contribute to the long-standing understanding that on the scales of microbial systems, the respiration reactions exert substantial control on biogeochemi- cal cycles by regulating the availability of the electron donors and acceptors, secondary minerals, adsorption sites, and alkalinity. Moreover, I develop a new conceptual model to simulate the preferential catabolic reaction pathways based on power produced in reactions. In contrast to common kinetic rate expressions, I demonstrate that new ther- modynamically based formulations can be applied to describe the microbial respiration of arbitrary large reaction networks. New approach substantially improves the robustness, transferability, and allows the generalization of the model-derived parameters. In Chapter 4, I show that on the regional scale, weather defines hydrodynamic flush rates and water circulation patterns, which, in turn, control the phosphorus transport in Lake Erie, Canada. Specifically, precipitation controls the release of phosphorus from the watershed in the spring, while wind governs the water circulation and transport of the phosphorus released from sediment in the central basin during summer. I also illustrate that climate and weather in the upper Laurentian Great Lakes regulate changes in the water level of Lake Erie. Overall, this thesis improves the fundamental understanding of the phosphorus cycle in lakes, which is being controlled by numerous biogeochemical and physical processes at various scales. In particular, I show that the climate has a cascading effect on the phosphorus cycle in lakes. First, climate controls regional precipitation, wind, and air temperature, which in turn control phosphorus supply from the watershed and basin- wide phosphorus transport. Second, being vulnerable to climate warming, the duration of ice cover impacts the phosphorus cycle through changes in light attenuation, water temperature, mixing regimes, and water column ventilation. Lastly, local environmental perturbations (e.g., pH, temperature, or redox state) define thermodynamic properties of the sediment, which control microbial metabolism and, therefore, the internal phosphorus loading. Finally, this thesis provides new open-source tools for reactive transport simula- tions in lakes as well as in saturated media. In addition to the coupled lake-sediment model developed in Chapter 3, I develop a computer program PorousMediaLab, which performs biogeochemical simulations in water-saturated media and described In Chapter 5. PorousMediaLab is the core component of the numerical investigations presented in the thesis. For example, PorousMediaLab is applied in Chapter 2 to design and test the initial reaction network, calculate fluxes at the sediment-water interface, and estimate re- action timescales. In Chapter 3, PorousMediaLab is used to simulate the reaction rates of batch and one-dimensional sediment column using a novel approach based on the thermo- dynamic switch function. In Chapter 4, PorousMediaLab is used to build a mass balance model and to improve the current understanding of the inter-basin exchange. Both tools are open-source, and they are available online.

This thesis examines the use of Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) for the study of lake ice with a particular focus on the estimation of ice thickness. Experiments were conducted in two lake regions: (1) sub-Arctic lakes located near Yellowknife and Inuvik in the Northwest Territories during March 2017 and 2019, and (2) MacDonald Lake, Haliburton, Ontario, which is known as a mid-latitude lake, during the ice season of 2019-2020. For both regions, GNSS-IR results are compared and validated against in-situ ice and on-ice snow measurements, and also with ice thickness derived from thermodynamic lake ice models. In the first experiment, GNSS antennas were installed directly on the ice surface and the ice thickness at each site was estimated by analyzing the signal-to-noise ratio (SNR) of the reflected GNSS signals. The GNSS-IR capability of ice thickness estimation tested on sub-Arctic lakes results in a root mean square error (RMSE) of 0.07 m, a mean bias error (MBE) of -0.01 m, and a correlation of 0.66. At MacDonald Lake, a GNSS antenna was mounted on a 5-m tower on the shore to collect reflected signals from the lake surface. The Least-Squares Harmonic Estimation (LS-HE) method was applied to retrieve higher SNR frequencies in order to estimate the depths of multiple layers within lake ice and the overlaying snowpack. Promising results were obtained from this experiment; however, ice thickness estimation using GNSS-IR at this mid-latitude lake site was found to be highly dependent on the presence or absence of wet layers such as slush at the snow-ice interface and wet snow above that interface. On colder days, when there was a lower chance for the formation of wet layers, ice thickness could be estimated with a correlation of 0.68, RMSE of 0.07 m, and MBE of -0.02 m. In addition, GNSS-IR showed the potential for determining the freeze-up and break-up timing based on the SNR amplitude of reflected signals. The novel work presented in this thesis points to the potential of using reflected signals acquired by recent (e.g. Cyclone Global Navigation Satellite System (CYGNSS) and TechDemoSat-1 (TDS-1)) and future GNSS-R missions for lake ice investigations.

Increased phosphorus (P) loadings from agricultural runoff into the Great Lakes can lead to eutrophication, resulting in harmful algal blooms and hypoxic conditions. Many studies have demonstrated that subsurface tile drains contribute to total P loss, particularly under no-till. However, most studies have been conducted on soils receiving synthetic fertilizers, and less is understood regarding P loss in tile drains following manure application and if and how tillage and/or manure placement can impact these losses. The goal of this study was to determine if different management practices i.e., conservation till, conventional till, and incorporation, mitigates P loss through tile drains following fall manure application over the Non-Growing Season (NGS). The objectives of this field-based study were to: 1) quantify annual runoff, and P loss from tile drains in a silt loam soil throughout the NGS; 2) investigate if losses differ between conventional and conservation tillage; and 3) determine if incorporation of manure impacts P loss in tile runoff. Tile discharge was monitored from 3 adjacent tile drains with different management treatments (annual till without incorporation, conservation till (with and without manure incorporation) over the span of 8 years (2011-2018), with water samples collected during runoff events for most years during this period. Two years that followed fall manure application (2014-15, 2017-18) were selected for more intensive study. Most P loss occurred during discrete hydrologic events over the NGS, predominantly during the first large discharge event. During this event deep annual tillage increased P loss compared to conservation tillage, with manure incorporation further reducing P loss resulting in differences in cumulative P loss in the tiles over the NGS. This study highlights the importance of in-field long term monitoring in order to capture temporal and spatial variability within a system and recommends that fall manure is incorporated to reduce P losses in tile drains.

Nutrient losses from agricultural operations contributes to the issue of eutrophication of freshwater systems. Although many studies have been conducted on diffuse nutrient losses from fertilizer application, there is a paucity of studies on point source phosphorus (P) loss from bunker silos. Furthermore, the build-up of legacy P in the landscape from historical land management practices can create critical source areas of P that contribute to P loads long after those practices cease. The goal of this thesis is to quantify the contribution of a dairy farm (dominated by bunker silo losses) to watershed P losses, and to monitor P concentrations in surface and groundwater across a riparian zone to characterize the sorption potential of its sediments and infer whether the riparian zone may be acting as a sink for P, or a source of previously retained (legacy) P to the stream. Stream discharge was monitored continuously throughout the study, and automatic water samplers were deployed in the stream above, and below the bunker silo to analyze soluble reactive P (SRP), total dissolved P (TDP), and total P (TP) on an event basis. The riparian zone was equipped with a series of nested wells and piezometers along a three transects to monitor groundwater P levels, and to determine the hydraulic conductivity of the riparian groundwater. A transect was also installed on the unaffected side of the transect as a reference. The farmyard contribution to watershed P losses over a one-year period was 32% (SRP) and 22% (TP). Cumulative loads over the entire study suggest that the farmyard P losses were 21.2 kg/ha SRP and 120 kg/ha TP. Peak P concentrations occurred during snowmelt and thaw events and were smaller during periods of baseflow. However, after the bunker silo was refilled in mid-summer months, both SRP and TP were considerably elevated. Large amounts of P were found to be stored in the riparian soil, however, estimated contributions of riparian P to the overall loads were negligible. This may be a result of missed flowpaths during site set-up, or an occurrence of upwelling of P in the streambed. The results of this research suggest that this particular farmyard bunker silo contributes large amounts of P to the adjacent stream on an annual basis. This study should be used as a starting point for future studies examining livestock farmyard nutrient losses.

Alpine regions contribute 60 % of annual surface runoff, playing an important role in regulating the global water balance. Many of the world’s major river networks originate from alpine headwater basins, popularizing mountains as the “Water Towers of the World”. The Rocky Mountains represent Western Canada’s “Water Tower” since they store and distribute water resources to over 13 million people across Western Canada and the Pacific Northwest USA. At the headwater, topography causes land surfaces to cycle in and out of shadows, creating distinct microclimates that strongly influence evapotranspiration (ET) and carbon fluxes. Yet, relatively few studies have observed the relationship between the energy, water, and carbon fluxes of mountain catchments; and have rather focused on periods of snow and ice cover. Therefore, understanding the contribution of subalpine wetlands to the water budget remains a leading hydrological need in mountain areas worldwide. This thesis attempts to address these knowledge gaps by investigating the influence of complex terrain on the spatial and temporal variability of shade across a subalpine wetland (2,083 m a.s.l.) in the Canadian Rocky Mountains and the effect of shade on seasonal flux dynamics. Meteorological and eddy covariance equipment was installed from June 7th to September 10th to establish baseline environment conditions and to monitor the turbulent and radiative fluxes over the 2018 snow free period. Hill shade and solar radiation models for clear-sky days were compared to field observations to understand how shade impacted the energy, water, and carbon fluxes. Water Use Efficiency (WUE) was used as a metric to understand the relationship between water and carbon cycling. Overall, shade shortened the growing season and prolonged snowmelt. Shade was greatest near the headwall and reduced cumulative solar radiation by 86.4 MJ over the study period. When shade was low and constant during the period of Stable Shade (June 7th – July 30th), it had a non-significant relationship with incoming solar radiation (K↓) and net radiation (Q*); however, when shade rapidly increased during the period of Dynamic Shade (July 31st – September 10th) it strongly influenced K↓ and Q*. On average, during Dynamic Shade, each hourly increase of shade per day, reduced K↓ and Q* by 32 W/m2 and 28 W/m2, equivalent to 13 % and 16 %, respectively. Water and carbon fluxes had a similar response to shade as the energy fluxes. Each hourly increase of shade reduced ET and Gross Primary Production (GPP) by similar margins: 17 % and 15 %, respectively. Therefore, WUE remained relatively unaffected by horizon shade, because shade equally reduced ET and GPP. These findings indicate that under uncertain future climate scenarios (i.e. increased risk of flood, drought, and forest fires), shade may be an important mechanism for moisture conservation in a variety of subalpine ecosystems that are at risk of late season water stress.

Canada’s Rocky Mountains provide a large and essential supply of freshwater to downstream ecosystems and communities. Previous research has demonstrated that warmer temperatures, associated with climate change, are expected to increase the recruitment of trees towards alpine zones, by way of tree islands and krummholz. Tree islands and krummholz are coniferous trees that grow in isolated patches. Tree islands are stunted and deformed, yet their stems grow above the shrub layer, leaving them exposed to winter snowdrifts, unlike krummholz, which grow stunted or in matts, below the snowpack. These trees are unique, relative to conifers below the treeline limit, as they have growth mechanisms which allow them to persist in areas that are otherwise too harsh for full treeline expansion. This thesis addresses the complex relationships between the spatial variability of evapotranspiration (ET) in tree islands and krummholz on a subalpine ridge slope in Kananaskis, Alberta. As well, relationships between these canopies and controls on ET, such as snowcover, meteorological fluxes and vegetation characteristics are assessed. By addressing these objectives, this study will reduce existing knowledge gaps on how forest transition zones in mountain ecosystems may contribute to ecosystem water loss, should these tree patches continue to prosper at higher elevations. Atmometers, which measure the rate of ET from heterogenous landcover to the atmosphere, were used at FRS to determine the rate of potential crop evapotranspiration (ETC) from krummholz and tree islands. ETC was then converted to actual evapotranspiration (ETA) using patch-specific correction coefficients (KC) in order to address the influence of canopy dynamics and water availability on ET. ETA during the growing season was greater in the krummholz (190 mm) than the tree islands (131 mm). Krummholz were observed to be moisture rich tree patches that were shorter in height and more exposed than tree islands. Because of this, krummholz ET was controlled by the advection of sensible heat transported from drier areas downward over the krummholz resulting in oasis-effect ET (QE > Q*). Horizontal advection of sensible heat from the taller tree islands to the shorter krummholz increases clothesline-effect ET at FRS. In addition, the exposure of the krummholz to the effects of solar radiation to the their subsurface increases the rate of early growing season ETA by increasing soil water evaporation. Tree islands, which extend above the annual snowpack were capable of sheltering windblown snow, increasing the amount of water available to the tree islands and krummholz for the growing season. Water balances for the tree islands and krummholz indicated that SWE was the primary source of water to the patches and did not suggest water deficits during the observed growing season. Tree island ET rates were controlled by the evaporation of intercepted precipitation (2 - 58%), and growth characteristic such as increased canopy density, which increased subsurface sheltering, reducing soil water evaporation, while maintaining inner-canopy VPD (increases transpiration). The results of this study improve our knowledge of how tree islands and krummholz will influence ecosystem water storage, especially in terms of ET, and determined what dominant controls exist on ET in subalpine systems. As climate change is expected to decrease annual snowpack levels and increase seasonal air temperatures, ET from tree island and krummholz may contribute to water deficits in subalpine ecosystems.

The spatio-temporal heterogeneity of seasonal snow and its impact on socio-economic and environmental functionality make accurate, real-time estimates of snow water equivalent (SWE) important for hydrological and climatological predictions. Passive microwave remote sensing offers a cost effective, temporally and spatially consistent approach to SWE monitoring at the global to regional scale. However, local scale estimates are subject to large errors given the coarse spatial resolution of passive microwave observations (25 x 25 km). Regression downscaling techniques can be implemented to increase the spatial resolution of gridded datasets with the use of related auxiliary datasets at a finer spatial resolution. These techniques have been successfully implemented to remote sensing datasets such as soil moisture estimates, however, limited work has applied such techniques to snow-related datasets. This thesis focuses on assessing the feasibility of using regression downscaling to increase the spatial resolution of the European Space Agency’s (ESA) Globsnow SWE product in the Red River basin, an agriculturally important region of the northern United States that is widely recognized as a suitable location for passive microwave remote sensing research. Multiple Linear (MLR), Random Forest (RFR) and Geographically Weighted (GWR) regression downscaling techniques were assessed in a closed loop experiment using Snow Data Assimilation System (SNODAS) SWE estimates at a 1 x 1 km spatial resolution. SNODAS SWE data for a 5-year period between 2013-2018 was aggregated to a 25 x 25 km spatial resolution to match Globsnow. The three regression techniques were applied using correlative datasets to downscale the aggregated SNODAS data back to the original 1 x 1 km spatial resolution. By comparing the downscaled SNODAS estimates to the original SNODAS data, it was found that RFR downscaling produced much less variation in downscaled results, and lower RMSE values throughout the study period when compared to MLR and GWR downscaling techniques, indicating it was the optimal downscaling method. RFR downscaling was then implemented on daily Globsnow SWE estimates for the same time period. The downscaled SWE results were evaluated using SNODAS SWE as well as in situ derived SWE estimates from weather stations within the study region. Spatial and temporal errors were assessed using both the SNODAS and in situ reference datasets and overall RMSEs of 21 mm and 37 mm were found, respectively. It was observed that the southern regions of the basin and seasons with higher downscaled SWE estimates were associated with higher errors with overestimation being the most common bias throughout the region. A major contribution of this study is the illustration that RFR downscaling of Globsnow SWE estimates is a feasible approach to understanding the seasonal dynamics of SWE in the Red River basin. This is extremely beneficial for local communities within the basin for flood management and mitigation and water resource management.

Non-point source anthropogenic nutrient loading through intensive farming practices is a global source of water quality degradation by creating harmful algal blooms in aquatic ecosystems. Phosphorus, as the key nutrient in this process, has received much attention in different studies as well as conservation programs aimed at mitigating the transfer of polluting nutrients to freshwater resources. Central to conservation initiatives developed to maintain and improve water quality is the application of the Conservation Practices (CPs), introduced widely as practical, cost-effective measures with overall positive impacts on the rate of nutrient load reductions from farmlands to freshwater resources. Crop rotation is one of the field-based BMPs applied to maintain the overall soil fertility and preventing the displacement of the topsoil layers by surface water runoff across the agricultural watersheds. The underlying concept in the application of this particular BMP is a deviation from the monoculture cropping system by integrating different crops into the farming process. This way, cultivated soils do not lose key nutrients, which are necessary for crop growth, and the overall crop productivity remains unchanged in the landscape. The successful implementation of crop rotation highly depends on planning the rotation process, which is influenced by a variety of environmental, structural, and managerial factors, including the size of farmlands, climate variability, crop type, level of implementation, soil type, and market prices among other factors. Each of these decision variables is subject to variation depending upon the variability of other factors, the complexity of watersheds upon which this BMP is implemented, and the overall objectives of the BMP adoption. This study aims to investigate two of these decision variables and their potential impacts on phosphorus load reductions through a scenario-based hydrologic modeling framework developed to iv assess the post-crop rotation water quality improvements across the Medway Creek Watershed, situated in the Lake Erie Basin in Ontario, Canada. These variables are the spatial pattern of crop rotation and its level of implementation, assessed at the watershed scale through the modifications made to the delineation of the basic Hydrologic Response Units (HRUs) in the modeling process as well as certain assumptions in the management schedules, and decision rules required for the integration of crop rotation into the proposed modeling framework and optimal placement of this non-structural BMP across the watershed. The main modeling package utilized in this study is the Soil and Water Assessment Tool (SWAT), used in conjunction with the ArcGIS and IBMSPSS tools to allow for spatial assessment and statistical analyses of the proposed hydrologic modeling results, respectively. Following in-depth statistical analyses of the scenarios, the results of the study elicit the critical role of both factors by proposing optimal ranges of application on the watershed under study. Accordingly, to achieve optimal implementation results compared to the baseline scenario, which has the zero rate of implementation, conservation initiatives in the watershed are encouraged to consider the targeted placement of crop rotation on half of the lands under cultivation. Despite, having a statistically significant impact on water quality compared to the baseline scenario, the random distribution scenario is less effective than the targeted scenario in mitigation of total phosphorus load. Similarly, compared to the medium rate of implementation the targeted placement in a higher proportion of the cultivated areas did not lead to statistically significant results but may be considered depending upon the purpose and scope of implementation.

Spatial data is characterized by rich contextual information with multiple characteristics at each location. The interpretation of this multifaceted data is an integral part of current technological developments, data rich environments and data driven approaches for solving complex problems. While data availability, exploitation and complexity continue to grow, new technologies, tools and methods continue to evolve in order to meet these demands, including advancing analytical capabilities, as well as the explicit formalization of geographic knowledge. In spite of these developments Discrete Global Grid Systems (DGGS) were proposed as a new comprehensive approach for transforming scientific data of various sources, types and qualities into one integrated environment. The DGGS framework was developed as the global data model and standard for efficient storage, analysis and visualization of spatial information via a discrete hierarchy of equal area cells at various spatial resolutions. Each DGGS cell is the explicit representation of the Earth surface, which can store multiple data values and be conveniently recognized and identified within the hierarchy of the DGGS system. A detailed evaluation of some notable DGGS implementations in this research indicates great prospects and flexibility in performing essential data management operations, including spatial analysis and visualization. Yet they fall short in recognizing interactivity between system components and their visualization, nor providing advanced data friendly techniques. To address these limitations and promote further theoretical advancement of DGGS, this research suggests the use of Q-analysis theory as a way to utilize the potential of the hierarchical DGGS data model via the tools of simplicial complexes and algebraic topology. As a proof of concept and demonstration of Q-analysis feasibility, the method has been applied in a water quality and water health study, the interpretation of which has revealed much contextual information about the behaviour of the water network, the spread of pollution and chain affects. It is concluded that the use of Q-analysis indeed contributes to the further advancement and development of DGGS as a data rich framework for formalizing multilevel data systems and for the exploration of new data driven and data friendly approaches to close the gap between knowledge and data complexity.

Background: In the subarctic Dehcho region of the Northwest Territories, many remote communities rely on traditional foods, including fish, to supplement more expensive store-bought options. Fish are an excellent source of omega-3 and omega-6 polyunsaturated fatty acids (n-3 and n-6 PUFAs, respectively), essential compounds that can only be obtained through the diet. Long-chain n-3 PUFAs, such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are especially important for human health. As the health benefits derived from consuming fish can be diminished by the risk imposed by exposure to contaminants, such as mercury, researchers and communities in the Dehcho region began a collaborative project in 2012 to quantify both fatty acid and mercury concentrations in fish. In the course of this work, it was found that concentrations of fatty acids in fish differed significantly among lakes in the Dehcho region. In freshwater ecosystems, fatty acids are produced by algae and bacteria and transferred up the food chain through consumption. The type and quality of fatty acids produced varies among primary producer taxa, meaning that fatty acid profiles in fish may vary among lakes due to variation in the composition of algal and bacterial communities, which in turn vary in response to abiotic conditions in lakes. Objectives: As some fish samples were stored for multiple years before processing, the first objective of this study was to determine if there was a relationship between concentrations of fatty acids and storage time at -20 degrees C. After determining which fatty acids were affected by storage time and how they were affected by storage time, the second objective was to update existing fish fatty acid profiles (analysed from samples collected 2013-2015) for the study lakes. The third objective was to determine whether there were differences in concentrations among lakes for several fatty acid groups of interest, including total fatty acids (TFA), n-3 and n-6 PUFAs, DHA, and EPA, and whether observed differences in fish fatty acid profiles could be explained by water chemistry and/or watershed characteristics among lakes. Methods: A total of 433 fish, including Burbot (Lota lota), Cisco (Coregonus artedi), Lake Trout (Salvelinus namaycush), Longnose Sucker (Catastomus catastomus), Lake Whitefish (Coregonus clupeaformis), Northern Pike (Esox lucius), Walleye (Sander vitreus), and White Sucker (Catastomus commersoni) were captured in 10 important subsistence lakes within the Dehcho region between the years of 2013 and 2018. Sampled lakes were located in three different eco-zones, the Hay River Lowlands, the Horn Plateau, and the Northern Alberta Uplands. Fish muscle tissue was frozen on-site and transported back to the University of Waterloo for laboratory analysis of both fatty acid and mercury concentrations. Water samples were collected at each lake to characterise lake chemistry (e.g. major nutrients, ions, dissolved organic carbon, etc.), and these data were compared to an existing dataset on watershed characteristics (e.g. lake area, watershed area, etc.). Results: In every fish species, DHA concentrations decreased exponentially with increasing storage time, while C:24:0, a saturated fatty acid, increased significantly with increasing storage time. Updated fish fatty acid profiles and mercury concentrations confirmed results found by Reyes et al (2017) and Laird et al (2018); Cisco, Lake Whitefish, Longnose Sucker, and White Sucker are the fish species with the highest fatty acid concentrations and lowest mercury concentrations. Concentrations of all fatty acid groups examined in Northern Pike were statistically different among lakes (TFA, n-3 and n-6 PUFAs, EPA, and DHA), while only some fatty acid groups in Lake Whitefish (TFA, n-6 PUFAs, and DHA) and Walleye (n-3 and n-6 PUFAs) varied significantly among lakes. Significant predictors of concentrations of fish fatty acids included both water chemistry and watershed characteristics, and fell into 3 distinct groups of variables: lake productivity (total phosphorus), indicators of carbon quality (UV254, specific UV absorbance, dissolved organic carbon, and total nitrogen), and catchment influence (chloride concentrations, calcium concentrations, and the ratio of lake perimeter to watershed area). Understanding factors that lead to variation in concentrations of fish fatty acids, both among lakes and because of storage practices, can inform predictions of the nutritional value of fish in other lakes, provide a baseline for assessing ongoing effects of climate-induced change, and allow community members to make informed choices about the fish that they are eating.

Lake Erie’s commercial and recreational walleye fishery is the largest of the Great Lakes, requiring effective management to maintain a sustainable and complex fishery. Lake Erie’s walleye fishery is composed of multiple spawning populations, which presents a management challenge. The movement patterns and recruitment of distinct walleye populations that make up the fishery must be considered by managers to avoid overexploitation and to maintain population diversity. The Grand River walleye population in Lake Erie’s eastern basin is considered a priority for rehabilitation due to blocked access to spawning habitat by a low-head dam and degraded habitat quality. The objectives of this study were to: i) investigate movement patterns of spawning walleye in the Grand River using acoustic telemetry; and, ii) investigate movement and habitat use of young-of-the-year (YOY) walleye in relation to the Dunnville Dam and surrounding habitat segments using stable isotope analysis. Between 2015 and 2018, 267 mature walleye were tracked in the Grand River using acoustic telemetry, and in fall of 2018 144 YOY walleye were sampled from the river via boat-mounted electrofishing. Both male and female mature walleye that were moved upstream of the Dunnville dam were found to actively migrate ~20-40 km up-river to areas with suspected suitable spawning substrate during the spring spawning season. Residence time of walleye above the Dunnville Dam and timing of return migrations suggest that the dam may be acting as an impediment to downstream movement. Of all the walleye tagged, 43% returned to the Grand River during at least one year subsequent to the initial spawning season during which they were tagged, and those that returned were detected at spawning habitat below the Dunnville Dam during March and April. Although differences in YOY walleye stable isotope signatures (carbon and nitrogen) were evident across sampling locations in the Grand River in fall of 2018, YOY walleye were not successfully sampled in 2019 and a description of the trophic baseline was needed to infer YOY walleye movements. Condition of YOY walleye sampled during the fall of 2018 was highest at the river mouth, which may indicate relatively favourable health conditions for YOY walleye at this location. The results of the biotelemetry study suggest that the removal of the Dunnville Dam or the construction of a functional fishway would increase access to potential additional spawning habitat, which may lead to an increase in successful spawning activity for the Grand River walleye population. Future research on YOY walleye in the southern Grand River will be necessary to enhance the understanding of how recruitment and year-class strength is impacted by movement barriers (i.e., Dunnville Dam) and variation in spawning and nursery habitat quality (i.e., abiotic and biotic stressors). Furthermore, additional analyses on mature walleye apparent annual survival and spawning site fidelity probabilities would further inform our understanding of Grand River walleye movement and support walleye management in Lake Erie’s eastern basin.

As wetlands around the world are being lost, policies are implemented to help protect further destruction and loss of valuable services that wetlands provide. In Alberta, wetland policy has been put in place with the goals of protecting the most valuable wetlands and replacing necessary loss of wetlands to maintain functional value. To help the policy meet its objectives, the Alberta Wetland Rapid Evaluation Tool-Actual (ABWRET-A) was developed and implemented in Alberta’s settled area in 2015 as a standardized way to give a value score via functional assessment to any wetland in the province, with the hopes that the most valuable wetlands will be conserved. These assessment tools are in constant need of review and improvement to make sure they are helping meet policy goals. I assess biases made in the selection for ABWRET-A calibration wetlands and determine how these biases affect ABWRET-A scoring to determine if subsequent scores provided by this tool are over or under estimating wetland value. I also assess the wetlands that underwent ABWRET-A evaluation and were drained or filled in under a permit in the 1.5 yr after ABWRET-A implementation in Alberta’s settled region to determine whether they mirror the calibration wetlands. I found that the calibration dataset comprised larger, more permanently ponded wetlands distributed closer to roads than the general wetland population. I also found that the calibration dataset included fewer bogs and more fens. I found that larger wetlands and wetlands classified as fens received higher ABWRET-A scores, whereas wetlands close to roads received lower scores. Consequently, I surmise that the scores being given out since ABWRET-A’s implementation are likely underestimates. This is corroborated by a lower distribution of scores in the wetlands permitted for drainage than policy recommends. The wetlands being targeted for permitted loss were also smaller, more road-proximate, and concentrated around major cities, implying permanent regional loss of those wetlands and their functions. Based on these findings, I make suggestions for improving ABWRET-A, including adding calibration sites to better capture the natural variability of wetlands in the area to improve ABWRET-A’s accuracy in estimating relative wetland value.

In agricultural watersheds across the world, decades of commercial fertilizer application and intensive livestock production have led to elevated stream nutrient levels and problems of eutrophication in both inland and coastal waters. Despite widespread implementation of a range of strategies to reduce nutrient export to receiving water bodies, expected improvements in water quality have often not been observed. It is increasingly understood that long time lags to seeing reductions in stream nutrient concentrations can result from the existence of legacy nutrient stores within the landscape. However, it is less understood how spatial heterogeneity in legacy nutrient dynamics might allow us to target implementation of appropriate management practices. In this thesis, we have explored the dominant controls of legacy nitrogen accumulation in a predominantly agricultural 6000-km2 mixed-landuse watershed. First, we synthesized a 216 year (1800 – 2016) nitrogen (N) mass balance trajectory at the subbasin scale accounting for inputs from population, agriculture, and atmospheric data, and output from crop production using a combination of census data, satellite imagery data, and existing model estimates. Using these data, we calculated the N surplus, defined as the difference between inputs to the soil surface from manure application, atmospheric deposition, fertilizer application, and biological N fixation, and outputs primarily from crop production. We then used the ELEMeNT-N model, with the estimates of the N mass balance components as the model inputs, to quantify legacy accumulation in the groundwater and soil in the study basin and 13 of its subbasins. Our results showed that from 1950, N surplus across the study site rose dramatically and plateaued in 1980. Agricultural inputs from fertilizer and biological nitrogen fixation were the dominant drivers of N surplus magnitude in all areas of the watershed. Model results revealed that 40% of the N surplus to the watershed since 1940 is stored as legacy N, and that the proportion of N surplus that is stored as legacy vary across the watershed, ranging from 33% to 69%. Where legacy tends to accumulate also varies across the watershed, ranging from 49% - 72% stored in soil, and 28% - 51% stored in groundwater. Through correlation analysis, we found that soil N accumulation tends to occur where there is high agricultural N surplus, and groundwater N accumulation tends to occur where mean groundwater travel times are long. We also found that using the model calibrated mean groundwater travel times as an indication of lag times, we can identify the length of lag time in various regions in the watershed to help inform long-term management plans. Our modeling framework provides a way forward for the design of more targeted approaches to water quality management.

A data driven approach was used in this study to investigate the drivers of nutrient water quality across the Laurentian Great Lakes drainage basin. Monitored time series of nutrient water quality and discharge were modelled using a dynamic regression-based model. Random forest machine learning was used as a framework to assess drivers of nutrient water quality, using mean annual flow-weighted concentrations (FWCs) and ratios calculated from modelled water quality, combined with spatial factors from monitored watersheds. Analysis revealed that landscape variables of developed land use, tile drained land, and wetland area played important roles in controlling nitrate and nitrite (DIN) and soluble reactive phosphorus (SRP) FWCs, while soil type and wetland area was important for controlling particulate phosphorus (PP) FWCs. Fertilizer and manure practices were important controls in nutrient ratios of SRP:Total Phosphorus (TP), and DIN:TP, with developed land use, manure application, and tile drained land important for the former, and developed land use and manure application (vs synthetic fertilizer application) important for the latter. Plots of feature contribution were generated to isolate the effect that spatial variables had in machine learning models and revealed underlying behaviour of important controls in driving nutrient water quality across the basin. Random forest models were further developed to predict FWCs and ratios of nutrients across all watersheds within the Great Lakes drainage basin. Modelled results revealed hot spots of high DIN, SRP and PP in the watersheds along the southeastern shores of Lake Huron, on the eastern watersheds of the Huron-Erie corridor, and in the southwestern watersheds of Lake Erie. High SRP:TP ratio hot spots were seen in watersheds along the southeastern shores of Lake Huron and along the eastern side of the Huron-Erie corridor. Hot spots of low DIN:TP ratios with high nutrient export were seen in the southwestern watersheds of Lake Erie, which has implications for harmful algal growth. Nutrient ratios across the Great Lakes watersheds compared similarly to other heavily human impacted catchments of the Baltic Sea and western Europe. Annual basin loads of DIN, SRP, and TP were estimated from random forest models for each year from 2000-2016. Calculated annual nutrient loadings of SRP and TP were consistent with other published values of Great Lakes watershed estimates and revealed highest loadings during 2011 when the largest recorded algal bloom in Lake Erie occurred to date. Overall, this data-driven analysis of nutrient water quality reinforces and refines our process understanding of nutrient pollution dynamics across the Great Lakes drainage basin.

Interest in numerical modeling of permafrost has increased over the past decade due to accelerating rates of permafrost thaw. Discontinuous permafrost regions are particularly susceptible to climate change since small increases in mean annual temperature may lead to significant permafrost thaw, landscape change, changes in hydrologic connectivity, and greenhouse gas (GHG) emissions. The influence of local heterogeneities on the short- and long-term evolution of permafrost bodies is poorly understood. In order to numerically simulate the freeze-thaw processes in heterogenous media, a robust numerical model is desirable to overcome potential instabilities induced by heterogeneity in soil thermal properties. Here, such a model is developed, supplemented by a careful evaluation of the impact of heterogeneity upon the soil freezing curve, and applied to investigate the influence of local heterogeneity upon discontinuous permafrost evolution. Numerical models of permafrost evolution in porous media typically rely upon a smooth continuous relation between pore ice saturation and sub-freezing temperature, rather than the abrupt phase change that occurs in pure media. Soil scientists have known for decades that this function, known as the soil freezing curve (SFC), is related to the soil-water-characteristic-curve (SWCC) for unfrozen soils due to the analogous capillary and sorptive effects experienced during both soil freezing and drying. Herein we demonstrate that other factors beyond the SFC-SWCC relationship can influence the potential range over which pore water phase change occurs. In particular, we provide a theoretical extension for the functional form of the SFC based upon the presence of spatial heterogeneity in both soil thermal conductivity and the freezing point depression of water. We infer the functional form of the SFC from many abrupt-interface 1-D numerical simulations of heterogeneous systems with prescribed statistical distributions of water and soil properties. The proposed SFC paradigm extension has the appealing features that it (1) is determinable from measurable soil and water properties, (2) collapses into an abrupt phase transition for homogeneous media, (3) describes a wide range of heterogeneity within a single functional expression, and (4) replicates the observed hysteretic behavior of freeze-thaw cycles in soils. SFCs are used in all the permafrost models that use a continuum phase-change criterion. Here, an efficient enthalpy-based continuum numerical approach is introduced for solving heat transfer problems with non-isothermal phase change. In order to simulate permafrost over time spans of several years, a robust and efficient model is required. In the present setting, the heat transfer problem is converted to a minimization problem, in which we minimize a potential function that characterizes the governing heat transfer PDE within a time discrete framework. The use of the trust region minimization algorithm proves desirable due to the highly nonlinear energy functional which also involves non-convex terms induced by phase change. Results obtained show satisfactory agreement with existing analytical solutions. Moreover, the grid and timestep convergence studies conducted to examine the dependence of the solution on mesh and timestep sizes indicate robust convergence rates. This is the first application of trust region energy minimization algorithm in permafrost simulation. The two-dimensional enthalpy-based numerical model with continuum phase-change is applied to study the effect of heterogeneity in the soil freezing point on conduction-driven talik formation. This model is rigorously verified against Lunardini's solution (Lunardini 1981), which is an analytical solution of the Stefan Problem with a non-isothermal phase-change criterion. Stochastic realizations of spatially correlated distributed soil thermal parameter fields are generated using the geostatistical software library (GSLIB) for a variety of correlation lengths and variances in material properties. These are used as input to the 2-D permafrost model under fully saturated conditions. The simulation results indicate that local heterogeneities have conditional effects on the formation of unfrozen zones and, eventually, talik. This influence is exacerbated under the presence of advective heat transfer, where small perturbations to the liquid water saturation can lead to preferential flow conduits. This work is extended by conducting sensitivity analysis to study the relative dependence of talik formation on different sources of heterogeneity (e.g. soil density and boundary conditions).

The installation of tile drainage systems in Southern Manitoba has been accelerating over the past two decades to improve crop production. Given current environmental and political concerns related to agricultural pollution and the eutrophication of Lake Winnipeg, the role that tiles may play in both runoff and nutrient loading from agricultural fields must be evaluated because tiles can also have environmental consequences due to their capacity to export significant quantities of pollutants such as phosphorus (P) and nitrogen (N) from croplands by acting as subsurface lateral conduit pathways. This study examined surface and subsurface runoff from tiled and non-tiled fields on a farm in Elm Creek, Manitoba from 2015 to 2017 to quantify edge of field runoff and nutrient losses, to characterize surface-tile connectivity through the vadose zone, and to characterize ditch-overland flow dynamics at the edge-of-field. Water samples were collected from field surfaces, tile drainage, groundwater and roadside ditches during runoff events that occurred throughout the open water season. In addition, soil samples were collected in 2017 and analyzed for inorganic P fractions and P availability. This thesis has shown that overland flow was the major pathway for runoff and nutrient (P and N) edge of field losses, and the presence of the tile drainage did not decrease the frequent occurrence of the overland flow due to the prevailing climate conditions and vertisolic clays in the Red River Valley. Tile drains were responsible for 11-28% annual runoff losses, < 5% annual P losses and 40-50% annual nitrate N losses. Thus, although tile drainage did not exacerbate the edge of field P losses, it has the potential elevate N losses. Tile drainage was often activated from top-down water front movement and tile flow activation was hastened by higher rainfall intensities and wetter antecedent moisture conditions. Significant tile drainage predominantly occurred in late spring under wet antecedent conditions when the water table was elevated. During such periods, the chemistry of tile drain effluent was similar to that of groundwater, which was low in P and high in N. In contrast, tile drainage in both early spring (snowmelt) and summer was small, although for different reasons. During snowmelt, when most runoff occurs in the Prairies, tile drainage was impeded by the presence of frozen ground and most runoff left fields as overland flow. Tile chemistry during this period reflected surface runoff, which was rich in P, indicating the presence of preferential flow through frozen ground. The chemistry of tile drainage was also rich in P and reflected surface runoff in summer, when rain fell on dry soils, also indicating preferential flow. Thus, although preferential flow between the surface and tile drains appears to have occurred in the vertisolic clays of the Red River Valley, it was associated with very small flow volumes and therefore small loads, whereas tile drain chemistry during periods when the majority of tile flow occurred resembled that of groundwater. This thesis has shown that tile drains will do little to modify water volume or chemistry during the snowmelt period, which dominates annual water cycles, due to the presence of frozen ground, and surface runoff will remain the greatest source of P loss from agricultural fields. This thesis has shown that some of the P loss from fields is due to direct losses from fields, but some may be mobilized during flooding due to water backing up in roadside ditches during snowmelt runoff, spring storms and massive thunderstorms. This suggests that an improved understanding of the role of ditch management on agricultural P loss is needed. This thesis has produced a comprehensive view of edge-of-field and in-field hydrochemical losses in tile drained fields in the Red River Valley. The outcomes of this thesis have implications for both water and nutrient management perspectives for farmers and policymakers.

In cold regions, the breakup of river ice can be a significant event, resulting in flooding and damage to communities. Given the severity of such events, it is desirable to be able to predict the timing and severity of breakup. Limited progress has been made on forecasting breakup related flooding as no deterministic model of the breakup process and ice jam formation exist. Current tools for predicting breakup rely on developing a relationship between the previous winter conditions and the current spring conditions, with the understanding that a rapid or large runoff with a thick ice cover has the potential for a more severe breakup than if ice has had time to melt. These tools are largely empirical, statistical, or soft computing methods which rely on historical data sets of discrete observations to relate the complex relationship between climate and hydrology to breakup conditions and are limited by access to the extensive data required. Within the current prediction methods, the application of hydrological models for forecasting breakup timing and severity is limited. Hydrological models can address some of the limitations of current tools, as they are able to simulate the complex relationships between climate and hydrology which has a strong influence on the breakup period. Additionally, hydrological models may be more practical in regions with limited data, as they can simulate variables of interest instead of relying on large historical data sets. This thesis demonstrates how a hydrological model can be used to predict the timing and severity of breakup, through the coupling of a 1D river ice model with a hydrological model. Emphasis is placed on the development of the hydrological model to ensure that it provides realistic results throughout the basin. The Liard basin, a large relatively data sparse river basin, in northern Canada is used as a case study. A thorough calibration strategy, based on an iterative, multi-objective approach is used in the development of the model. The final model exhibits strong performance in both calibration and validation throughout the basin. A simple 1D river ice model in MATLAB is coupled with the hydrological model. The hydrological model can forecast the timing of breakup well based on the timing of the initial rise in the hydrograph. Breakup severity is predicted using a simple threshold model based on ice thickness, flow, and accumulated shortwave radiation. The prediction method was applied to an independent location as verification of the methodology with promising results.

Flows of reactive nitrogen (N) have significantly increased over the last century, corresponding to increases in the global population. The pressures on the N cycle include human waste, fossil fuel combustion as well as increasing food production (i.e., increasing fertilizer consumption, biological N fixation, and livestock manure production). The result is humans causing a 10-fold increase in the flow of reactive N globally. The influx of anthropogenic N into aquatic environments degrades water quality, alters fresh and saline ecosystem productivity, and poses an increasing threat to drinking water sources. In the U.S., decades of persistent hypoxic zones, created by elevated concentrations of nitrate from the landscape, have altered ecosystem trophic structure and productivity. Additionally, increasing N contamination of groundwater aquifers places over 20% of the U.S. population at increased risk of diseases and cancers. Despite billions of dollars of investment in watershed conservation measures, we have not seen proportional improvements in water quality. It has been argued that delayed improvements in water quality can be attributed to legacy stores of N, which has accumulated in the landscape over many decades. There is considerable uncertainty associated with the fate of N in the landscape; however, studies quantified increasing stores of N in the subsurface, suggesting increasing stores of N in groundwater aquifers, in soil organic nitrogen pools, and the unsaturated zone. Nevertheless, the spatial distribution of legacy N across the conterminous U.S. is poorly quantified. Here, we have synthesized population, agricultural, and atmospheric deposition data to develop a comprehensive, 88-year (1930 to 2017) dataset of county-scale N surplus trajectories for the U.S. N surplus, defined as the difference between N inputs and usable N outputs (crop harvest), provides insight into the trends and spatial distribution of excess N in the landscape and an upper bound on the magnitude of legacy N accumulation. Our results show that the spatial pattern of N surplus has changed drastically over the 88-year study period. In the 1930s, the N inputs were more or less uniformly distributed across the U.S., resulting in a few hotspots of N surplus. The following decades had sharp increases in N surplus, driven by the exponential use of fertilizer and combustion of fossil fuels. Contemporary N surplus distribution resembles a mosaic of varying degrees of excess, concentrated in the heavily cultivated areas. To understand dominant modes of behavior, we used a machine learning algorithm to characterize N surplus trajectories as a function of both surplus magnitudes and the dominant N inputs. We find ten primary clusters, three in crop dominated landscapes, four in livestock dominated landscapes, two in urban dominated landscapes, and one in areas minimally impacted by humans. Using the typologies generated can facilitate nutrient management decisions. For example, watersheds containing urban clusters would benefit from wastewater treatment plant upgrades. In contrast, those dominated by livestock clusters would have more success in managing nutrients by implementing manure management programs. The estimates of cumulative agricultural N surplus in the landscape highlights agronomic regions that are at risk of large stores of legacy N, possibly leading to groundwater and surface water contamination. In these agronomic regions, the average cumulative N surplus exceeds 1200 kg-N/ha by 2017. Despite having minimal agricultural activity in urban areas, urban fertilizer use has led to an average cumulative N surplus of over 900 kg-N/ha. While our estimates are an upper bound to legacy stores, significant uncertainty remains regarding the magnitude of the estimate of N accumulation. However, our results suggest that legacy N is at varying degrees, impacting most counties in the U.S. The significant investment and corresponding lack of returns can lead to disillusionment in farmers, watershed managers, and the general public. Developing such N surplus typologies helps improve understanding of long-term N dynamics. Beyond refining the supporting science, appropriately communicating uncertainties and limitations of water quality improvements to the stakeholders, authorities, and policymakers are essential to continuing efforts to improve national water quality.

It has been predicted that approximately 65% of the developing world and 85% of the developed world will be living in cities by 2050. Toronto, the largest city in Canada and the fourth largest in North America, is expected to double in population in the next 50 years. Although such rapid urbanization can lead to enormous social, economic, and environmental change, little is understood about how population growth in Toronto and the “Golden Horseshoe” region around Lake Ontario will impact the ecological systems of Southern Ontario. In our study, we are particularly interested in the ways in which increasing population densities in the Greater Toronto Area are impacting nutrient flows across Southern Ontario’s urban/rural continuum and how changing nutrient dynamics may lead to increasingly impaired water quality in Lake Ontario and beyond. In this work, we utilize a mass balance approach to quantify the flow of nutrients through urban, suburban, and agricultural areas of the Greater Toronto Area. A wide range of factors are considered, including human behaviour, domestic animals, stormwater management, and wastewater treatment processes. The present results suggest that any study of urban metabolism must take into account not only nutrient flows within urban boundaries, but must also identify externalities of urban development associated with a range of processes, from global trade to regional waste management.

Intensification of farming operations and increased nutrient application rates have led to higher crop yields and greater food security. At the same time, widespread use of commercial nitrogen (N) and phosphorus (P) fertilizers and large-scale livestock production have led to unintended environmental consequences, including eutrophication of both coastal and inland waters, threats to drinking water, and increased production of N2O, a potent greenhouse gas. In the past, crop and livestock production were typically more integrated, allowing most livestock to be fed by local crops, and most livestock manure to be applied directly to nearby cropland. Under current intensive agriculture practices, however, there is frequently a spatial decoupling of crops and livestock, leading to hot spots of manure production and a lack of opportunities for cost-efficient and environmentally sensitive disposal. In recent years, there has also been increased interest in the use of both farm and regional-scale bioreactors to convert excess manure to energy, thus exploiting a renewable energy source and increasing the potential to recycle animal waste. In the present work, I develop a spatially distributed optimization approach to identify hotspots of manure production, and, using both economic and environmental criteria, evaluate the economic feasibility of (1) transporting manure for spreading on cropland to meet established nutrient requirements, and (2) constructing biogas reactors to process excess manure in areas where long-range transport is found to be infeasible. This work is focused on manure redistribution, and potential for biogas construction at the continental US scale. In order to identify the spatial disconnect between livestock and crop production, I developed a gridded data set where each cell was 6 km x 6 km and calculated the crop requirements and manure production in each cell. After finding the P requirements in each cell, I found that 530,000 tonnes of phosphorus in manure was located in areas where, if applied, it would be in excess of the local crop requirements. I then examined the feasibility of transporting manure from excess locations (cells) to other locations to use as fertilizer by formulating an optimization problem to maximize the financial benefits of transporting the manure. Savings from transporting manure was calculated as the financial benefit from buying less mineral fertilizer minus the cost of transporting the manure. The solution to this optimization problem shows that transporting manure was able to reduce the excess phosphorus applied to fields by at least 88% with savings of up to $3 billion USD. Finally, I examined the costs and benefits of using the remaining excess manure (after transportation for fertilizer) as fuel to operate biogas plants. For this, I formulated an optimization model to site biogas plants across the continental US such that net profits from the biogas plants were maximized. Biogas net profits were defined as the money made from selling electricity minus the annualized costs for constructing and operating the biogas plants and transporting the manure to the biogas plants. The solution to this problem shows that constructing and operating 387 biogas plants yielded a net profit of $100 million USD and would utilize all of the manure remaining after transportation for fertilizer. This 100% utilization rate of excess manure would have great environmental benefits in terms of removing potential sources of non-point source pollution from farms that would otherwise be available to runoff into waterways.

Internal phosphorus loads, from bottom sediments to surface waters, are often comparable in magnitude to external phosphorus loads, particularly in water bodies with a history of high external phosphorus inputs from point and non-point sources. The benthic release of phosphorus can be influenced by several factors including pH, redox potential, temperature, microbial activity and the concentration of competitive anions at or near the sediment-water interface. Dissolved silicate occurs ubiquitously in natural waters and may act as a competitive ion to phosphate. Nonetheless, prior to the work in this thesis, the effect of silicate on internal phosphorus loading remained poorly understood. This thesis addresses several of the mechanisms through which silicate may influence the mobilization of aqueous phosphate from sediments in aquatic environments. The thesis starts with a thorough literature review of phosphorus biogeochemical cycling in relation to eutrophication, sediment-surface water interactions, mineralogy, competitive anions and microbial activity (Chapter 1). Next, adsorption/desorption of phosphate on/from goethite, a model ferric (hydr)oxide mineral, is investigated in the absence and presence of dissolved silicate. The influence of dissolved silicate on phosphate adsorption is evaluated through laboratory experiments and application of the CD-MUSIC model (Chapter 2). The results show that increasing concentrations of silicate decrease phosphate adsorption, leaving more phosphate in the aqueous phase. The competitive effect of dissolved silicate is more pronounced under alkaline conditions. Subsequently, phosphate desorption experiments were conducted under dynamic pH conditions in the presence and absence of silicate (Chapter 3). The experimental results show that the gradual transition from acidic to alkaline conditions induces the desorption of phosphate adsorbed to goethite under acidic conditions. The presence of silicate in the phosphate/goethite system does not affect phosphate desorption, because of the stronger surface complexation of phosphate to goethite. In addition to adsorption and desorption processes, the co-precipitation of phosphate with iron and the potential subsequent dissolution of these co-precipitates as a result of changing physico-chemical conditions may also control the mobility of phosphate in aquatic environments. The effects of dissolved silicate on the co-precipitation of phosphate with iron and the reactivity of the resulting solids are examined (Chapter 4). Ferric (co)-precipitates (i.e., Fe-P-Si) with variable Si:Fe ratios, were synthesized either via oxidation of Fe2+(aq) or by increasing the pH of Fe3+(aq) solution. The solids were characterized by a combination of chemical and spectroscopic techniques including attenuated total internal reflectance Fourier transform infrared spectroscopy (ATR-FTIR) and X-ray powder diffraction spectroscopy (XRD). Similar solid phase P:Fe ratios were found in co-precipitates formed from solutions with different dissolved silicate concentrations, regardless of the method of preparation. This suggests that the interactions between phosphate and iron during co-precipitation were not affected by dissolved silicate. The ferric (co)-precipitates were subsequently reductively dissolved abiotically in buffered ascorbate-citrate solution to determine their reactivity under reducing conditions. The kinetic data show that the co-precipitates with higher Si:Fe ratios were more recalcitrant to dissolution. For co-precipitates synthesized via oxidation of Fe2+(aq), reductive dissolution experiments were also conducted in the presence of the facultative anaerobic iron reducing bacteria Shewanella putrefaciens. XRD analyses of the residual solids imply that solids with the higher Si:Fe ratios may be more resistant to microbially mediated reductive dissolution. The relative reactivities of the co-precipitates obtained by the two synthesis methods are also addressed in Chapter 4. In Chapter 5, the effect of silicate on the mobility of phosphate in a natural sediment was evaluated via flow-through experiments. The results show that dissolved silicate enhances the mobility of phosphate at the sediment-water interface. Ferric (co)-precipitates were formed at the oxic surface of sediment columns via the oxidation of ferrous iron supplied with upflowing solutions containing variable silicate concentrations. The subsequent dissolution of these co-precipitates under imposed anoxic conditions at the sediment-water interface indicates that the co-precipitates formed at higher dissolved silicate concentrations were more reactive towards reductive dissolution. These results are therefore in apparent contradiction to those observed in Chapter 4. The ferric (co)-precipitates (i.e., Fe-P-Si) evaluated in Chapter 4 were prepared from solutions containing high concentrations of iron, phosphate and silicate, by imposing either rapid aeration or pH increase. These conditions were selected to maximize the yield from the syntheses. The synthesis methods in Chapter 4 are therefore most representative of aquatic environments where co-precipitation occurs rapidly (e.g., groundwater springs) and the concentrations of these dissolved constituents are fairly high. However, in many other aquatic environments, the diffusion-controlled release of Fe2+(aq) from the deeper sediments results in the gradual oxidation of Fe2+ at the sediment-water interface under oxic conditions. This process is typical in lake sediments with minimal advective exchange between surface water and groundwater. This gradual oxidation (at relatively low concentrations of Fe2+) results in the slow formation of ferric (co)-precipitates which may be dissimilar to those synthesized herein and discussed in Chapter 4. The ferric (co)-precipitates synthesized with the flow-through column system in Chapter 5 may be better analogues of slow forming co-precipitates in diffusion dominated or moderately advection influenced aquatic sediments than those synthesized in Chapter 4. Finally, to elucidate the likely importance of the various influences of dissolved silicate on phosphate mobility investigated in this thesis, concentrations of dissolved phosphate and silicate as well as pH data are extracted from the US National Water Information System (NWIS) network (data shown in Chapter 1 and Chapter 2). The NWIS data along with combined experimental and modeling results suggest that silicate-mediated phosphate mobilization is likely a commonly occurring process at the sediment-water interface of lakes and reservoirs. This thesis also demonstrates the multiple roles of silicate on the mobilization of phosphate in aquatic environments, and improves our fundamental knowledge of iron, phosphorus and silicon cycling in freshwater environments.

Excess phosphorus (P) loading increases the frequency of harmful algal blooms (HABs), posing severe threats to drinking water security and aquatic ecosystems. Efforts to reduce the inputs of P to Canadian agricultural soils started in the late 1970s-early 1980s, and were initially successful, but surface water P loading became persistent again in the 2000s. HABs were a problem in the southern Laurentian Great Lakes (LGL) before the initial nutrient mitigation efforts, and the re-emergence of HABs in Lake Erie in the 2000s was in part a result of legacy P that had accumulated in soils and groundwater in agricultural watersheds. Legacy P exists as a result of historical inputs of P, typically fertilizer used in excess of crop needs. Consequentially, even after reducing P inputs, legacy P continues to be exported from soils after several decades. In Chapter 2, a large-scale mass balance was conducted for the Ontario watersheds to locate and quantify agricultural and other anthropogenic P inputs from 1961 to 2016, utilizing existing datasets as well as historical reconstructions of P inputs to the landscape. This scale of P mass-balance has not been completed before in Ontario. The mass balance model was implemented into a Geographical Information System (GIS) platform to delineate potential areas with legacy P accumulation and depletion within the landscape. These maps identified areas with high P inputs and large potential stores of legacy P. Historically, southwestern Ontario has had the densest agriculture and populated areas in Ontario and has had high P inputs over time. County-scale trends such as shifts to intensive livestock or crop-based agriculture, or increasing urbanization were also identified. In Chapter 3, the fate and transport of P and the possible development of P legacies was explored in the context of risk. P export is influenced by environmental conditions in soil, as such, there is spatial variance in the likelihood that P will runoff or accumulate in soils. The environmental conditions may therefore be used to represent the vulnerability of the system and the risk to either lose or accumulate P. The cumulative agricultural P surplus map was used in conjunction with vulnerability maps to construct soil P risk maps. Different vulnerability models were explored, and ultimately soil erosion potential maps were used to identify vulnerable areas with a high risk of P losses to surface water and areas with a high risk of P accumulation in soil. It was determined that there was a higher risk of P accumulation in soil along the coast of Lake Erie, and it is possible that P legacies exist in these areas. The results inform nutrient management and abatement strategies and the adaptive implementation of conservation practices.

The work presented in this thesis aimed to demonstrate the use of remote geophysical methods and numerical modelling to address questions related to hydrogeological processes in freezing soils. Two different study areas and research questions were investigated: In chapter 1, Landsat 4-5 TM and RapidEye-3 datasets were used to identify groundwater discharge zones in the Central Mackenzie Valley of the Northwest Territories. Given that this area is undergoing active shale oil exploration, identification of groundwater discharge zones is of great importance. Discharge zones represent groundwater-surface water interaction points that are potential pathways for contaminants associated with hydraulic fracturing to move. Following the works of Morse and Wolfe (2015), a series of image algorithms were applied to imagery for the entire Central Mackenzie Valley, and for the Bogg Creek Watershed (a sub watershed of the CMV) for selected years between 2004 and 2017. The algorithm series extracted ‘icings’ from the images. Icings (also called aufeis) are surface ice lenses where groundwater discharges in the winter months, then freezes. Icings were statistically examined for all of the selected years to determine whether a significant difference in their occurrence and size existed. It was concluded that there was a significant difference in the spatial distribution of icings from year to year, but that there were several places where icings were recurring. During a field visit in August of 2018, high resolution thermal imagery was captured for several of these locations and it was found that groundwater was also discharging in the summer. This provides strong evidence to suggest that the recurring icings represent springs from which groundwater discharges year-round. These springs represent ideal locations to monitor the quality of discharging groundwater following the establishment of fracking operations. Furthermore, identifying these monitoring points remotely is expected to have drastically reduced the field efforts that would have been required to find them in situ. This thesis demonstrates the value of remote geophysical methods for hydrogeological applications, particularly in areas that have limited accessibility. The second component of this thesis establishes a conceptual model that describes the processes which influence winter subsurface pipe and water main breakage in Southern Ontario. Recent winters in this region have seen a dramatic increase in the number of pipe/main breaks that occur above the regional water table, but below the freezing front. Though subsurface pipe breakage is well-understood in saturated conditions or in the case when it is encompassed within the freezing zone, it is not well understood in the conditions of unsaturated soil which is not frozen, but contains a frozen zone above it. It is hypothesized that differential stress may be exerted on a pipe as soils of different frost heaving potential freeze and expand above the pipe. Once a certain overburden pressure is exceeded and upward frost heaving is no longer possible, it is expected that subsequent freezing will exert force onto the soil skeleton (and therefore pipe) below it. Two specific conceptual problems are developed in this work, and the processes which affect the distribution of stress in each case are described in detail. Then, a plan for the numerical implementation of each problem using the finite element model ABAQUS is given. Though the numerical models are not shown in this work, the conceptual models provide the framework which is necessary to implement a successful numerical model. The problem of pipe breakage in the unsaturated zone is complex and requires the coupling of equations which describe groundwater flow, stress-strain in a porous medium, and heat transport. Therefore, it was vital for an in-depth conceptual understanding of the problem to be established before introducing the numerical application.

Fresh water supplies in mountainous regions are at risk as snow and ice stores continue to decline under rising global temperatures, earlier winter snowmelt and changing climate regimes. Alpine forests are of particular importance due to their hydrological connectivity within watersheds controlling groundwater base flow, influencing evapotranspiration (ET) and snow storage dynamics. A change in the water availability to subalpine vegetation via changes in winter snowpack accumulation and quantities or differing summer precipitation (P) regimes could have a drastic effect on the long-term health of these forests. This makes it imperative to understand and quantify their hydrological connectivity within these watersheds. Study sites located at Fortress Mountain in Kananaskis, Alberta are composed of co-occurring coniferous tree stands of Abies lasiocarpa and Picea engelmannii. Little is known about water use dynamics of these species at high elevations, specifically the quantity and timing of transpiration (T) in addition to the water sources most important for T during the entire length of the growing season. This study used a combination of hydrological and meteorological tools to address coniferous subalpine tree water use behaviours before, during and after the growing season (June-September). Methodologies focussed on determining seasonal T patterns using the non-invasive stem-heat balance method to determine sap flow and eddy covariance to capture stand ET. The source water of the studied trees was determined using δ18O and δ2H stable water isotopes and further partitioned using the MixSIAR Bayesian Mixing Model (BMM). Groundwater monitoring wells, soil tensiometers, P gauges, and meteorological stations were used to determine baseline environmental conditions. Stable water isotopes δ18O and δ2H were collected from all source waters (P, snow cover, soil water, groundwater) in addition to xylem water samples from the coniferous trees within the study area. Understanding tree response to P and drying events was the main objective addressed, yielding stark differences between the growing seasons of 2016 and 2017. Stand T was higher in 2017 (165 mm) than 2016 (118 mm) despite a much drier and warmer season (155 mm of rain in 2017 compared to 283 mm in 2016). A deeper, sustained snowpack in 2017 coupled with higher net radiation allowed for higher T rates. Paired with δ18O and δ2H stable isotope source partitioning, this study was able to identify soil water as the most important source to season-long tree productivity, with groundwater the most important for early growing season. Well-drained soils and shallow depth to bedrock inhibited groundwater access for the studied trees after the snowmelt period concluded. Thus soil moisture supplied a majority of water to the tree population during mid growing season, determined both hydrometrically and isotopically. Dry conditions in 2017 showed a clear trend between soil moisture levels and tree water use, with 2016 having almost double the soil moisture and tree productivity in the tail end of the growing season. By closely examining the patterns of subalpine tree water use, we can begin to clarify how these important ecosystems services will be impacted under a changing climate in addition to helping us better manage our forest and freshwater resources.

Peatlands cover approximately 50% of the total landscape of the Western Boreal Forest, which includes the sub-humid Boreal Plains (BP) ecozone. The BP experiences persistent water deficit conditions, promoting anaerobic conditions, which has the potential to increase decomposition, transforming the peatlands from carbon sinks to carbon sources. With evapotranspiration (ET) being the dominant source of water loss in the BP, peatland persistence is hydrologically precarious, and as such, it is necessary to understand the dynamics and controls on ET within these systems. Due to the heterogeneity of the landscape, surrounding upland forests often shelter peatlands from wind. This results in spatially varying evaporative rates, which can influence surface moisture and vegetation regimes across a peatlands surface. High-resolution turbulent models allow for such flow scenarios to be resolved as they resolve flow in a 3D domain. Therefore, high-resolution turbulent models are essential in assessing the spatial variability of stresses placed on surface scalars such as ET, by displacement height transition. This study uses a canopy resolving large-eddy simulation (RAFLES) to study the impact of displacement height transitions on surface-atmosphere exchanges of moisture within peatlands of the BP. The dimensions, vegetation structure and energy dynamics of the modeled peatlands were generated from observations of natural peatlands of the BP. Within the sheltered region leeward of a backward-facing step transition, the simulated peatlands experienced higher resistances to surface-atmosphere exchanges of moisture when compared to the reattachment and recovery regions. However, this trend was muted when the surface roughness of the peatland was increased as the roughness lowered the overall resistance of the surface. This study also found that the length of the peatland did not influence the flow reattachment dynamics within the peatland. However, it was observed that the peatlands with a narrow shape and a curved front-facing step geometry resulted in faster regional wind velocities. Understanding the turbulent dynamics within heterogeneous landscapes can help to control the rate and variability of surface to atmosphere exchanges of moisture within disrupted and reclaimed landscapes which can increase the predictability of moisture demands within future landscapes.

Carbon storage in northern peatlands is estimated to be ~795 Tg, equivalent to ~40% of atmospheric CO2. Peatlands are dominant features of the Western Boreal Plains (WBP), which are experiencing a regime shift to a warmer and drier climate, as well as an increase in forest fire disturbance. Burning of the upper layers of rich organic matter peat releases enormous quantities of C to the atmosphere. The projected response of peatlands to forest fire is concerning, but widely understudied and could be of the utmost importance for the biogeochemical function and future net C balance of peatland. Impacts of climate change driven drying on peatland nutrient dynamics have been explored previously, however, the impacts of wildfire on nutrient dynamics have not been examined. This study assessed the impact of wildfire on N and P bioavailability and nutrient mineralization, plant nutrients balance, and the C and macronutrient stoichiometry and stock in a fen one-year post-wildfire by comparing a Burned and Unburned area. The results show that bioavailable P increased up to 200 times in surface water leachate, 125 times in groundwater and 5 times in peat. Surface ash leachate had increased concentrations in ammonium (NH4+) and nitrate (NO3-), and through groundwater mobility, the entire fen experienced increased bioavailable N. Mineralization of N and P were minimal at the Burned sites, relative to Unburned sites. Fire affected plant nutrient limitation patterns, switching from dominantly N-limited to NP co-limited in moss and P-limitation in vascular species. Burned site C stock (~14000 kg/ha) was higher relative to the Unburned site, which also increased CN and CP ratios. These findings suggest that long-term effects of elevated C, N, and P concentrations on plant productivity and decomposition must be re-evaluated for fire disturbance to understand the resiliency of peatland biogeochemistry post-wildfire. Environmental controls, including hydrologic, biologic, and edaphic variables modified by the fire and their effect on CO2 fluxes have not been studied holistically. In this thesis, I studied a treed fen burned during the Horse River wildfire in Fort McMurray, AB, comparing CO2 fluxes between a Burned and Unburned area of the fen. We see that both gross ecosystem productivity (GEP) and total respiration (Rtot) were reduced in magnitude at the Burned sites in comparison to the Unburned site, with peak fluxes in the Unburned site occurring in late June, whereas the Burned site CO2 fluxes peaked later in the growing season. GEP and net ecosystem exchange (NEE) increased in carbon uptake in the Burned sites along a depth of burn (DOB) gradient, with the deepest burned areas having an increased potential to uptake more CO2 than the Unburned site. The data also showed that both bioavailable P and moss recolonization were highest in the deepest burned areas. Unburned environmental controls on CO2 fluxes were dominated by soil temperature, whereas the Burned sites CO2 fluxes were controlled by leaf area index. One-year post-wildfire, the deepest burned areas had between 5-200 times greater concentration of P than the Unburned site, the most moss recolonization, and the greatest CO2 uptake, showing that deeper burning could potentially increase the recovery trajectory and resiliency of northern peatlands after fire disturbance.

Watershed urbanization profoundly alters the hydrologic characteristics of urban rivers compared to their rural counterparts. This change in hydrologic conditions in combination with alterations to the sediment supply regime in urban watersheds leads to adjustments to channel form and the widespread degradation of urban rivers. Urban river management increasingly attempts to balance the societal needs of flood and erosion control, while simultaneously improving the ecological health or waterways. Two common types of river management include stormwater management (SWM), which focuses on the attenuation of urban floods, and in-stream restoration, which attempts to reconstruct stable and ecologically favourable channels. However, current urban river management designs lack consideration of the key process responsible for channel stability and habitat availability: bedload sediment transport. Two reasons for this shortcoming are the lack of bedload sediment data in urban watersheds and the consequent gap in understanding of the bedload transport dynamics of urban rivers. Consequently, the degradation of urban rivers persists. This project investigates bedload transport dynamics in urban rivers with different management scenarios to focus on four themes: (1) how urbanization affects bedload transport dynamics and its relationship to channel morphology, (2) how to best predict bedload transport dynamics in urban rivers, (3) how current urban river management strategies change the transport dynamics of rivers, and (4) how to improve bedload sediment monitoring technology. This project focuses on the grain-scale bedload transport dynamics of coarse material because it links to the morphodynamics and ecological processes of channels, it provides insights on the exact controls and spatial variability of bedload transport, and the responses to individual flood events can be directly measured. The overarching goal of this study is to contribute to improved urban river management strategies that focus on adaptive management and interdisciplinary approaches. Bedload sediment transport was monitored using RFID tracer stones in three streams with different hydrologic settings: rural, urban with no SWM, and urban with SWM. High-resolution water level data confirmed the hydrologic differences expected from the three watershed conditions, as well as channel enlargement characteristic of urban rivers. Results demonstrate that the morphologic differences between the study streams can be linked to changes in the grain-scale bedload transport dynamics of the streams. Bedload transport is accelerated in the urban stream due to an increase in the frequency of bedload mobilization, particularly of coarse sediment sizes. In contrast, SWM hasdecreased the bedload transport to an immobile and armoured state indicative of a competence-limit transport regime. Results are used to make recommendations for improved urban river management. Results from the bedload tracking were used to build predictive models of tracer displacements. A new variable that captures both the mobility and travel length of bed particles is introduced. Several flow metrics developed in the literature in rural and laboratory settings are calculated, and their ability to predict tracer displacements in the three streams is tested. Scaling tracer travel lengths by mean channel width collapses the data into a single, strong relationship with cumulative energy expenditure, providing a single model that can be used across systems with different watershed conditions. To assess the impact of an in-stream riffle-pool channel reconstruction on bedload sediment dynamics, bedload transport and morphologic change was monitored in adjacent unrestored and restored reaches of an urban channel. Results reveal that the restored reach is stable and self- maintaining, mirroring bedform maintenance processes in natural riffle-pool streams. However, the construction is more successful at slowing down the transport of coarse sediment more than fine sediment, leading to a coarse sediment discontinuity that may be contributing to accelerated channel adjustment beyond the limits of the constructed riffle-pool sequences. This project highlights the importance of considering the entire channel corridor when designing and monitoring restoration projects. A large limitation of bedload sediment tracking technology is the inability to determine the vertical position of tracers, which hinders the ability to study vertical mixing and translate tracer data into bedload transport rates. A new Radio Frequency Identification (RFID) bedload tracer stone is presented, along with results of laboratory performance tests. This new bedload tracer improves upon existing bedload sediment monitoring technology by providing the ability to measure the burial depth of tracers without disturbing the bed. An important contribution of this study is the extensive dataset of bedload transport collected in urban rivers. This study attempts to move away from descriptive differences in the characteristics of urban rivers compared to rural streams, and towards a process-level understanding of the anthropogenic effects on river systems. Grain-scale bedload transport theory, developed in rural and laboratory settings, is applied to urban settings to gain insights into the effects of urbanization and common river management strategies on the geomorphic processes of urban rivers. Recommendations for improved urban river management are developed from the results of this thesis.

Background: Fish is an important food source for many Indigenous communities within the Dehcho and Sahtú regions of Northwest Territories (NWT). However, environmental toxicants, such as mercury (Hg), can bioaccumulate and biomagnify to reach detectable levels in fish, particularly among predatory fish species. From a public health perspective, Hg exposure and biomonitoring assessments are invaluable assets that can highlight potential risk factors and evaluate risks associated with Hg toxicity. With the current body of knowledge, exposure assessment models conservatively account for Hg bioaccessibility in fish to be 100% because of the limited information available. Past literature have also noted statistical differences in Hg bioaccessibility with respect to food source as well as food preparation methods. As such, there is a need for more Hg biomonitoring and Hg bioaccessibility assessments to elucidate the relationships between dietary and demographic determinants with respect to Hg exposure. Objectives: Hg biomonitoring component. 1) To determine internal Hg exposure levels (blood and hair) from six Indigenous communities in the Dehcho and Sahtú regions of NWT. 2) To conduct t-test analysis for internal Hg exposure levels with respect to sex. 3) To construct simple linear regression models between internal Hg exposure levels with the following factors: i) age ii) non-piscivorous fish consumption iii) piscivorous fish consumption. Hg bioaccessibility component. 1) To characterize Hg concentrations for uncooked and dried by smoking (dried/smoked) lake whitefish (Coregonus clupeaformis). 2) To determine Hg bioaccessibility in uncooked and dried/smoked lake whitefish. 3) To conduct t-test analysis to determine Hg concentration and Hg bioaccesibility differences between uncooked and dried/smoked whitefish. Methods: Hg biomonitoring component. 150 Dene and Métis participants, between the ages of 6-79, were recruited from Deline, Fort Providence, Hay River Reserve, Kakisa, West Point First Nation and Jean Marie River of NWT. Blood and hair samples were collected for internal Hg exposure assessments. Participants were also asked to provide demographic information and complete a food frequency questionnaire (FFQ) for describing potential determinants of Hg exposure. T-test analysis and log-linear simple regression models were used to identify individual relationships. Hg bioaccessibility component. Five whole lake whitefish (Coregonus clupeaformis) were used to investigate Hg bioaccessibility. Subsamples were subjected to two treatments: uncooked and dried/smoked. All dried/smoked and uncooked fish samples underwent a single phase (gastric only) IVBA model treatment. All Hg concentrations were expressed as wet weight and determined using an Ohio-Lumex portable mercury analyzer with pyrolyzer attachment. Results: Hg biomonitoring component. Hg levels in blood and hair were highly linearly-correlated with one another (r= 0.854, p<0.05) and all of the participants had blood and hair Hg levels below 20μg/L and 6μg/g respectively. T-test analysis found no statistical differences (p>0.05) when comparing blood Hg levels of male and female participants (2.711 ± 2.616 µg/L and 2.667 ± 4.079 µg/L respectively) Similarly, T- test analysis found no statistical differences (p>0.05) when comparing hair Hg levels of male and female participants (0.868 ± 0.925 µg/L and 0.671 ± 0.866 µg/g respectively). According to the simple regression model’s findings, age was the only statistically significant factor and was positively associated with blood and hair Hg levels (B=0.020 and p<0.01; B=0.011 and p<0.01). Based on simple regression model’s findings, piscivorous fish consumption was not significantly associated with either Hg levels found within blood and hair (B= 0.068 and p>0.05; B= 0.023 and p>0.05). Similarly, non-piscivorous fish consumption does not appear to be significantly associated with blood and hair Hg levels (B= 0.022 and p>0.05; B= 0.035 and p>0.05) Hg bioaccessibility component. After culinary treatment, dried/smoked whitefish samples retained, on average, 46% of their original mass (i.e., due to moisture loss). As a result of moisture loss, the initial Hg concentration in dried/smoked whitefish samples (0.194 mg/kg) was significantly greater (p=0.002) relative to uncooked whitefish samples (0.080 mg/kg). When comparing Hg bioaccessibility, dried/smoked whitefish (53%) was statistically lower (p< 0.001) compared to uncooked whitefish (102%). Despite the reduction in Hg bioaccessibility for dried/smoked whitefish, the bioaccessible Hg concentration in dried/smoked whitefish (0.101 mg/kg) was significantly greater (p< 0.05) compared to uncooked whitefish samples (0.081 mg/kg). The loss of moisture mass and the increased density of Hg within dried/smoked whitefish samples were likely responsible for the increase in total and bioaccessibile mercury in dried/smoked whitefish. As such, dried/smoked whitefish samples had greater amounts of Hg digested and solubilized in aqueous solution. Conclusion: Both simple regression and static IVBA models’ findings have provided additional information for characterizing Hg exposure among participating communities in Northwest Territories. Based on the study population’s demographic determinants and dietary patterns, age is a key factor for understanding and predicting Hg exposure in the Northwest Territories. Bioaccessibility findings suggest source apportionment models that incorporate differences in Hg concentration between raw and smoked/dried fish may also need to account for the corresponding difference in bioaccessibility. For further generalizability of IVBA model’s findings, additional in vitro bioaccessibility studies using different fish species are needed. This research is important because it highlights dietary factors, such as cooking method, are potentially important to human exposure. Inclusion of such factors to source apportionment model may help further characterize the relationship between internal Hg exposure levels detected in biomarkers and the external exposure estimates.

Worldwide phenomena called algae bloom has been recently a serious matter for inland water bodies. Temporal and spatial variability of the bloom makes it di cult to use in-situ monitoring of the lakes. This study aimed to evaluate the potential of Sentinel-3 Ocean and Land Colour Instrument (OLCI) and Sentinel-2 Multispectral Instrument (MSI) data for monitoring algal blooms in Lake Erie. Chlorophyll-a (Chl-a) related products were tested using NOAA-Great Lakes Chl-a monitoring data over summer 2016 and 2017. Thematic water processor, fluorescence line height/maximum chlorophyll index (MCI) and S2 MCI, plug-in SNAP were assessed for their ability to estimate Chl-a concentration. We processed both Top of the Atmosphere (TOA) reflectance and radiance data. Results show that while FLH algorithms are limited to lakes with Chl-a < 8 mg m-3, MCI has the potential to be used effectively to monitor Chl-a concentration over eutrophic lakes. Sentinel-3 MCI is suggested for Chl-a > 20 mg m-3 and Sentinel-2 MCI for Chla > 8 mg m-3. The different Chl-a range limitation for the MCI products can be due to the different location of the maximum peak bands, 705 and 709 for MSI and OLCI sensors respectively. TOA radiances showed a signi cantly better correlation with in situ data compared to TOA reflectances which may be related to the poor pixel identi cation during the process of pixel flagging affected by the complexity of Case-2 water. Our fi nding suggests that Sentinel-2 MCI achieves better performance for Chl-a retrieval (R2 = 0.90). However, the FLH algorithms outperformed showing negative reflectance due to the shift of reflectance peak to longer wavelengths along with increasing Chl-a values. Although the algorithms show moderate performance for estimating Chl-a concentration; this study demonstrated that the new satellite sensors, OLCI and MSI, can play a signi ficant role in the monitoring of algae blooms for Lake Erie.

Agricultural tile drains are a source of phosphorus (P) contributing to eutrophication. Preferential flowpaths can rapidly transport P to tile drains, but their activation in different soil textures and under variable seasonal conditions (antecedent moisture conditions and presence of soil frost) is not well understood. Subsurface placement of fertilizer has been proposed as a management practice to reduce P loss, compared to surface applications. However, how subsurface placement reduces P loss is not well understood. The goal of this thesis is to relate subsurface flowpaths and fertilizer placement to identify source and transport mechanisms controlling P movement to tile drains, across soil textures and seasonal conditions. A lab experiment was done on intact soil monoliths (clay, silt loam) to investigate interactions between fertilizer placement, subsurface flowpaths, and soil frost. Conservative water tracers (Brˉ, Clˉ and D₂O) applied through successive events identified matrix flow as the dominant flowpath in unfrozen silt loam, while preferential flow dominated in unfrozen clay and in both soil types under partially frozen conditions. Subsurface placement of fertilizer reduced dissolved reactive P losses by 60% in silt loam and 64% in clay over the simulated non-growing season compared to surface broadcast applications. A field study used blue dye as a tracer to investigate subsurface flowpaths in clay and silt loam plots under wet and dry conditions. Dye stain patterns were analyzed to determine the relative importance of matrix and macropore flow. Soil samples were collected to determine soil P distribution post-fertilization. Preferential flow occurred under all soil texture and moisture conditions. Dry clay soil showed the deepest staining (92 ± 7.6 cm), followed by wet clay (77 ± 4.7 cm). In silt loam soil, depth of staining did not differ between wet (56 ± 7.2 cm) and dry (50 ± 6.6 cm) conditions. Soil water-extractable P distribution varied with fertilizer application in the top 10 cm of the soil profile, but did not differ at depth. Together, the results of this research suggest subsurface placement is a suitable practice for minimizing subsurface nutrient loss, by reducing contact between the nutrient supply and preferential flowpaths, particularly in clay soils prone to preferential flow. This work provides an improved understanding of subsurface flowpaths carrying P to tile drains, and more broadly, solute transport through preferential flowpaths.

In the coupled human-environment system, humans play a central role in creating various environmental problems, and in turn, are impacted by these environmental consequences. In Canada, water quality degradation caused by agricultural activities has become a severe problem for a long time. It has been noted that the application of pesticides, manure and fertilizers have led to an increasing amount of chemicals and other pollutants in surface runoff which eventually converge into surface water bodies and result in water eutrophication. To maintain water quality and develop a sustainable agricultural system, Best Management Practices (BMPs) have been suggested. However, the high complexity of the agriculture system makes it difficult for policymakers and researchers to monitor and evaluate the performance of BMPs across large spatial scales and develop appropriate improvement strategies accordingly. Under these circumstances, agent-based models (ABM) stand out for their ability to deal with the complexities in the agri-environment system. To better understand the dynamics of farmer’s decision-making on BMP application under different socio-economic and environmental situations, an ABM has been developed to simulate the decision-making processes in the Upper Medway Creek subwatershed in this study. The ABM uses an optimizing decision-making structure that relies on choice by highest utility. In addition, the ABM integrates a weighted sum function to evaluate the influences of economic, environmental and social factors on farmers’ decision-making. Results from the model pre-test were compared to those obtained from a random generator to examine how does the developed ABM perform against the random generator. Then, a sensitivity analysis has been performed using the one-factor-at-a-time method to examine the impacts of different potential interventions, including government subsidies and educational activities, on farmers’ decision-making for certain BMP adoptions. The results demonstrated that the developed ABM is robust in simulating farmers’ decision-making on BMP application within the Upper Medway Creek subwatershed. According to the sensitivity analysis, providing subsidies and improving knowledge level of BMPs have positive effects on the implementations of certain BMPs in general. While comparing to improving knowledge levels of BMPs, providing subsidies makes greater contribution to motivating farmers to adopt BMPs. For each BMP, a subsidy rate, which indicates the proportion of implementation costs needs to be subsidized to effectively encourage the BMP adoption, has been suggested. The results of this study provide a better understanding of how different socio-economic conditions affect farmers’ decision-making on BMP adoptions and offer insights for policymakers to develop effective strategies incentivising farmers’ adoptions of BMPs and further preserving water quality in the Upper Medway Creek subwatershed.

Oil sands mining operations began in 1967, but the onset of a monitoring program to assess water and sediment quality in the Athabasca River watershed began 30 years later. Consequently, no knowledge of pre-industrial, baseline conditions exists upon which current river sediment quality can be compared. This has undermined an ability to determine the relative importance of contaminants supplied by natural processes versus pollution to the Athabasca River by rapid growth of oil sands development. In this study, a paleolimnological approach was used to analyze sediment cores from five flood-influenced lakes located upstream and downstream of oil sands operations within the Alberta Oil Sands Region (AOSR). Loss-on-ignition and organic carbon and nitrogen elemental analyses were used to differentiate periods of strong and weak Athabasca River flood influence. In addition, the temporal changes in concentrations of bitumen-associated metals vanadium (V) and nickel (Ni) were explored at each lake. A pre-industrial baseline was developed using pre-1967 sediment concentrations of V and Ni, normalized to aluminum concentration, from lakes in the AOSR to estimate the natural range of variability of these metals. When normalized metals concentrations in recently deposited flood-influenced sediment were compared to the pre-industrial baseline, no evidence of enrichment in the river-derived stratigraphic intervals was detected. However, significant enrichment of bitumen-related metals V and Ni (up to 2- and 1.6-fold above the baseline, respectively) was observed in weakly flood-influenced sediment in the two floodplain lakes located closest to the most active mining operations (< 10 km), indicating local atmospheric pollution. Athabasca River sediment data collected by regional monitoring programs RAMP (1997-2002) and JOSM (2012-2014) were examined in the context of the newly developed baselines and showed enrichment of V (1.2-1.7x baseline) and Ni (1.2-2.0x baseline) at some of the river monitoring sites, usually proximal to tributary outflows. This research indicates that sediment profiles from floodplain lakes along the lower Athabasca provide valuable information as pre-industrial depositional areas of natural sediment metals. Paleohydrological analyses, however, indicate that flood-influence at many of these lakes is declining, coincident with oil sands growth, and so many of the lakes no longer frequently capture flood sediments. Nonetheless, metal-specific baselines using the pre-1967 data can be used to detect enrichment in modern sediments of the floodplain lakes and in river sediment monitoring data, the latter previously criticized for inadequate baseline knowledge, and which also now serves as a foundation for ongoing river sediment monitoring.

Background: Country food consumption by Indigenous peoples is associated with improved nutrition, food security, and lower rates of chronic diseases (Kuhnlein, Burlingame, & Erasmus, 2013); however, these foods can also pose potential risks of exposure to contaminants such as mercury and cadmium (Berti, 1997). Elevated fish mercury concentrations in some lakes in the Sahtú region of the Northwest Territories (NWT) resulted in a series of consumption notices that suggested people limit their consumption of walleye, northern pike, and lake trout from specific lakes in the region (NWT, 2016). Therefore, as part of a health communication component was added. This component was designed to assess participants risk perceptions and awareness of current consumption notices and health messages, to provide baseline data to evaluate the impact of consumption notices, to determine how health messages are currently developed, and to make recommendations to create more targeted communication strategies. Objectives: The research objectives of this thesis are to: 1) Assess participants’ risk perceptions and awareness of current consumption notices and health messages; 2) Provide baseline data to evaluate the impact of consumption notices and health messages over time; 3) Understand how consumption notices and health messages are currently developed and communicated to communities by the Government of the Northwest Territories; and 4) Make recommendations that aim to improve and create more effective communication strategies with focus in the Sahtú Region of the Northwest Territories based on knowledge synthesized from terminology workshop(s), surveys, and interviews. Methods: This project involves a Health Messages Survey, two terminology workshops, community interviews, and stakeholder interviews. Participants were invited to take part in a Health Messages Survey where they were asked questions about their perceptions of contaminants, whether they had heard or seen consumption notices, in addition to their preferences for future health messages based on trust and accessibility. Two terminology workshops were implemented in Délı̨nę, where key terms from the human biomonitoring project were identified, in order to translate or interpret into Slavey language. Twelve interviews were conducted with community members in Délı̨nę to document perceptions and stories regarding contaminants. Interviews also took place in Yellowknife with stakeholders from the Government of the Northwest Territories, the Federal Government, the Dene Nation, and the Giant Mine Oversight Board. Results: Based on the data collected by surveys (n=43), interviews (n=19) and two terminology workshops (n=27) during a two-year period from 2017-2018, we found that country foods are extremely valued in communities for many reasons. Each participant reported eating country foods and many preferred solely relying on country foods rather than: i) store-bought foods; or ii) a mix of store-bought foods and country foods. During community interviews, each participant expressed their gratitude for the country foods that they eat and wished that they could eat it more often. In the terminology workshops, the cultural importance of trapping, fishing, and hunting was discussed. Many participants had heard or seen messages that promoted the nutritional value of country foods and had also learned from their Elders and family members that these foods have been a source of sustenance for thousands of years. The majority of participants had heard or seen messages about fish with high levels of mercury in their country foods from radio, researcher or scientists, and friends and family. These participants also expressed their fear that contaminants may have impacted the country food that they most preferred to eat. The fear that participants expressed invoked change by the Government of the Northwest Territories (GNWT. The GNWT, the primary disseminator of health messages, has since changed their communication strategy from only consumption notices, to fact sheets and general fish consumption guidelines. Conclusion: This project demonstrated the importance of country food in the Sahtú region. In the Sahtú country foods connects people to the traditional and cultural practices that were passed on by parents and Elders which provides a sense of community and of belonging. This project provides baseline data based on surveys, interviews and terminology workshops for a clearer picture of how these participants perceive contaminants, where they currently get their health information, and who they would trust to receive this type of information. More research in the area of northern health and risk communication is necessary in order to compare results to other regions across northern Canada and to determine best practices. In order to create more culturally significant and relevant contaminant health messages, researchers and government must invest time into determining ways that will be well received and heard by the communities they are working with and be adaptable based on the region, a community. For instance, in the Sahtú region, this could include using methods such as the local radio, friends and relatives, researchers or other health workers (nurses, etc.) as the most trusted sources to receive this type of information. Another consideration is assuring that these messages are understood when disseminated. In communities of the Sahtú, this means that providing information about contaminants must be translated in local languages or having translators to disseminate results of research projects. We can use these results as a baseline to compare with other health communication projects happening in northern Canada and other circumpolar Indigenous communities and regions. This thesis has provided further context in disseminating contaminant and health and risk communication messaging and suggestions to improve communication strategies.

Water quality in many regions of the Great Lakes Basin (GLB) has deteriorated due to numerous anthropogenic drivers, including increases in agricultural area, increased fertilizer use, intensive livestock production, and increases in human population densities. Excessive nutrient inputs from both point and non-point sources have accelerated eutrophication in inland watersheds and in receiving water bodies, and policy goals have recently been set to reduce phosphorus loading to Lake Erie by as much as 40%. Under such pressures, it is crucial to better our understanding of nutrient transport across the GLB and to identify key watershed drivers of both seasonal and annual nutrient loading from watersheds to the lakes. In this research study, I have utilized numerous metrics to characterize nutrient dynamics in Great Lakes Watersheds across a gradient of human impacts and have attempted to identify key controls on biogeochemical signatures. As a part of this work, I paired water quality data from over 200 Great Lakes watersheds with land use and climate data to identify dominant controls on stream nutrient concentrations at the annual, seasonal, and event scales. At the annual scale, standardized regression analysis identified significant relationships between flow-weighted concentration (FWCs) and selected catchment characteristics. FWCs were found to be strongly linked to land-use variables such as combined agricultural and urban land, wetlands and tile drainage. Our quantification of these relationships was used to create spatial maps of annual nutrient concentrations and loads and to identify nutrient hotspots across the GLB. Specifically, high nutrient concentrations and export were observed in the Maumee and Sydenham River catchments, whereas lower concentrations and loads were found in Lake Superior catchments. At the seasonal scale, three primary seasonal nutrient regimes were identified: (1) ‘in-phase’ (positive correlation between monthly concentrations and discharge), (2) ‘out-of-phase’ (negative correlation), and (3) ‘stationary’ (no significant relationship). While in-phase seasonality was found to be the most common concentration regime for watersheds with higher levels of agricultural land use, nitrate seasonality in particular was found to be muted in watersheds with the highest agricultural land use, but to be more extreme in watersheds with less agriculture but higher amounts of forested area and higher wetland densities. Out-of-phase seasonality was found to be significantly associated with higher population densities and higher percent urban areas. At the event-scale, concentrations were found to be more variable with discharge for phosphorus than for nitrate. Additionally, Lake Erie showed significantly lower concentration variability in relation to discharge compared to all the other Lakes. As the Lake Erie basin also has higher agricultural land use than the other lakes, the more chemostatic concentration dynamics in these watersheds appears to be linked to agricultural nutrient use and suggests that agricultural nutrient legacies may be an important driver of current patterns in nutrient delivery to the lakes.

Background: Fish are often rich in essential micronutrients including omega-3 fatty acids (FA), and are a cultural and dietary staple in traditional food systems of First Nations communities in the Canadian subarctic. Country foods including fish contribute to improved food security, and promote the cultural sovereignty of First Nations communities. However, these foods are often a primary route of exposure to methylmercury, an environmental contaminant that can pose significant adverse health risks. Objectives: The objectives of this study are to: 1) Determine the concentration of total mercury (Hg) and long chain omega-3 polyunsaturated fatty acids (omega-3 PUFAs) in the muscle tissue of various wild-caught freshwater fish species harvested from eight lakes in the Dehcho region, Northwest Territories (NWT); 2) Construct a probabilistic, population-based retrospective dose reconstruction model to assess dietary omega-3 PUFA intake and Hg exposure across several Dehcho First Nations communities; 3) Characterize and quantify sources of mercury and omega-3 PUFA exposure from country food consumption, and identify key contributors from the diet; and 4) Assess the utility and accuracy of the probabilistic exposure model at estimating population-level profiles of risk and cardioprotective benefit using biomarkers for omega-3 PUFAs in blood plasma and Hg in hair. Methods: Samples from eight freshwater fish species [Burbot (Lota lota; also known as Loche or Mariah), Cisco (Coregonus artedi; also known as Herring), Lake Trout (Salvelinus namaycush), Lake Whitefish (Coregonus clupeaformis), Longnose Sucker (Catostomus catostomus), Northern Pike (Esox lucius; known locally as Jackfish), Walleye (Sander vitreus; also known as Pickerel) and White Sucker (Catastomus commersoni)] were harvested from eight lakes of the Dehcho (including Ekali, Trout, Sanguez, Tathlina, McGill, Gargan, Mustard, and Kakisa Lakes) in August of 2013, 2014, and 2015. Omega-3 PUFA levels in fish tissue were determined by a lipid extraction on pulverized, homogenized fish tissue and quantified using a gas chromatograph with a flame ionization detector, and freeze-dried, homogenized fish muscle tissue samples were analyzed and quantified for total Hg using a Direct Mercury Analyzer. Fish mercury and fatty acid profiles were paired with primary species-specific country food consumption data collected during the Contaminants Biomonitoring Study in the Northwest Territories Mackenzie Valley using a food frequency questionnaire (FFQ). A retrospective probabilistic dose reconstruction model was developed using Oracle Crystal Ball™ advanced risk modeling software, to simultaneously characterize intake of methylmercury (MeHg) and n-3 PUFAs through country food consumption, and to estimate both the population proportion at risk of exceeding the tolerable daily intake for MeHg, and those not meeting adequate intakes for PUFAs. A two-dimensional Monte Carlo analysis was conducted and profiles of exposure to MeHg and n-3 PUFAs were generated. Results from the model output were compared to toxicological and nutritional data from the results of the Contaminant Biomonitoring Study in the Northwest Territories Mackenzie Valley. Results: Mean HgT concentrations within piscivorous fishes (e.g. Northern Pike, Walleye, Burbot and Lake Trout) were up to 7.3 times higher than observed in benthivorous and planktivorous fishes (e.g. Cisco, Lake Whitefish, and Sucker). Further, EPA+DHA concentrations in Lake Trout were up to 4.5 times higher than observed in other piscivorous fish species harvested, including Burbot, Northern Pike, and Walleye. Significant differences were noted for mercury and fatty acid profiles in fish between lakes. Negative correlations were observed between mercury and fatty acids for Burbot, Northern Pike and Walleye. Stratifying by species, mean DHA:HgT ratios for Lake Whitefish and Cisco were up to 8.7-fold higher than in piscivorous fish species including Northern Pike, Walleye and Burbot. As an exception, Lake Trout, demonstrating higher omega-3 PUFAs than other species, had accordingly higher fatty acid:mercury ratios. Based on fatty acid and mercury levels in fish species of the Dehcho, and results from the FFQ, estimates for mercury exposure from fish consumption among the Dehcho population indicated that up to 7% of trials exceeded the pTWI of 3.29 μg/kg/week (0.47 μg/kg/day). In contrast, only 0.5% of respondents within participating Dehcho communities exceeded Health Canada’s recommended guidance value of 6 mg/kg. Mean hair mercury was 0.74 mg/kg, with a geometric mean of 0.38 mg/kg. Generally, only a small proportion of trial values exceeded Dietary Reference Intakes for fatty acid subgroups DHA, EPA+DHA, and total omega-3 PUFAs. Similarly, the Omega-3 Index of participants indicated levels of EPA+DHA that fell within the category associated with very low cardioprotective benefits. Sensitivity analyses indicated that input variables corresponding to Lake Whitefish were strong drivers of fatty acid intake across all fatty acid subgroups, while the proportion of the population consuming Northern Pike and Walleye were primary drivers of exposure to methylmercury intake. Conclusion: Probabilistic models provide an important lens for characterizing the risks and benefits from country food consumption in First Nations communities of the Dehcho region. Future studies in probabilistic human health risk assessment will incorporate a component to the model that characterizes risk not only within the general population, but in demographics most vulnerable to the risks associated with mercury exposure, including young children, women of childbearing age, and pregnant women. Any consumption notices and advisories that outline recommendations to modify country food use must consider the multitude of sociocultural, nutritional and spiritual benefits of these foods in subarctic Indigenous populations.