Forest disturbance from wildfire, pathogens, or forest harvesting removes the forest canopy, increasing the total precipitation that reaches the forest floor (Williams et al., 2019; Burles and Boon, 2011; Boon, 2012; Pugh and Small, 2012; Varhola et al., 2010), often altering the dominant flow pathways, increasing streamflow quantity, and changing the timing of flows in forested watersheds (Stednick, 1996; Scott, 1993; Bearup et al., 2014; Winkler et al., 2017). However, large variability has been observed in streamflow responses following disturbance due to differences in disturbance type, vegetation type, precipitation regimes, and soil moisture storage (Brown et al., 2005; Stednick, 1996). Some studies in Alberta's Rocky Mountains have reported little, if any, change in streamflow following disturbance (Williams et al., 2015; Harder et al., 2015; Goodbrand and Anderson, 2016; Andres et al., 1987), but the mechanisms and watershed features (e.g., bedrock, surficial geology, and wetlands) potentially responsible for the lack of flow response have received little attention. It has been suggested that watersheds exhibiting a lack of change in streamflow following disturbance might be associated with a large storage capacity and complex subsurface flow pathways (Harder et al., 2015), but the higher-order controls regulating these muted responses remain unclear.

Runoff generation has been extensively studied in regions with relatively impermeable bedrock overlain by shallower soils, which has led to broadly accepted conceptualizations of runoff dynamics (e.g., old water contributions to streamflow, macropore flow, and subsurface streamflow generation – McGlynn et al., 2002; fill and spill hypothesis – Tromp-van Meerveld and McDonnell, 2006; hillslope–stream connectivity – Jencso et al., 2009). However, these conceptualizations may not apply to regions with more complex structural controls on runoff such as permeable bedrock, deeper soils, or where multiple subsurface systems interact. Runoff generation in Alberta's Rocky Mountains has added complexity because of the combination of both permeable sedimentary bedrock (highly fractured and faulted) and an overlying layer of deep, heterogenous glacial till (3 m deep, on average, and up to 10 m deep; AGS, 2004; Waterline Resources Inc., 2013). This is in contrast to regions such as the southern Rocky Mountains in Colorado (Sueker et al., 2000; Cowie et al., 2017) and Montana (Jencso et al., 2009, 2010; Nippgen et al., 2015) that are often dominated by less permeable metamorphic or igneous bedrock and thinner soils. While runoff generation processes may differ from these regions, some inferences can be drawn from studies in regions with either permeable bedrock or deep soils alone. Watersheds with high bedrock permeability have been associated with longer subsurface flow pathways and the slow release of stored water to streams during baseflow (Uchida et al., 2006; Liu et al., 2004; Pfister et al., 2017). Uchida et al. (2006) reported that a watershed with greater bedrock permeability had larger aquifer storage, and the subsequent release of stored water maintained baseflow later in the year. Similarly, Liu et al. (2004) showed that the recession limb of the annual hydrograph in the Colorado front range Rocky Mountains was driven by baseflow released from fractured bedrock, but Cowie et al. (2017) also stressed the importance of talus slopes as a source of streamflow in the same alpine watershed. Deep soils and till deposits with large storage capacities have also been shown to sustain baseflows during drier periods (Floriancic et al., 2018; Shanley et al., 2015). Deep sediment deposits in the Poschiavino watershed, in Switzerland, were associated with greater storage capacity and higher winter baseflows compared to watersheds with shallow sediment deposits (Floriancic et al., 2018). Similarly, deep basal till in the Sleepers River watershed in Vermont was associated with large storage capacity and low permeability that promoted the extended maintenance of baseflow (Shanley et al., 2015).

While these studies illustrate the influence of permeable or fractured bedrock, deep soils, or till on baseflows, few studies have explored the combination of these storage zones on streamflow contributions (Burns et al., 1998; Dalke et al., 2012; Shaman et al., 2004). Burns et al. (1998) characterized the difference in deep (bedrock) and shallow (soils and till) flow systems in the Catskill Mountains in New York, a region with both glacial till and permeable sedimentary bedrock. Baseflow was maintained by discharge from perennial springs which originated from bedrock fractures, rather than contributions from the soil and till flow system (Burns et al., 1998). Conversely, fragipan layers contributed to differing flow systems under dry vs. wet antecedent conditions in central New York (state), USA (Dalke et al., 2012). Storm flow was generated from deep flow pathways below the fragipan during dry conditions and near surface flow pathways during wet conditions. Comparatively little research on runoff generation processes has been conducted in the Canadian Rocky Mountains, in part due to deep snow that is present for much of the year (October–May). While some studies have shown the importance of groundwater contributions to streamflow in alpine watersheds in the Rocky Mountains (Hood and Hayashi, 2015; McClymont et al., 2010), the additional complexity imposed by highly heterogeneous glacial till and permeable bedrock in sub-alpine and upper montane watersheds has limited more extensive research on runoff dynamics of this region. As a first attempt to conceptualize runoff generation in Alberta's Rocky Mountains, Spencer et al. (2019) quantified storage and precipitation–runoff relationships from hydrometric data. Results indicated that runoff generation was strongly governed by the interaction of two zones of storage – soil and till storage and bedrock storage. The alpine region and sub-alpine/upper montane region were also identified as two separate hydrologic response units that differed in timing and flow pathways for runoff response. While Spencer et al. (2019) developed a conceptualization of runoff generation for this region, they concluded that coupled flow and tracer approaches would be needed to reduce uncertainty in estimated flow contributions from each storage zone.

Chemical signatures of source water (rain, snowmelt, soil water, hillslope groundwater, till groundwater, and bedrock groundwater) and stream water can be used to determine which sources are contributing to streamflow during different flow conditions using end-member mixing analysis (EMMA; Christophersen and Hooper, 1992). The key assumptions for EMMA are that (1) the tracers are conservative, (2) the mixing process is linear, (3) source chemistry does not change temporally or spatially over the period or area studied (Inamdar, 2011; Hooper, 2003), and (4) all sources have been identified and have the potential to contribute to streamflow. Many studies have used EMMA to conceptualize when different geologic components are contributing to the stream (e.g., James and Roulet, 2006; Cowie et al., 2017; Ali et al., 2010). However, this approach has been most successful in smaller watersheds (1 km2) because of more constrained variation in source water at smaller spatial scales (Hoeg et al., 2000). Large watersheds could be characterized based on smaller sub-watersheds (James and Roulet, 2006) if the sub-watersheds are homogeneous. Others have concluded that where source water displays large variation or assumptions cannot be met, runoff processes should be described qualitatively (Inamdar et al., 2013; Hoeg et al., 2000; Correa et al., 2019).

To expand on recent work carried out in the same study area by Spencer et al. (2019), this study aims to advance the conceptualization of runoff generation in the Rocky Mountains in Alberta, Canada. The objectives of this study were to (1) characterize how sources of stream water (rain, snowmelt, soil water, hillslope groundwater, till groundwater, and bedrock groundwater) vary spatially, across four sub-watersheds of a Rocky Mountain watershed, and temporally, from spring snowmelt to the start of the next year's snow accumulation period, and (2) determine the relative contributions of source water to the stream from spring to fall for each sub-watershed. This study should help inform the current conceptualization of runoff generation in northern Rocky Mountain watersheds.

Star Creek watershed (10.4 km2; Fig. 1) is located in the eastern slopes of Canada's Rocky Mountains, a region with fractured sedimentary bedrock overlain by glacial till. Average annual precipitation was 720 mm at Star Main (1482 m above sea level – a.s.l.) and 990 mm at Star Alpine (1732 m a.s.l.; Spencer et al., 2019). The area-weighted average annual precipitation (2005–2018) was 950 mm, using the Thiessen polygon method and nine precipitation gauges at a range of elevations in and surrounding Star Creek; 50 %–60 % of the precipitation falls in the form of snow (Spencer et al., 2019). Soils are Eutric Brunisols (Canadian System of Soil Classification; also known as Eutric Cambisols in the Food and Agriculture Organization system) approximately 1 m deep, on average. Star Creek is underlain by unsorted and uncompacted glacial till, which is generally less than 3 m deep with an estimated total area of 2.4 km2 (AGS, 2004; Fig. 1). Some clay-rich till layers, likely from localized glacial ice melt features, occur intermittently throughout the watershed, resulting in heterogeneous and uneven distribution of glacial till throughout the watershed. Sedimentary geologic formations (Upper Paleozoic formation, Belly River–St. Mary Succession, and Alberta Group formation) are primarily composed of shale and sandstone (AGS, 2004) and are highly fractured due to folding and faulting (Waterline Resources Inc., 2013).

Star Creek includes two main sub-watersheds, namely Star East (3.9 km2; 1537–2628 m a.s.l.) and Star West (4.6 km2; 1540–2516 m a.s.l.). Unvegetated talus slopes (0.50 km2 in Star East and 0.53 km2 in Star West, digitized from orthoimages) and exposed bedrock form the upper portion of alpine zones in both sub-watersheds (Figs. 1 and 2). Talus slopes terminate in the alpine and transitional forested regions of the watershed, but streams or tributary features flowing from talus slopes have not been observed. There is also no evidence of permafrost, ice lenses, or rock glaciers, unlike in other Rocky Mountain regions (Cowie et al., 2017; Clow et al., 2003; Hood and Hayashi, 2015; McClymont et al., 2010). Star West has a larger alpine region, with cirque till deposits (estimated area of 0.14 km2; AGS, 2004), that includes a narrow marshy area proximal to the stream that holds water throughout the summer and drains into the main channel that is primarily bedrock in the upper reaches. The Star East alpine region is smaller and more constricted than in Star West (Fig. 2) and is comprised mostly of a grassy meadow, with the stream originating from springs where the water table reaches the soil surface and is incised in colluvium with large boulders. In the lower reaches, streams in both sub-watersheds are composed of a series of step pools incised in alluvium and colluvium, with some areas of exposed bedrock.

Two historical streamflow gauging sites exist in each sub-watershed – a lower site (Star West Lower – SWL; Star East Lower – SEL) near the confluence of the two sub-watersheds (1540 m a.s.l.) and an upper site (Star West Upper – SWU; Star East Upper – SEU) located at approximately 1690 m a.s.l. in the sub-alpine transition zone (Fig. 1). The sub-alpine and upper montane zones are dominated by sub-alpine fir (Abies lasiocarpa) and Engelmann spruce (Picea engelmannii) above forests dominated by lodgepole pine (Pinus contorta) at lower elevations (Dixon et al., 2014; Silins et al., 2009). Vegetation in upper and lower watersheds (Fig. 1) are distinguished by a transition between higher-elevation alpine heath/shrub vegetation and sub-alpine fir-dominated forests in the upper watersheds and lodgepole-pine-dominated forest in the lower watersheds.

End-member mixing analysis (EMMA) was used to visualize multivariate source water and stream chemistry by reducing the dimensionality of the data with principal component analysis (PCA; Christophersen and Hooper, 1992). In addition, there were multiple subjective decisions required prior to EMMA, such as choosing tracers/ions and defining sources. Bivariate plots and tracer variability ratio (TVR) were used to determine if tracers were appropriate to use in the analysis. First, a matrix of bivariate plots of stream chemistry data (ion concentrations), used most commonly in geographical hydrograph separations, was used to determine if ions were conservative in nature (Hooper, 2003). A linear relationship between tracers can be interpreted as a sign of conservative relationships. Second, TVR, used most commonly in sediment source apportionment studies, was used to determine if the difference in ion concentrations between groups was larger than the variation within a source group (Pulley et al., 2015). TVR was calculated using the following equation for each tracer and compared between each source group pair: 1x̃max-x̃minx̃min×100meanCVsource1,CVsource2, where x̃max is the maximum median tracer concentration of either source group, x̃min is the minimum median tracer concentration of either source group, and CV is the coefficient of variation (Pulley et al., 2015; Pulley and Collins, 2018). TVR should be greater than two to be considered appropriate for use in mixing calculations (Pulley and Collins, 2018), although, depending on the data set in question, a greater threshold may be adopted to make the tracer selection more stringent and to help reduce the numbers of tracers included in further data processing.

Box and whisker plots and linear discriminant analysis (LDA) were used to remove the subjectivity of the defining sources (Ali et al., 2010; Pulley and Collins, 2018). Box and whisker plots were used as a visual means of discriminating between sources. LDA was then used to determine if the combined sources exhibited sufficiently robust statistical separation (Pulley and Collins, 2018). LDA optimizes separation between the centroid of group clusters by partitioning the variation across each tracer and weighing that variation into two axes (reducing the dimensionality). Other statistical classification methods, such as hierarchical clustering or k-means clustering, were not appropriate because source categories were known a priori. The data were processed in R (R Core Team, 2014), using the lda function in the MASS package (Venables and Ripley, 2002) to reduce dimensionality and assess the separation visually and the klaR (stepwise function; Weihs et al., 2005) package to model the data and determine the ability to separate groups statistically. The stepwise function models the data while removing individual tracers iteratively. The backwards direction was used in an attempt to maintain the most tracers with the lda method and ability to separate criterion.

After the sources were characterized, the stream water was processed using principal component analysis (PCA; prcomp function in R; R Core Team, 2014) as a method of dimensionality reduction to create a two-dimensional (2D) mixing space (Christophersen and Hooper, 1992). Stream water was standardized (subtracting the mean and dividing by the standard deviation for each sampling point) for each tracer, to create equal variance between chemical components, and used to create a correlation matrix. PCA was conducted on the correlation matrix to calculate eigenvectors and eigenvalues. Standardized stream water was then projected into the end member mixing space by multiplying by eigenvectors. Ideally, two principal components (PCs) explained most of the variation in the data and were used to generate a 2D mixing space, which corresponds to three sources in EMMA (Hooper, 2003). Other studies have used the rule of one to determine how many dimensions, and therefore sources, should be used to create the mixing space (Ali et al., 2010; Barthold et al., 2011). For this study, the mixing space was set to two dimensions for ease of visualization but used all appropriate sources, as presented by Inamdar et al. (2013), to provide a full description of potential source contributions. Source water was then standardized, using stream water means and standard deviations for each ion, and projected into the 2D mixing space as defined by the stream water (Christophersen and Hooper, 1992; Hooper, 2003). Stream water sources should create an outer boundary or polygon around all stream water samples if all sources were correctly identified and adequately sampled.

Twice monthly stream water and source water samples collected in Star Creek, from April to October in 2014 and 2015, have been used here to conceptualize runoff generation in Alberta's Rocky Mountains. Results from this study allow for a detailed examination of temporal patterns in source water chemistry and a qualitative description of source water contributions to stream water. While our intention was a quantitative estimate of source water contributions to streamflow using an unmixing routine, two of the key assumptions for EMMA, the chemical composition of sources does not change (1) over the timescale considered or (2) with space (Hooper, 2003; Inamdar, 2011), were violated in this data set. Source water chemistry varied greatly across the watersheds. For example, when all hillslope samples from each sub-watershed were projected into the mixing space created by stream water at the watershed outlet (SM), large variability was evident between sites (Fig. 12). While there was some overlap between some sites (SWL and SEU), SWU was clearly different than the other hillslope samples. As a result, source water from within individual sub-watersheds was used to reduce the uncertainty associated with large spatial variability in source water chemistry. However, the variability within sites was also quite large. The CV of source water tracer concentrations was often larger than the CV of the stream water tracer concentrations (there should be little to no variation in source water over time; James and Roulet, 2006; Inamdar, 2011), particularly for K+. The occasions where source water CV was smaller than stream water CV for most ions were for seeps in SEU and SEL, bedrock groundwater in SWL and SEL, and hillslope and riparian water in SWU. Chemical signatures of source water have been shown to vary seasonally and annually (Rademacher et al., 2005) and spatially across sub-watersheds in southern Quebec, Canada (James and Roulet, 2006). As a result, James and Roulet (2006) suggested that only source water from within individual sub-watersheds of interest should be used in unmixing calculations. Inamdar et al. (2013) further argued that mixing proportions should not be calculated because multiple assumptions are often violated and can lead to significant errors in unmixing proportions. Rather, temporal and spatial variation in stream water and source water should be examined and used to describe or to develop a physically based conceptualization of runoff mechanisms.

The inability to run the unmixing routine (stream water fell outside the bounds of the source water) also hindered the use of some tracer selection methods. Other studies have often used the selection criteria presented in Barthold et al. (2011), but the unmixing routine is required for this method. Rather, the TVR and LDA have been presented as effective parameters to subjectively determine if tracers are included in the analysis and if sources are well separated or grouped appropriately, respectively (Pulley et al., 2015, Pulley and Collins, 2018, and others – see the comprehensive review in Collins et al., 2017).

Despite the violation of assumptions, notable temporal trends in source water chemistry were observed in snowmelt, riparian water, hillslope and soil water, and bedrock groundwater and their contributions to stream water can be generalized for all sub-watersheds in Star Creek in a number of ways. The water that was stored in the hillslope over winter (or reacted water) was likely the first to reach the stream in the early spring prior to high flow as snowmelt started to saturate the landscape (Figs. 6 and 7). Temporal patterns in stream water chemistry also showed a spike in concentrations of some ions (e.g., Ca2+ concentration in Fig. 11) in the stream in the early spring, as this reacted water mobilized prior to the onset of the snowmelt freshet. Although three snowmelt samples in 2014 showed similar ionic pulses early in the snowmelt season to those reported in the Colorado Rocky Mountains (Williams et al., 2009), the concentrations were notably less than from all other sources and, thus, not likely an important source of the observed early season increase in stream water concentration of some ions. Rather, the delivery of reacted water to the stream at the onset of snowmelt is likely similar to the flushing mechanism observed in the Turkey Lakes watershed in central Ontario, Canada (Creed and Band, 1998), where high nitrogen concentrations were observed prior to peak streamflow. McGlynn et al. (1999) observed the displacement of old water to the stream at the onset of snowmelt in the Sleepers River research watershed in Vermont, USA, and suggested this was due to a small volume of snowmelt being added to a large storage of water already in the subsurface.

This initial displacement of reacted water was followed by a dilution effect, where large volumes of low-concentration snowmelt mixed with soil water and contributed to streamflow. Snowmelt was the major event that produced a water table response in all hillslope wells and connected the hillslopes to the stream (Fig. 11; Spencer et al., 2019). The initial snowmelt period was also the only time that overland flow was observed at the study site. Other studies have also reported that snowmelt creates a dilution response in the stream (Rademacher et al., 2005; Cowie et al., 2017). Conversely, the opposite has been observed whereby a previously disconnected source was connected to the stream and caused an increase in solute concentrations (McNamara et al., 2005). Although this was the main period of hydrologic connectivity in Star Creek (Fig. 11), we did not observe an increase in stream water ion concentrations associated with newly connected sources. Hillslope groundwater and soil water chemistry should reflect the dilution from snowmelt and the subsequent drying of the hillslope, thereby increasing ion concentrations in the soil water from spring to fall. This corresponding temporal pattern in hillslope groundwater chemistry was observed in SWU (Fig. 7) and for soil water chemistry in SEL and SEU (Fig. 9).

Source water contributions to the stream were more similar within Star East (SEL and SEU) and Star West (SWL and SWU) sub-watersheds than between alpine/sub-alpine (SEU and SWU) and upper montane (SEL and SWL) sub-watersheds. PCA plots for SEL and SEU showed that stream water chemistry was most similar to precipitation in May and June, whereas a dilution effect from snowmelt occurred in June and July in SWL and May to July in SWU. It should be noted that although dilution was observed, stream water was less diluted than snowmelt alone. Snowmelt water interacts with the soil as it moves through the subsurface to the stream, directly influencing the chemical composition of the snowmelt contributions to streams (Sueker et al., 2000). The delayed response in SWL and SWU is consistent with the watershed storage estimates from baseflow recession analysis (Spencer et al., 2019) that suggested that the west fork sub-watersheds had a larger storage capacity than the east fork sub-watersheds. Accordingly, more water would be required to fill the storage before saturation or hydrologic connectivity could occur.

Differences in the east and west forks were also evident in the hysteresis pattern in stream water chemistry from spring to fall. Star West sub-watersheds had a counterclockwise pattern, whereas Star East sub-watersheds had a clockwise pattern. In general, this is an artefact of the PCA analysis driven by the specific ions that defined each PC (Table 1). In Star East, the first PCs were dominated by anions, and the second PCs were dominated by SO42- (negative relationship). While the first PCs for Star West were dominated by anions, the second PCs included a mix of anions and cations and SO42- with a positive correlation, thereby producing an opposite hysteresis pattern. Although this is an artefact of the PCA analysis, it was ultimately due to slight variations in the sources contributing to the streams at different times during the flow season. For instance, in SWL and SWU, stream water was chemically similar to riparian water in the fall (Fig. 6); whereas, in SEL and SEU, stream water was similar to hillslope groundwater in August but fell outside the boundaries created by the identified sources in September and October (Fig. 7). Details on the possible processes underlying these differences are described below.

Temporal variations in riparian water in SWU were observed from spring to fall and followed the same pattern observed in stream water chemistry in May compared to September/October (Fig. 7). It is not clear if the stream chemistry responded to variations in riparian chemistry or if riparian water responded to stream chemistry, but these pools of water were likely mixing to create similar temporal patterns rather than reflecting those of hillslope water chemistry influencing the stream. The timing of riparian and stream water level responses may be used to help clarify these patterns in future research. Other studies have shown the importance of the riparian zone for buffering stream water chemistry from inputs from other sources, particularly for individual hydrologic events (McGlynn and Seibert, 2003; Grabs et al., 2012). Further research needs to be conducted to estimate the extent of the riparian area and the potential volume of water that may contribute to streamflow in Star East compared to Star West.

A groundwater seep in SEL and SEU followed similar temporal patterns to stream water from spring to fall (Fig. 7) and may provide insights into the sources of stream water in September and October. Consistently cool, but low, specific conductivity of the groundwater seeps suggest they likely reflected a deeper bedrock groundwater source different than the bedrock groundwater well. Although most of the stream is situated within the same geologic formation, there may be differences in bedrock groundwater chemistry associated with heterogeneous sedimentary layers or contact time in the upper versus lower watershed (Freeze and Cherry, 1979). Temperature signals from other seeps suggested some were fed by shallow subsurface water or till groundwater (larger fluctuations in temperature; Fig. 5; Taniguchi, 1993), yet they had high specific conductivity and similar ion concentrations to bedrock and hillslope groundwater. This suggests that there are likely many complex subsurface flow pathways, making it difficult to differentiate between subsurface sources, but it is possible, therefore, that additional bedrock groundwater sources were contributing to streamflow in Star East in September and October. Other tracers, such as oxygen and hydrogen isotopes, or non-conservative tracers, such as nitrogen and dissolved organic carbon, may help to better differentiate between seeps, hillslope groundwater, and bedrock groundwater (e.g., Cowie et al., 2017; Ali et al., 2010; Orlova and Branfireun, 2014). Additional observation wells in the bedrock and till would be required to characterize more thoroughly the variability in groundwater across the watershed (Rinderer et al., 2014).

Topographic transitions and convergent zones have been associated with groundwater contributions to streamflow (Covino and McGlynn, 2007; Hjerdt et al., 2004). While minimal groundwater discharge occurred over the mountain front recharge zone in Humphrey Creek, southwest Montana, USA, considerable groundwater discharge was observed in the valley bottoms (Covino and McGlynn, 2007). Large increases in the concentration of stream water ions that may be associated with strong groundwater upwelling were not evident between April and October or along the length of the stream. However, chemical signatures of groundwater seeps suggest that some bedrock groundwater may not be distinguishable from stream water. Consequently, these transitions may not be visible in water chemistry along the length of the stream.

Although contamination of the till groundwater well limited the inferences that could be made from its water chemistry, water levels in the bedrock and till groundwater wells do provide some insights into potential contributions of till groundwater to streamflow. Water table depth in the till groundwater was more responsive than bedrock groundwater level in the spring, though the overall rise in water level in the bedrock was slightly greater. Despite the flashier response earlier in the year, till groundwater levels remained elevated longer than bedrock groundwater, resulting in a slower recession (slower drainage) in the till groundwater well in the summer (Fig. 13). Similar to the post-glacial landscape in the Sleepers River watershed (Shanley et al., 2015), slower drainage from till groundwater may be partly responsible for maintaining streamflow during late summer or fall. The temperature range of some seeps sampled along the stream length were similar to till groundwater, but ion concentrations were similar to hillslope and bedrock groundwater. It is likely that glacial till chemistry is similar to hillslope and bedrock groundwater, given that they are situated above and below the glacial till layers. Heterogeneous glacial till deposits with different physical characteristics were also linked to the variable release of stored water, and thus the variability in baseflow, in the Scottish Highlands (Blumstock et al., 2015). Glacial till in the Rocky Mountains can be highly spatially variable, likely promoting multiple flow pathways (Langston et al., 2011). Clay lenses can create perched water tables that have different response times to the rest of the till matrix (Evans et al., 2000) or create complex groundwater flow pathways (Freeze and Witherspoon, 1967). Further research is needed to help differentiate between bedrock groundwater and till groundwater and their contribution to stream water during low flows.

Stream and source water were collected over the 2014 and 2015 water years and visualized using principal components analysis to conceptualize the runoff generation processes in the Canadian Rocky Mountains. While strong variability in source water chemistry limited our ability to quantitatively estimate the relative contributions of multiple water sources to the stream using an unmixing routine, the analyses used here enabled a strong qualitative description of precipitation, hillslope water, and bedrock groundwater source contributions to streamflow. This allowed us to both indirectly observe and infer key runoff generation processes in watersheds with a complex lithological structure characteristic of the highly permeable bedrock and glacial till of the Alberta Rocky Mountain region.

Stream water chemistry in four sub-watersheds of Star Creek showed that Star East (SEL and SEU) and Star West (SWL and SWU) sub-watersheds were more similar than alpine/sub-alpine (SEU and SWU) and upper montane (SEL and SWL) physiographic zones. In general, higher-concentration reacted water reached the stream first at the onset of spring melt in all sub-watersheds. This was followed by a dilution effect as the snowmelt saturated the landscape, and the hillslope was connected to the stream. Fall baseflows differed between Star East and Star West forks. Star West stream water was once again similar to hillslope water or riparian water, but Star East stream water plotted outside the boundary of the measured sources. Seep water temperatures were cool and had low variability, suggesting that it may be deeper bedrock water contributing to the stream. Slower recession rates (and likely lower hydraulic conductivity) in the till groundwater well than in the bedrock groundwater well suggest that water recharged into the till groundwater may be slowly released to the stream. Contamination of the till groundwater well made it unclear when it was contributing to the stream, but groundwater table fluctuations suggested it is likely contributing during late summer or fall. More research on the variability in bedrock and till groundwater chemistry is needed to clarify the difference between these sources and their contributions to streamflow throughout the year. However, it is clear from this research that multiple subsurface flow systems lead to the slow leakage of bedrock and till groundwater to the stream, promoting higher baseflows in this region compared to regions with shallow soils and impermeable bedrock where groundwater stops flowing in the summer.