Abstract. The destructive nature of the ∼ 590 000 ha Horse river wildfire in the Western Boreal Plain (WBP), northern Alberta, in May of 2016 motivated the investigation of the hydrometeorological conditions that preceded the fire. Historical climate and field hydrometeorological data from a moderate-rich fen watershed were used to (a) identify whether the spring 2016 conditions were outside the range of natural variability for WBP climate cycles, (b) explain the observed patterns in burn severity across the watershed, and (c) identify whether fall and winter moisture signals observed in peatlands and lowland forests in the region are indicative of wildfire. Field hydrometeorological data from the fen watershed confirmed the presence of cumulative moisture deficits prior to the fire. Hydrogeological investigations highlighted the susceptibility of fen and upland areas to water table and soil moisture decline over rain-free periods (including winter), due to the watershed's reliance on supply from localized flow systems originating in topographic highs. Subtle changes in topographic position led to large changes in groundwater connectivity, leading to greater organic soil consumption by fire in wetland margins and at high elevations. The 2016 spring moisture conditions measured prior to the ignition of the fen watershed were not illustrated well by the Drought Code (DC) when standard overwintering procedures were applied. However, close agreement was found when default assumptions were replaced with measured duff soil moisture recharge and incorporated into the overwintering DC procedure. We conclude that accumulated moisture deficits dating back to the summer of 2015 led to the dry conditions that preceded the fire. The infrequent coinciding of several hydrometeorological conditions, including low autumn soil moisture, a modest snowpack, lack of spring precipitation, and high spring air temperatures and winds, ultimately led to the Horse river wildfire spreading widely and causing the observed burn patterns. Monitoring soil moisture at different land classes and watersheds would aid management strategies in the production of more accurate overwintered DC calculations, providing fire management agencies early warning signals ahead of severe spring wildfire seasons.

Oil sands development within the Athabasca Oil Sands Region (AOSR) has accelerated in recent decades, causing alteration to natural ecosystems including wetlands that perform many vital ecosystem functions such as water and carbon storage. These wetlands comprise more than half of the landscape, and their distribution and local hydrology are the result of interactions among a subhumid climate, topography, and spatially heterogeneous surficial and bedrock geology. Since hydrology plays a fundamental role in wetland ecological functioning and determines wetland sensitivity to human disturbances, the characterization of anthropogenic impacts on wetland hydrology in the AOSR is necessary to assess wetland resilience and to improve current best management practices. As such, this paper reviews the impacts of oil sands development and related disturbances including infrastructure construction, gravel extraction, and land clearing on wetland function in the AOSR. Hydrologic disturbances in wetlands in the AOSR include changes to soil hydrophysical properties that control water table position, the interruption of recharge–discharge patterns, and alteration of micrometeorological conditions; these in turn govern wetland ecological structure and wetland ecosystem processes (e.g., evapotranspiration, nutrient cycling). Given that anthropogenic disturbance can affect natural wetland succession, long-term hydrological monitoring is crucial for predicting the response of these ecosystems to varying levels of human impact.

Peatlands are a dominant land feature in the Athabasca Oil Sands Region (AOSR) of the Western Boreal Plain (WBP), comprising >50% of the total land area, many of which are moderate-rich fens. The carbon stocks of moderate-rich fens in the WBP are susceptible to degradation through anthropogenic-and climate-related factors, yet, few studies have aimed to understand their hydrologic function. This study, located in a meltwater channel belt characterized by relatively thin outwash sand and gravel (~6 m) underlying the peat, provides the first hydrological assessment of a moderate-rich fen in the AOSR. The lithology, hydrological function and groundwater geochemistry all point to the dominance of a local flow system supplying groundwater to the fen areas, evidenced by a thick (~16 m) and shallow (~7 m below ground surface) aquitard underlying the outwash, restricting hydrological connectivity between the fen and underlying regional aquifers. Vertical hydraulic gradients between the peat and underlying outwash aquifer, and horizontal hydraulic gradients between the fen and upland varied in response to both short-term and seasonal precipitation trends. Groundwater discharge to the fen was enhanced during wet periods characterized by high rainfall. Conversely, flow reversals (groundwater recharge; fen to underlying aquifer and upland), and subsequently, enhanced fen water table drawdown persisted during extended dry periods. This local groundwater flow-system influences recharge/discharge patterns at Poplar Fen, with hydraulic head in the underlying outwash aquifer highly susceptible to fluctuations in the presence and absence of precipitation-driven recharge from adjacent uplands. Moderate-rich fens similar to that studied here will likely become more susceptible to drying in the future due to a changing climate, leading to enhanced water table drawdown, peat oxidation and subsequent decomposition, vulnerability to wildfire, and seral succession to a more ombrogenous peatland system.

Growth of natural resource development in northern Canada has raised concerns about the effects on downstream aquatic ecosystems, but insufficient knowledge of pre-industrial baseline conditions continues to undermine ability of monitoring programs to distinguish industrial-derived contaminants from those supplied by natural processes. Here, we apply a novel paleolimnological approach to define pre-industrial baseline concentrations of 13 priority pollutant metals and vanadium and assess temporal changes, pathways and sources of these metals at a flood-prone lake (SD2) in the Slave River Delta (NWT, Canada) located ~500 km north of Alberta's oil sands development and ~140 km south of a former gold mine at Yellowknife, NWT. Results identify that metal concentrations, normalized to lithium concentration, are not elevated in sediments deposited during intervals of high flood influence or low flood influence since onset of oil sands development (post-1967) relative to the 1920-1967 baseline established at SD2. When compared to a previously defined baseline for the upstream Athabasca River, several metal-Li relations (Cd, Cr, Ni, Zn, V) in post-1967 sediments delivered by floodwaters appear to plot along a different trajectory, suggesting that the Peace and Slave River watersheds are important natural sources of metal deposition at the Slave River Delta. However, analysis revealed unusually high concentrations of As deposited during the 1950s, an interval of very low flood influence at SD2, which corresponded closely with emission history of the Giant Mine gold smelter indicating a legacy of far-field atmospheric pollution. Our study demonstrates the potential for paleolimnological characterization of baseline conditions and detection of pollution from multiple pathways in floodplain ecosystems, but that knowledge of paleohydrological conditions is essential for interpretation of contaminant profiles.

Abstract. The destructive nature of the ~ 590,000 ha Horse River Wildfire in the Western Boreal Plain (WBP), northern Alberta in May of 2016 motivated the investigation of the hydrometeorological conditions that preceded the fire. Historical climate and field hydrometeorological data from a moderate-rich fen watershed were used to identify a) whether the spring 2016 conditions were outside the range of natural variability for WBP climate cycles; b) explain the observed patterns in burn severity across the watershed; and c) identify whether fall and winter moisture signals observed in peatlands and lowland forests in the region are indicative of fire susceptibility. Field hydrometeorological data from the fen watershed confirmed the presence of cumulative moisture deficits prior to the fire. Hydrogeological investigations highlighted the susceptibility of fen and upland areas to water table and soil moisture decline over rain-free periods (including winter), due to the watershed's reliance on supply from localized flow systems originating in topographic highs. Subtle changes in topographic position led to large changes in groundwater connectivity, leading to greater organic soil consumption in wetland margins and at high elevations. The 2016 spring moisture conditions measured prior to the ignition of the fen watershed were not illustrated well by the Drought Code (DC) when standard overwintering procedures were applied. However, close agreement was found when default assumptions were replaced with measured duff soil moisture recharge and incorporated into the overwintering DC procedure. We conclude that accumulated moisture deficits dating back to the summer of 2015 led to the dry conditions that preceded the fire. The infrequent coinciding of several hydrometeorological conditions, including low autumn soil moisture, a modest snowpack, lack of spring precipitation, and high spring air temperatures and winds, ultimately led to the Horse River wildfire spreading widely and causing observed burn patterns. Monitoring soil moisture at different land classes and watersheds would aid management strategies in the production of more accurate overwintered DC calculations, providing fire management agencies early warning signals ahead of severe spring wildfire seasons.