• Yukon River basin;
  • boreal;
  • dissolved organic matter;
  • hydrology;
  • permafrost


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[1] Groundwater discharge to rivers has increased in recent decades across the circumpolar region and has been attributed to thawing permafrost in arctic and subarctic watersheds. Permafrost-driven changes in groundwater discharge will alter the flux of dissolved organic carbon (DOC) in rivers, yet little is known about the chemical composition and reactivity of dissolved organic matter (DOM) of groundwater in permafrost settings. Here, we characterize DOM composition of winter flow in 60 rivers and streams of the Yukon River basin to evaluate the biogeochemical consequences of enhanced groundwater discharge associated with permafrost thaw. DOC concentration of winter flow averaged 3.9 ± 0.5 mg C L−1, yet was highly variable across basins (ranging from <1 to >20 mg C L−1). In comparison to the summer-autumn period, DOM composition of winter flow had lower aromaticity (as indicated by specific ultraviolet absorbance at 254 nm, or SUVA254), lower hydrophobic acid content, and a higher proportion of hydrophilic compounds (HPI). Fluorescence spectroscopy and parallel factor analysis indicated enrichment of protein-like fluorophores in some, but not all, winter flow samples. The ratio of DOC to dissolved organic nitrogen, an indicator of DOM biodegradability, was positively correlated with SUVA254and negatively correlated with the percentage of protein-like compounds. Using a simple two-pool mixing model, we evaluate possible changes in DOM during the summer-autumn period across a range of conditions reflecting possible increases in groundwater discharge. Across three watersheds, we consistently observed decreases in DOC concentration and SUVA254 and increases in HPI with increasing groundwater discharge. Spatial patterns in DOM composition of winter flow appear to reflect differences in the relative contributions of groundwater from suprapermafrost and subpermafrost aquifers across watersheds. Our findings call for more explicit consideration of DOC loss and stabilization pathways associated with changing subsurface hydrology in watersheds underlain by thawing permafrost.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[2] Permafrost temperatures have warmed across the northern hemisphere in recent decades in response to increasing air temperatures [Romanovsky et al., 2010]. Regional and local thawing have been documented in arctic and subarctic regions of Alaska [Jorgenson et al., 2001, 2006], Canada [Payette et al., 2004; Camill, 2005], and Russia [Oberman, 2008]. Thawing of permafrost will likely have a profound impact on watershed hydrology, and in particular, the magnitude and seasonality of river discharge from arctic and subarctic basins. For instance, Walvoord and Striegl [2007] document an increase in groundwater contribution to streamflow (base flow) of 0.7–0.9% yr−1over the last 30 years in the Yukon River Basin (YRB; Alaska and Canada), presumably due to basin-wide permafrost thaw and subsequent changes in watershed hydrology.Smith et al. [2007] report an increase in minimum daily flow for 111 rivers across Russia, reflecting an increase in groundwater mobilization associated with a changing ground thermal regime. Analyses by St. Jacques and Sauchyn [2009]of 23 rivers in the Northwest Territories (Canada) highlight a long-term increase in winter streamflow, consistent with post-thaw increases in soil water infiltration and depth of subsurface flow paths. Model simulations have projected widespread permafrost thaw over the next century [e.g.,Lawrence et al., 2008], suggesting continued enhancements in groundwater circulation and contribution to river discharge in arctic and subarctic basins [Walvoord et al., 2012]. Addressing the impact of such upward trends in base flow on the chemical composition of arctic and subarctic rivers is a primary focus of the study presented here.

[3] Permafrost-driven changes in subsurface flow paths and fluxes will influence the flux of dissolved organic carbon (DOC) from terrestrial to aquatic and marine ecosystems [Frey and McClelland, 2009], which is an important component of the carbon cycle in arctic and subarctic regions [McGuire et al., 2009, 2010]. In the YRB, the discharge-normalized flux of DOC has decreased during summer and autumn over the past 30 years [Striegl et al., 2005]. This observation has been attributed to increases in active layer thickness, depth of subsurface flows, and water residence time [Walvoord and Striegl, 2007; Lyon and Destouni, 2010]. As a result, DOC concentration in soil water and groundwater could be reduced through biological (i.e., enhanced mineralization with longer transit times) or physical processes (i.e., sorption to mineral surfaces [Kawahigashi et al., 2004, 2006]). Similar hypotheses have also been proposed to explain spatial differences in stream DOC concentration across watersheds with different permafrost extents [MacLean et al., 1999; Petrone et al., 2006]. This study builds upon previous hydrology and DOC studies in the YRB and adds to the overarching question of biogeochemical consequences of permafrost thaw by providing a spatial and seasonal analysis of dissolved organic matter (DOM) with which to improve future predictions and expand the understanding of anticipated shifts in DOM characteristics.

[4] In the present study, we examine in detail the chemical composition of DOM during winter flow of streams of the YRB. We assume that the chemical composition of winter flow (November through April) is reflective of a groundwater signal because freezing surface temperatures during this period prohibits surficial and shallow subsurface flow. Observed seasonal shifts in DOM composition are also examined and used as a basis for evaluating the potential biogeochemical impacts of a continued upward trend in groundwater contribution to total river discharge [Walvoord and Striegl, 2007]. We use a range of analytical approaches to characterize DOM, including optical properties (ultraviolet absorbance, fluorescence) and chemical fractionation. Using DOM chemistry in conjunction with stream and groundwater discharge data, we use a simple mixing model to evaluate the effects of changing groundwater discharge on stream DOM composition across future permafrost thaw scenarios. We also evaluate the utility of DOM in differentiating hydrologic conditions in watersheds underlain by discontinuous permafrost. Unlike many other arctic rivers that flow south to north with permafrost-free headwaters and downstream areas not yet affected by permafrost thaw, the Yukon River flows east to west draining large expanses of discontinuous permafrost, most vulnerable to thaw. This characteristic and the general lack of engineered flow control make the YRB an ideal, natural system to assess current and near future consequences of permafrost thaw. Conclusions drawn here for the YRB may provide valuable insight for anticipating future responses in other circumboreal systems.

2. Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

2.1. Study Sites

[5] The YRB covers 854,700 km2 of northwest Canada and Alaska, and primarily composes remote wilderness (Figure 1). Mean annual air temperature throughout the YRB ranges from −6 to −1°C (1971 to 2000, from Alaska Climate Research Center (ACRC), Precipitation in interior Alaska is spatially variable, averaging 235–380 mm yr−1 in the continental climate of interior Alaska, 400–500 mm yr−1 in the Yukon Delta area (1971–2000, ACRC), and greater than 1000 mm yr−1 in the headwaters in the St Elias Mountains [Striegl et al., 2007]. Much of the YRB is underlain by permafrost, which is actively warming and thawing throughout the basin [Osterkamp and Romanovsky, 1999; Osterkamp, 2007]. Approximately 30% of the YRB is covered by low-lying wetlands [Brabets et al., 2000]. Watersheds of the YRB vary with respect to hydrology, parent material and source water contributions to stream flow. Blackwater streams (e.g., Porcupine River) typically drainwatersheds underlain by permafrost and have extensive wetland coverage. In clearwater streams (e.g., Clearwater Creek), groundwater discharge dominates stream flow. In glacially fed streams (e.g., Tanana River), flow originates as meltwater from alpine glaciers and snowfields. Despite being glacially fed, groundwater discharge to the Tanana River accounts for more than 25% of annual flow [Walvoord and Striegl, 2007].


Figure 1. Map of sampling locations in the Yukon River basin of interior Alaska.

Download figure to PowerPoint

[6] Parent material underlying watersheds of the YRB is spatially variable. Nearly 75% of the Yukon –Tanana upland area is underlain by rocky colluvium, which is composed primarily of quartz-mica schists beneath a relatively thin loess mantle [Ping et al., 2006]. Thick loess deposits (>20–30 m thick) are also present, accumulating in unglaciated regions of the YRB during the late-Pleistocene and early Holocene [Muhs et al., 2003]. Alluvial deposits are common along floodplains and terraces of large glacial rivers (e.g., Tanana River), and are typically composed of a loamy cap above sandy and gravelly substrate [Ping et al., 2006]. Glacial deposits (e.g., outwash, till, moraine) are also common, particularly along the northern foothills of the Alaska Range [Ping et al., 2006].

[7] Permafrost extent in the YRB is largely discontinuous (50–90% of area) with localized regions of sporadic (10–50%) and isolated permafrost (<10%) in lowlying areas such as Tanana and Koyukuk Flats [Jorgenson et al., 2008]. Continuous permafrost (>90%) also exists in northern parts of the basin, such as in the Porcupine River basin. Thermokarst features have been documented throughout the YRB, with collapse-scar fens, bogs, and thermokarst lakes identified as the most common modes of thaw [Jorgenson et al., 2007]. Wildfire frequency and severity have increased in recent decades in Alaska's boreal region [Turetsky et al., 2011] and may exacerbate rates of permafrost thaw relative to climate warming alone [O'Donnell et al., 2011a].

2.2. Study Overview

[8] We describe the chemical composition of DOM during winter flow from 60 streams draining subcatchments of YRB and points along the main stem of the YRB (Figure 1). We also describe samples from two representative groundwater sources (Fox Spring, Tape Well, both located in or near Fairbanks). For this study, we define winter flow as the period spanning November 1 through April 30, following the convention used by Striegl et al. [2005]. For samples collected and processed between 2001 and 2010, data were available through the U.S. Geological Service National Water Information System (NWIS, Web interface: Water Data for the Nation; In March 2011, we collected winter streamflow samples from 25 additional sites to better characterize spatial variability in river DOM character throughout the YRB. To examine the unique chemical composition of DOM during winter, we compared and contrasted DOM properties during winter versus summer-autumn (July 1–October 31) for several representative sites (Porcupine, Salcha, and Tanana Rivers; Hess Creek; Yukon River at Eagle, Stevens Village and Pilot Station). Summer-autumn data were also obtained through the NWIS website. Under-ice samples were obtained using an ice-auger powered either by hand or by a Stihl 8D powerhead. Samples were characterized for DOC (n = 173 samples) and dissolved organic nitrogen (DON) concentrations (n = 57 samples), UV-visible absorbance (n = 173 samples at 254 nm; n = 47 samples for all other wavelengths), fluorescence (n = 72 samples), and XAD8/XAD4 fractionation (n = 104 samples). We also developed a simple mixing model to evaluate river DOM composition in response to projected changes in permafrost thaw and groundwater discharge.

2.3. DOC Concentration and DOM Chemical Fractionation

[9] DOC concentrations were determined using an O.I. Analytical Model 700 TOC Analyzer via the platinum catalyzed persulfate wet oxidation method [Aiken et al., 1992]. DON was calculated as the difference between total dissolved nitrogen (TDN) and inorganic nitrogen species (nitrate (NO3) + ammonium (NH4+)). TDN was determined using a Skalar Model Formacs Total Nitrogen Analyzer and NH4+was determined using Dionex Model DX-300 and DX-500 Ion Chromatographs with conductivity detectors following the methods ofSmith et al. [2005]. NO3 concentrations were either determined using ion chromatography or using colorimetric methods following Antweiler et al. [1996].

[10] Stream water samples were chromatographically separated into different fractions: hydrophobic acids (HPOA), hydrophobic neutrals (HPON), hydrophilic organic matter (HPI), and transphilic acids (TPIA) using Amberlite XAD-8 and XAD-4 resins [Aiken et al., 1992]. The resins preferentially sorb different classes of organic acids based on aqueous solubility of the solute, chemical composition of the resin, resin surface area, and resin pore size. The amount of organic matter within each fraction, expressed as a percentage of the total DOC concentration, was calculated using the DOC concentration and the sample mass of each fraction. UV-Vis absorbance was run on each of the major DOC fractions. The standard deviation for the mass percentages of the fractionation was ±2%.

2.4. Optical Properties of Chromophoric DOM

[11] For our analyses, we used several approaches for reporting and analyzing the optical properties of chromophoric DOM (CDOM). UV-Vis absorbance was measured at room temperature using a quartz cell with a path length of 1 cm on an Agilent Model 8453 photo-diode array spectrophotometer. Based on convention and ease of comparison with literature data, measured decadal absorbance values (A) were converted to absorption coefficients which can be expressed in both decadal units (a), as is common in the freshwater and wastewater literature, and Naperian units (α), which is the convention among marine chemists. The following equation shows the conversion to Naperian absorption coefficient.

  • display math

where A(λ) is the decadal absorbance at a specified wavelength (λ) and l is the cell path length in meters [Green and Blough, 1994]. CDOM absorption coefficients in Naperian units are reported at wavelengths of 254, 350, and 440 nm. Absorption coefficients at these wavelengths have been shown to correlate strongly with riverine DOC concentration, particularly where allochthonous organic matter inputs dominate [Spencer et al., 2012]. The absorption coefficient at 350 nm (α350), in particular, has been shown to be strongly correlated with lignin phenol concentrations [Spencer et al., 2008], which are derived from vascular plant material.

[12] Spectral slope (S) was calculated by fitting an exponential equation to the absorption spectra between 275 and 295, 290–350, and 350–400 nm using

  • display math

where ag(λ) is the absorption coefficient of CDOM at a specified wavelength, λref is a reference wavelength, and S is the slope fitting parameter [Helms et al., 2008; Spencer et al., 2008]. The spectral slope of the 275–295 nm region (S275–295) has been shown to be negatively correlated with the molecular weight of DOM [Helms et al., 2008]. Prior studies have shown the spectral slope between 275 and 295 nm (S275–295) to be sensitive to changes in DOM source (e.g., riverine versus estuarine versus open ocean and composition) [Spencer et al., 2007]. Slope ratio (SR; 275–295-nm slope:350–400-nm slope) was also calculated, and has also been shown to be strongly correlated with DOM molecular weight [Helms et al., 2008].

[13] Specific UV absorbance (SUVA254) was determined by dividing the decadal UV-Vis absorption coefficient per meter atλ = 254 nm by DOC concentration. SUVA254, which is typically used as an index of DOC aromaticity, provides an “average” absorptivity at λ = 254 nm for the DOC [Weishaar et al., 2003]. SUVA254 is reported in units of L mgC−1 m−1. Weishaar et al. [2003] also showed that UV absorbance values can be influenced by the presence of iron. To account for potential effects of iron on UV absorbance, we measured total iron concentrations on a subset of winter flow samples. We applied correction factors based on experimental work by B. Poulin and G. Aiken (unpublished data, 2012), who observed a significant positive correlation between α254 and the concentration of Fe3+, reflected by the equation A254-corrected = A254-measured 0.0687*[Fe3+] (R2 = 0.98; P < 0.0001). Using this relationship, we corrected UV absorbance values at 254 nm for samples where we measured total iron concentration. pH was also measured in the field on most samples, averaging 7.25 ± 0.60 (standard deviation, n = 126, range = 5.60–8.83).

[14] Fluorescence excitation-emission matrices (EEMs) were measured on select samples and a subset of major chemical fractions (HPOA, TPIA, HPI) at room temperature using a Jobin-Yvon Horiba Fluoromax-3 fluorometer. Samples were diluted to minimize inner filter effects with deionized water, when necessary, to a UV absorbance atλ= 254 nm of 0.2 absorbance units (1 cm cell). Using a 5-nm slit width, EEMs were collected over an excitation range of 240–450 nm every 5 nm, and an emission range of 300–600 nm every 2 nm. Scans were corrected for instrument optics, inner filter corrected, Raman area normalized, Raman normalized blank subtracted, and multiplied by the dilution factor if necessary [Murphy et al., 2010]. Fluorescence index (FI) was determined as the ratio of the intensities at excitation (ex) and emission (em) wavelengths ex370/em470 and ex370/em520 [Cory et al., 2010].

[15] EEMs were fit to the previously validated 13 component parallel factor analysis (PARAFAC) model presented in Cory and McKnight [2005]to identify the relative abundance of fluorescing components of the EEMs collected for this study. Two of the components (C8 and C13) were proposed by Cory and McKnight to be associated with protein-like fluorophores, whereas components C2 and C4 are associated with aquatic humic substances [McKnight et al., 2001; Cory and McKnight, 2005].

2.5. Modeling the Effects of Increasing Groundwater Discharge on DOM Composition

[16] We developed a simple mixing model to test the effects of increasing proportion of groundwater discharge to total streamflow on DOC concentration and DOM composition. We selected three rivers (Porcupine River, Salcha River, Tanana River) that differ with respect to DOM composition and the proportion of groundwater discharge to streams at present [Walvoord and Striegl, 2007; Brabets and Walvoord, 2009]. Our overall objective was to track DOM across a range of conditions reflecting projected increases in groundwater discharge. For the end-member mixing model, we solved the following equations

  • display math
  • display math

where [DOM] represents DOC concentration or a DOM chemical property in stream water (STREAM), groundwater (GW), or shallow soil water or runoff (SW), and f is the fraction of stream water derived from either GW or SW. We prescribed values of 0.05, 0.25, 0.50, 0.75 and 1.00 for fGW to capture the full range of potential groundwater fractions, though recognizing the unlikelihood of achieving close to 1.00 fGW, even for a totally permafrost-free condition. We focused our analysis on changes in chemical composition during the summer-autumn period (July through October), because the relative influence of groundwater on flow during the snowmelt period (May–June) is minimal.

[17] Findings from this exercise are intended to gain basic understanding of system behavior and place bounds on potential changes in DOM with increased groundwater discharge. Absolute values of predicted change have limitations due to the influence of assumptions required for analysis. First, we assume no change in summer-autumn flow, which is in agreement with historical records [Brabets and Walvoord, 2009] and supported by some climate projections that predict only modest increases for the YRB [Aerts et al., 2006]. Second, we assume that DOC and DOM properties respond linearly to increased groundwater discharge. Recent literature, however, suggests that a nonlinear response to warming-driven changes in DOC production, transformation, mineralization, and permafrost-carbon release may be expected [Christ and David, 1996; Neff and Hooper, 2002; Dutta et al., 2006; Waldrop et al., 2010]. Third, we assume that two end-members are sufficient to constrain summer-autumn conditions. In actuality, post-thaw changes in hydrology are likely very complex [e.g.,Carey and Woo, 2000], particularly in the discontinuous permafrost zone, and cannot be accounted for with this data set.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

3.1. Spatial and Temporal Patterns of DOC Concentration and SUVA254

[18] DOC concentration of winter flow averaged 3.94 ± 0.46 mg C L−1 (n = 173 samples; data are means ± one standard error) across 60 streams or sampling locations in the YRB (Table 1). DOC concentrations were generally less than 1 mgC L−1in groundwater-dominated and glacial streams of the upper Tanana River Basin (e.g., Clearwater Creek, Delta River) and of the southern Brooks Range (e.g., Middle Fork of the Koyukuk River). DOC concentrations were highest in Ketchum Creek (average = 29.3 mgC L−1) and in small streams of the Forty-Mile River basin (range = 25.4–61.6 mgC L−1). SUVA254 of winter flow averaged 2.2 ± 0.1 L mgC−1 m−1 (n = 169 samples) across 58 streams or sampling locations in the YRB (Table 1). SUVA254was lowest in groundwater-dominated streams, such as Clearwater Creek, which averaged 0.8 ± 0.1 L mgC−1 m−1. Glacial rivers (e.g., Delta River, Gerstle River, Johnson River) and samples collected from groundwater sources (e.g., Fox Spring, Tape Well) also had low SUVA254 values (0.7–1.0 L mgC−1 m−1). Winter flow at Hess Creek had the highest SUVA254 values, averaging 4.1 ± 0.2 L mgC−1 m−1 (n = 6 samples).

Table 1. Site Locations, Dissolved Organic Carbon (DOC) Concentrations, and Specific Ultraviolet Absorbance (SUVA254) Values for the Winter Flow Study Sites (2001–2011)
StreamLatitude (°N)Longitude (°W)Watershed Area (km2)DOC (mgC L−1)SUVA254 (L mgC−1 m−1)
Beaver Creek Watershed
Beaver Creek above Victoria Creek65.8056146.648633152.5 ± 0.42.1 ± 0.1
Beaver Creek near Michel Lake66.2217146.762261643.1 ± 0.22.1 ± 0.1
Beaver Creek below Fossil Cr.65.4928147.6370-0.91.3
Birch Creek Watershed
Birch Creek at Steese Highway Bridge65.7111144.333355685.8 ± 0.32.9 ± 0.2
Birch Creek above Forks66.2691145.4953-8.13.6
Ptarmigan Creek65.4396145.5271493.32.0
Twelve Mile Creek above Birch Creek65.4001145.74022313.22.4
Circle Hot Springs Creek65.4849144.6364<50.91.4
Birch Creek above Preacher Creek66.0977144.7474-9.53.1
Deadwood Creek65.5393144.7644213.91.8
Preacher Creek above Birch Creek66.1297144.8355-8.93.4
Boulder Creek65.5711144.8843812.41.1
Mammoth Creek65.5490145.18049614.22.2
Ketchum Creek65.5101144.6915-29.32.1
Chatanika River Basin
Upper Chatanika River65.2889146.38753421.61.8
Faith Creek65.2903146.54731582.82.1
Sourdough Creek65.2775146.64831764.00.9
Long Creek65.2201147.0759-3.41.8
Crooked Creek near Chatanika65.2044147.2271171.51.8
Chatanika River65.1922147.255311813.32.0
Kokomo Creek65.1745147.2838853.01.0
Forty-Mile River Basin
West Fork Denison River63.8899142.2365-61.62.6
Taylor Creek63.4229142.4859-25.42.5
Hess Creek Watershed
Erickson Creek65.5740148.94216810.42.7
Richardson Creek65.6497149.0829-3.72.3
Hess Creek at Dalton Highway Bridge65.6656149.0968171515.7 ± 0.74.1 ± 0.2
Koyukuk River Basin
Koyukuk River, Middle Fork67.4405150.0816-1.00.1
Dalton North67.7991149.8034-3.62.7
Porcupine River Basin
Porcupine River near Fort Yukon66.9906143.1378764021.9 ± 0.12.1 ± 0.2
Sheenjek River66.7400144.5367-1.71.5
Porcupine River 9.5 Miles upstream from mouth66.6518145.11321085173.7 ± 0.32.1 ± 0.1
Tanana River Basin - Blackwater
Shaw Creek64.2605146.1078-5.2-
Salcha River64.4700146.931056202.4 ± 0.22.4 ± 0.1
Chena River at Two Rivers64.8460146.968724272.3 ± 0.21.8 ± 0.1
Little Chena River64.8565147.40539632.4 ± 0.12.6 ± 0.6
Chena River at Nordale Rd64.8465147.409837812.6-
Goldstream Creek64.9117147.83202505.61.6
Cripple Creek near Ester64.8489147.9229-2.8 ± 0.02.6 ± 0.1
Tolovana River65.4714148.267136010.6 ± 1.73.1 ± 0.5
Tatalina River65.3293148.30781994.02.7
Washington Creek65.1509147.85661217.23.0
Tanana River Basin - Glacial/Clearwater
Johnson River63.7016144.6398-0.71.0
Gerstle River63.8071144.9370-1.50.8
Clearwater Creek at Campground64.0540145.4333-0.7 ± 0.00.8 ± 0.1
Delta River above Tanana River confluence64.1517145.8450-0.6 ± 0.10.8 ± 0.4
Tanana River above Delta64.1567145.8484349640.8 ± 0.31.1
Tanana River at Nenana64.5653149.0917663011.4 ± 0.02.0 ± 0.2
Yukon River Mainstem
Yukon River at Eagle64.7894141.78942939542.1 ± 0.102.1 ± 0.1
Yukon River above Circle65.7264144.0578-2.02.3
Yukon River at Fort Yukon66.5613145.2718-1.12.3
Yukon River near Stevens Village65.8756149.71785083972.0 ± 0.12.0 ± 0.1
Yukon River at Pilot Station61.9344162.88068313583.0 ± 0.12.4± 0.1
Miscellaneous Sites
American Creek at Eagle64.7056141.30831742.9 ± 0.12.2 ± 0.1
Chandalar River67.0969147.1844241641.21.1
Fox Spring64.9640147.6248-1.40.7
Tape Well64.89712147.81114-2.61.3
Fort Hamlin Hills66.0333150.1333-3.93.0
No Name Creek66.0333150.13333115.0 ± 2.73.2 ± 0.2
Dalton Central66.6167150.6917-1.52.6
Fish Creek66.5422150.7971-1.62.5

[19] We compared DOC concentration and SUVA254 values of winter flow (November 1 through April 30) against flow during summer and autumn (July 1 through October 31) for six study sites where groundwater discharge has changed in recent decades (Figures 2a and 2b). Across all six study sites, DOC concentration was greater during summer-autumn than winter (Figure 2a). Seasonal differences in DOC concentration were most pronounced in the Porcupine River, with summer-autumn and winter averaging 8.22 ± 0.64 and 1.93 ± 0.06 mgC L−1, respectively. In the Tanana River, seasonal differences were less pronounced, with summer-autumn and winter averaging 2.82 ± 0.33 and 1.40 ± 0.04 mgC L−1, respectively. We observed larger differences in DOC concentration between summer-autumn and winter periods at sites where the contribution of groundwater to annual flow was relatively low (Porcupine and Salcha Rivers). The difference in SUVA254values was relatively consistent between summer-autumn and winter flows (Figure 2b), averaging 3.0 ± 0.2 and 2.2 ± 0.0 L mgC−1 m−1, respectively, across all six study sites.


Figure 2. Comparison of winter and summer-autumn (a) DOC concentration, (b) SUVA254, (c) HPOA %, and (d) HPI % values for six study sites in the Yukon River basin. We define seasons in the YRB following the conventions of Striegl et al. [2005], with winter samples collected between November 1 to April 30, and summer-autumn samples collected between July 1 and October 31. The percentages under Figure 2c are present-day groundwater contributions to annual flow, and the percentages under Figure 2d reflect recent annual changes (% yr−1) in groundwater contributions to streamflow as reported by Walvoord and Striegl [2007].

Download figure to PowerPoint

3.2. Spatial and Temporal Patterns in Major Chemical Fractions of DOC

[20] HPOA was the dominant chemical fraction of DOC during winter flow, averaging 43 ± 1% across 26 study sites (n = 104 samples; Table 2). HPI and TPIA compose a smaller proportion of total DOC, averaging 25 ± 1% and 20 ± 0%, respectively, across all study sites. In spite of these general patterns, we observed considerable and unexpected variability in DOM composition of winter flow across our study sites. For instance, the HPOA fraction of winter flow was highly variable across study sites, with Hess Creek averaging 55 ± 0% and Clearwater Creek averaging 23 ± 3%. The HPI fraction of winter flow was highest at the Tanana River at Nenana (44 ± 4%) and lowest at Hess, Beaver, and Birch Creeks, which averaged 16 ± 3%, collectively. The TPIA fraction was less variable across sites, ranging from 15 to 26%.

Table 2. Major Chemical Fractions of Dissolved Organic Carbon (DOC) and Fraction-Specific Specific Ultraviolet Absorbance (SUVA254) Values for Winter Flowa
SiteHPOA (%)HPOA SUVA254 (L mgC−1 m−1)TPIA (%)TPIA SUVA254 (L mgC−1 m−1)HPI (%)HPI SUVA254 b (L mgC−1 m−1)
  • a

    Hydrophobic acids (HPOA); transphilic acids (TPIA); hydrophilic organic matter (HPI).

  • b

    Some HPI SUVA254 values were omitted due to the apparent influence of iron on UV absorbance values.

Beaver Creek above Victoria44 ± 22.4 ± 0.716 ± 41.5 ± 0.419 ± 61.2 ± 0.4
Beaver Creek near Michel L.47 ± 22.4 ± 0.115 ± 41.4 ± 0.416 ± 51.7 ± 0.5
Birch Creek above Preacher522.9202.217-
Birch Creek at Steese Highway Bridge49 ± 12.7 ± 0.918 ± 12.2 ± 0.016 ± 2-
Boulder Creek492.4201.820-
Chatanika River432.4201.9280.9
Chena River at Two Rivers38 ± 42.5 ± 0.223 ± 12.2 ± 0.027 ± 2-
Clearwater Creek23 ± 31.8 ± 0.218 ± 21.8 ± 0.340 ± 51.4 ± 0.4
Delta River24 ± 02.2 ± 0.026 ± 11.8 ± 0.1--
Hess Creek55 ± 03.3 ± 0.020 ± 12.6 ± 0.416 ± 1-
Little Chena River48 ± 32.7 ± 0.123 ± 22.2 ± 0.126 ± 0-
Mammoth Creek482.7171.9211.2
No Name Creek443.4162.429-
Porcupine River 9.5 miles upstream47 ± 22.7 ± 0.420 ± 12.1 ± 0.221 ± 11.3 ± 0.1
Porcupine River near Fort Yukon41 ± 22.9 ± 0.120 ± 32.7 ± 0.627 ± 21.2 ± 0.1
Preacher Creek492.9232.218-
Salcha River44 ± 22.6 ± 0.021 ± 22.4 ± 0.232 ± 11-
Tanana River at Nenana41 ± 22.7 ± 0.125 ± 42.0 ± 0.144 ± 41.2 ± 0.2
Tolovana River49 ± 33.6 ± 0.221 ± 12.9 ± 0.219 ± 2-
Twelve Mile Creek433.0212.1260.9
Washington Creek483.4152.327-
Yukon River at Eagle49 ± 53.1 ± 0.423 ± 32.1 ± 0.033 ± 81.8 ± 0.1
Yukon River at Pilot Station47 ± 02.9 ± 0.019 ± 02.0 ± 0.123 ± 21.7 ± 0.1
Yukon River near Stevens43 ± 22.7 ± 0.021 ± 22.0 ± 0.136 ± 41.2 ± 0.1

[21] We further characterized CDOM by measuring UV absorbance and calculating SUVA254 values for the major chemical fractions of DOM (Table 2). SUVA254 values for the HPOA fraction were quite variable, with the lowest values observed at Clearwater Creek (1.8 ± 0.2 L mgC−1 m−1) and the highest at the Tolovana River (3.6 ± 0.2 L mgC−1 m−1). SUVA254 values for the HPI fraction were also variable across sites, ranging from 1.2 to 1.8 L mgC−1 m−1.

[22] In general, the HPOA fraction comprised a larger percentage of DOC during summer-autumn than during winter (Figure 2c). Seasonal differences were most pronounced at the Porcupine and Salcha Rivers, where HPOA averaged more than 50% during summer-autumn and less than 45% during winter. Differences in the percentage of HPOA were less pronounced between summer-autumn and winter at the Yukon River at Eagle. The HPI fraction comprised a larger percentage of DOC during winter than during summer-autumn flow (Figure 2d). Seasonal differences in HPI were most pronounced in the Tanana River at Nenana, where HPI averaged 24 ± 1 and 44 ± 5% during summer-autumn and winter, respectively. Interestingly, HPI during winter was more variable (coefficient of variation, or CV, = 0.34) than during summer (CV = 0.15).

3.3. DOM Fluorescence Properties of Winter Flow

[23] Excitation-emission spectra revealed distinct fluorescence characteristics of CDOM during winter flow when compared to samples collected during summer (Figure 3). In the Porcupine River (Figures 3a and 3b), we observed a distinct peak in the tryptophan-like region that originates between an excitation maximum of 270–280 nm and an emission maxima of 300–325 nm, and has previously been associated with the presence of proteins [Coble et al., 1998]. By contrast, the tryptophan-like peak is absent in the summer-autumn sample from the Porcupine River. In the Tanana River (Figures 3c and 3d), we observed two distinct peaks originating at an emission maximum of 350 nm: one at excitation maxima between 290 and 310 nm, and the other between 230 and 250 nm. It remains unclear what the composition and origin of these two peaks are, but they are present in the winter samples (n = 2) and absent in summer-autumn samples (n = 4) of the Tanana River. In Hess Creek (Figures 3e and 3f), we did not observe any notable differences in CDOM fluorescence properties between winter and summer-autumn samples.


Figure 3. Excitation-emission matrices (EEMs) comparing summer-autumn and winter stream samples from (a, b) the Porcupine River, (c, d) the Tanana River at Nenana, and (e, f) Hess Creek.

Download figure to PowerPoint

[24] Applying the Cory and McKnight [2005]13-component PARAFAC model to our EEMs, we were able to detect compositional differences in winter flow samples across 54 study sites (Table 3). Components C2 and C4, associated with aquatic humic substances, comprised the largest proportion of fluorescent CDOM accounting for 17 ± 1 and 19 ± 3%, respectively. On average, protein-like components (sum of components C8 and C13) accounted for 8 ± 6%. Protein-like components appear to comprise a larger proportion of fluorescing CDOM during winter than during spring snowmelt (mean = 3.1 ± 2%) or the summer-autumn period (mean = 2.7 ± 0.1%) [O'Donnell et al., 2010]. The protein-like components were also more variable across winter flow study sites than the humic-like components. The CV of components C8 and C13 were 0.98 and 0.81, respectively, whereas the range of CV values for the other 11 components ranged from 0.09 to 0.39. We observed very low proportion of protein-like components (∼4%) in several streams, including Beaver Creek, Birch Creek, and the Porcupine River. The highest proportion of protein-like compounds was observed in the Tanana River at Nenana, accounting for 35% of the fluorescing CDOM.

Table 3. Summary of Fluorescence Components Based on Cory and McKnight [2005] for Winter Flow Samples Quantified Through PARAFAC Analysis
Site NameC1 (%)C2 (%)C3 (%)C4 (%)C5 (%)C6 (%)C7 (%)C8 (%)C9 (%)C10 (%)C11 (%)C12 (%)C13 (%)FI
Beaver Creek above Victoria Creek10176195672458931.54
Beaver Creek below Fossil Cr.9176205572457931.51
Beaver Creek near Michel Lake91761946724581031.54
Birch Creek above Forks91652256733571021.50
Birch Creek above Preacher Creek10175205672358931.49
Birch Creek at Steese Highway Bridge10176195662358931.47
Boulder Creek91541835472681161.46
Chandalar River91771946624571031.57
Chatanika River10176195762368941.49
Circle Hot Springs Creek81661934672471261.51
Clearwater Creek81752255642571041.46
Crooked Creek near Chatanika11175216671367731.45
Dalton Central71741945451581271.41
Dalton North9195195573358941.46
Deadwood Creek8185195582458931.55
Dietrich River5147122141614615141.54
Erickson Creek81642157532681141.41
Faith Creek10175195663368841.43
Fish Creek11186195672457821.48
Fort Hamlin Hills8194225671358831.44
Fox Spring816131423648571321.84
Gerstle River81651744572781171.47
Goldstream Creek81761847523591141.53
Hess Creek at Dalton Highway Bridge10175215563357851.41
Johnson River91742045552681041.47
Ketchum Creek9175185762378941.43
Kokomo Creek814317453111778101.37
Koyukuk River, Middle Fork81781522674571181.58
Little Chena River715521466103471201.53
Long Creek9166205553358961.43
No Name Creek9184215753258951.40
Porcupine River 9.5 Miles upstream10175205771358831.48
Porcupine River at Yukon River91751957623581051.46
Porcupine River near Fort Yukon91661856622471261.48
Preacher Creek above Birch Creek91752156723571031.49
Ptarmigan Creek10174196663288841.39
Richardson Creek9174215761369941.43
Salcha River at Richardson Hwy11186195672358921.47
Sheenjek River91671946624581041.55
Sourdough Creek413323202195531191.40
Tanana River above Delta91741848623571161.45
Tanana River at Nenana311169203800220271.51
Tape Well41482722385341901.49
Tatalina River8195215572358921.49
Taylor Creek9174185763289841.43
Tolovana River9184205754258951.40
Twelve Mile Creek above Birch Creek10174206762268841.37
Washington Creek9184205763258951.40
West Fork of the Denison River10174205860269931.43
Yukon River above Birch Creek91761957613681151.45
Yukon River at Eagle91651956522581071.43
Yukon River at Pilot Station10185175673368931.41
Yukon River near Stevens Village91752057613581041.47
Yukon River, South Channel, near Fort Yukon91851957613681041.46
Standard Deviation21231214111240.07

[25] The fluorescence index (FI) was used to infer DOM origin (terrestrial versus microbial) for both whole-water samples and major chemical fractions of DOM. FI averaged 1.47 ± 0.07 across 54 winter flow study sites (Table 3), which reflects a mixture of terrestrial- and microbially derived CDOM [McKnight et al., 2001]. The highest FI value was observed at Fox Spring (1.84), a groundwater well near Fairbanks, Alaska, reflecting a potentially higher contribution of microbially derived organic matter to the DOC pool [McKnight et al., 2001]. The FI values for major chemical fractions of DOM (HPOA, HPI, TPIA) were remarkably consistent across winter flow study sites (Figure 4a). FI values for the HPOA fraction averaged 1.39 ± 0.03 across four large watersheds, whereas FI values for the HPI and TPIA fractions averaged1.57 ± 0.13 and 1.59 ± 0.03, respectively. These findings indicate that the HPOA fraction is closely linked to terrestrial organic matter sources. We also observed a negative correlation between fraction-specific FI and fraction-specific SUVA254 values, indicating a strong link between DOM origin and aromaticity for winter flow samples (Figure 4b).


Figure 4. (a) Comparison of fraction-specific fluorescence index (FI) for four study sites with the Yukon River basin. (b) The relationship between fraction-specific FI and fraction-specific SUVA254 for all winter flow sites. Error bars represent ± one standard error.

Download figure to PowerPoint

3.4. Chromophoric DOM Composition of Winter Flow

[26] The absorption coefficient at 350 nm (α350) has been shown to serve as an inexpensive proxy for lignin phenol concentration in both marine [Hernes and Benner, 2003] and freshwater environments [Spencer et al., 2008]. For winter flow samples in the YRB, α350 averaged 9.87 ± 2.87 m−1, with values ranging from 0.11 to 68.32 m−1 (Table 4). The lowest values were associated with groundwater- (e.g., Clearwater Creek) and glacial-dominated streams and rivers (e.g., Johnson River), whereas the highest values were associated with blackwater streams (e.g., Hess Creek, Taylor Creek, West Fork of the Denison River). Using the relationship betweena350 and lignin phenol concentration reported by Spencer et al. [2008], our findings indicate that lignin phenol concentrations of winter flow range from less than 1 to more than 120 μg L−1 across all study sites. The spatial variability in a350 for winter flow samples across the streams of the YRB is similar in magnitude to the temporal variability in a350 for the Yukon River near its mouth (Yukon River at Pilot Station) [Spencer et al., 2008].

Table 4. Summary of Chromophoric DOM Properties in Winter Flow
Sitea254 (m−1)a350 (m−1)S275–295 (×10−3 nm−1)SR
Beaver Creek Watershed
Beaver above Victoria Creek12.07 ± 2.342.27 ± 0.2616.32 ± 0.810.95 ± 0.15
Beaver below Fossil Creek2.590.2914.640.7
Beaver near Michel Lake31.80 ± 15.213.89 ± 0.5216.20 ± 0.400.99 ± 0.05
Birch Creek Watershed
Birch above Forks66.3920.0513.260.76
Birch above Preacher Creek67.6217.7515.130.82
Birch at Steese Hwy Bridge38.03 ± 1.379.4614.610.76
Boulder Creek6.030.8323.69-
Circle Hot Springs Creek2.870.8318.590.98
Deadwood Creek15.692.8219.310.98
Ketchum Creek144.9529.3517.210.92
Preacher Creek69.6220.6913.480.78
Ptarmigan Creek15.203.0417.400.86
Twelve Mile Creek16.963.3718.230.90
Chandalar River3.020.3318.790.83
Chatanika River Watershed
Chatanika River14.742.7617.910.88
Crooked Creek6.431.314.590.76
Faith Creek13.983.0515.600.88
Kokomo Creek7.151.1519.671.14
Long Creek13.583.4816.200.96
Sourdough Creek8.592.814.40-
Forty-Mile River Basin
Taylor Creek143.8833.0215.650.88
West Fork of the Denison River373.2668.3219.000.95
Groundwater Wells/Springs
Fox Spring2.240.4231.560.81
Tape Well7.452.2410.780.99
Hess Creek Watershed
Erickson Creek65.0722.6412.70-
Hess Creek128.59 ± 23.6052.0811.700.63
Richardson Creek19.124.5115.950.96
Porcupine River Basin
9.5 miles upstream from mouth18.21 ± 2.223.13 ± 0.3517.38 ± 0.790.85 ± 0.02
Sheenjek River5.951.0318.640.94
Tanana River Basin - Blackwater
Goldstream Creek20.803.3419.650.96
Little Chena River14.46 ± 2.631.7118.811.11
Tanana River Basin - Glacial/Groundwater
Clearwater Creek1.12 ± 0.180.11 ± 0.0039.31 ± 25.320.56
Gerstle River2.820.5525.75-
Johnson River1.310.1821.080.94
Tanana River near Delta2.030.3816.270.98

[27] S275–295 averaged 18.32 ± 1.24 × 10−3 nm−1 for winter flow samples across all study sites, which is greater than the mean range for samples reported for 30 watersheds in the United States (13–16.50 × 10−3 nm−1) [Spencer et al., 2012]. The steeper spectral slope of YRB samples suggests that on average, CDOM of winter flow is of lower molecular weight and aromaticity than the majority of riverine CDOM in the U.S. The shallowest S275–295 values (<15 × 10−3 nm−1), which are associated with high molecular weight and aromatic DOM, were observed at Hess Creek, Erickson Creek, and some sites within the Birch Creek watershed. The steepest S275–295 values (>20 × 10−3 nm−1) were observed at groundwater-dominated sites (Clearwater Creek, Fox Spring) and glacial rivers (Gerstle and Johnson Rivers).S290–350 averaged 17.40 ± 0.63 × 10−3 nm−1 and generally followed the same patterns across winter flow study sites as S275–295. SR, which is negatively correlated with DOM molecular weight and aromaticity [Helms et al., 2008], averaged 0.91 ± 0.03 across all winter flow samples (n = 38). SR values ranged from 0.56 to 1.14 for winter flow, which encompasses nearly the entire range of SR values reported by Spencer et al. [2012] for the conterminous United States across the hydrograph.

3.5. Relationships Among Different DOM Properties

[28] The ratio of DOC to DON was positively correlated with SUVA254 values for winter flow samples collected across a range of subcatchments within the YRB (R2 = 0.88, P = 0.002; Figure 5a). DOC:DON and SUVA254 values were highest at Hess Creek, averaging 37 ± 9 and 4.0 ± 0.9 L mgC−1 m−1, respectively. Winter flow samples collected from glacial- and groundwater-dominated rivers of the Tanana River basin were characterized by the lowest DOC:DON and SUVA254 values, averaging 12 ± 1 and 0.9 ± 0.l L mgC−1 m−1, respectively. Water collected from groundwater wells had similar chemical composition, at least with respect to DOC:DON ratio and SUVA254, as winter flow samples from the Tanana River basin. However, FI values differed between winter flow samples from the Tanana River basin and groundwater well samples, averaging 1.46 ± 0.02 and 1.66 ± 0.24, respectively. DOC:DON ratio was also negatively correlated to the percentage of protein-like compounds (C8 + C13), as determined by PARAFAC analysis of fluorescence EEMs (exponential equation: DOC:DON = 13 + 359e−0.73*%Protein, R2 = 0.43, P < 0.0001; Figure 5b).


Figure 5. (a) Linear regression relating mean DOC:DON ratio and mean SUVA254values for different watersheds within the YRB. Error bars represent ± one standard error. Groundwater samples were collected from Fox Spring, Tape Well and all samples from Clearwater Creek near Delta Junction, AK. (b) Exponential relationship between DOC:DON ratio and the percentage of protein-like compounds, as determined by PARAFAC analysis of fluorescence EEMs.

Download figure to PowerPoint

[29] In the YRB, CDOM parameters have been linked to DOC concentration and DOM composition [Spencer et al., 2008, 2009], creating the potential for improved flux estimation using in situ monitoring of optical properties [Saraceno et al., 2009; Pellerin et al., 2012]. For instance, Spencer et al. [2009] observed a strong relationship between DOC concentration and absorption coefficients across several wavelengths for the samples collected across the hydrograph at Yukon River at Pilot Station (site 60; Table 1). Here, we examined the relationship between DOC concentration and Naperian absorption coefficients at 254, 350, and 440 nm (Figure 6) for winter flow across study sites. We observed strong relationships between DOC concentration and these CDOM parameters (R2 values ranging from 0.76 to 0.89 at P < 0.0001) across all winter flow study sites with available data. Correlations were stronger at lower wavelengths and were characterized by steeper slopes at higher wavelengths.


Figure 6. Linear regression between DOC concentration and a254, a350, and a440 for winter flow samples across study sites. Linear equations and statistics describing these equations are as follows: a254: DOC = 0.6929 + 0.1325 a254 (R2 = 0.89, P < 0.0001); a350: DOC = 0.6865 + 0.6595 a350 (R2 = 0.81, P < 0.0001); and a440: DOC = 0.6532 + 3.3538 a440 (R2 = 0.76, P < 0.0001).

Download figure to PowerPoint

3.6. Sensitivity of DOM to Permafrost Thaw and Increasing Groundwater Discharge

[30] Using a two-pool mixing model (equations (3) and (4)), we explore potential changes in DOC concentration and DOM composition to possible changes in groundwater discharge associated with permafrost thaw (Figure 7). Based on the model results, DOC concentration is predicted to decrease with increasing fraction of streamflow derived from groundwater discharge during the summer-autumn period (July through October;Figure 7a). Our findings suggest that the magnitude of decrease in DOC concentration will likely vary spatially throughout the basin. For instance, in the Porcupine River, DOC concentration could decline by nearly 80% from a present-day average of 8.2 ± 2.6 mg C L−1 (where fGW is 0.05) to an absolute minimum of 1.9 ± 0.1 mg C L−1 (where fGWis 1.00). By contrast, DOC concentration in the Tanana River could decline no more than ∼50% from a present-day average of 2.8 ± 1.5 mg C L−1 (where fGW is 0.16) to 1.4 ± 0.1 mg C L−1 (where fGW is 1.00). SUVA254 values also decline with increasing groundwater fraction, and the magnitude of change is relatively consistent across the three representative catchments (Figure 7b). In the Porcupine River, SUVA254decreases from a present-day average of 3.1 ± 0.4 (wherefGW is 0.05) to 2.1 ± 0.3 L mgC−1 m−1 (where fGW is 1.00). The HPI fraction increases with increasing groundwater discharge (Figure 7c). In the Porcupine River, we estimate modest maximum increases in the HPI fraction, ranging from 18 ± 2% (where fGW is 0.05) to 27 ± 4% (where fGWis 1.00). In the Tanana River, we estimate a more substantial change, with HPI fraction increasing from 20 ± 5% under present-day conditions to 44 ± 8% (wherefGW is 1.00).


Figure 7. Results from simple two-pool mixing analysis exploring the sensitivity of (a) DOC concentration, (b) SUVA254, and (c) HPI during summer-autumn flow to changes in the fraction of total streamflow derived from groundwater discharge. Error bars represent ± one standard error for present-day groundwater discharge and for the 100% treatment.

Download figure to PowerPoint

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

4.1. Dissolved Organic Matter Composition of Winter Flow

[31] DOM compositions associated with winter flow in rivers of the YRB are unique when compared to the summer-autumn period. Our findings illustrate several consistent trends when comparing DOM composition of winter flow and summer-autumn flow across basins. First, DOM is less aromatic during winter than during summer and autumn, as reflected by the temporal variation in SUVA254 values. This observation is in agreement with prior studies of the Yukon River main stem [Striegl et al., 2005], major tributaries of the Yukon River [Striegl et al., 2007; Spencer et al., 2008], smaller catchments of the YRB [O'Donnell et al. 2010], and several rivers in the Russian arctic [Neff et al., 2006; Mann et al., 2012]. Second, HPOA comprises a smaller fraction of DOC during winter than during summer and autumn. Lignin-derived compounds, such as lignin phenols which are biomarkers of terrestrial vascular plants [Spencer et al., 2008], are a major constituent of the HPOA fraction [Templier et al., 2005]. Sorption may result in the selective depletion of the HPOA fraction [Cronan and Aiken, 1985], which has been documented in permafrost soils of the Siberia tundra [Kawahigashi et al., 2006]. However, microbial degradation may also play an important role in reducing the fractional amount of HPOA comprising riverine DOM. Third, HPI comprises a larger fraction of DOC during winter than during summer and autumn. The HPI fraction is typically associated with high rates of DOC mineralization [Wickland et al., 2007]; however, its relative enrichment in winter flow suggests that HPI can be preserved in groundwater over decades to centuries [Michel et al., 2007]. The TPIA fraction which is of intermediate polarity and distinct composition in relation to HPOA and HPI [Templier et al., 2005], was less variable across study sites. Together, these findings indicate a seasonal shift in the chemical structure and reactivity of DOM across seasons, as driven by changes in source-water contributions to streamflow [O'Donnell et al., 2010].

[32] In spite of these consistent differences in DOM across seasons, we also observed considerable and unexpected spatial variability in DOM composition in winter flow across streams of the YRB. DOC concentration and CDOM properties of winter flow varied widely, spanning the entire range of previously reported values for the YRB [Spencer et al., 2008, 2009]. Some of the variability in DOM composition of winter flow is clearly governed by spatial variation in source water inputs and hydrology across watersheds (e.g., blackwater versus glacial versus groundwater-fed streams) [O'Donnell et al., 2010]. It is likely that this spatial variation across watersheds is also driven, in part, by the relative contributions of groundwater from suprapermafrost and subpermafrost aquifers [Bense et al., 2009; Ge et al., 2011; Walvoord et al., 2012]. Prior groundwater investigations in permafrost regions have identified a range of groundwater sources contributing to stream discharge that vary with respect to inorganic geochemistry [Clark et al., 2001]. Our findings indicate that dissolved organic matter composition may also be useful for distinguishing groundwater sources, particularly shallow and deep flow paths.

[33] While prior studies have documented the importance of protein-like compounds in winter flow relative to summer-autumn flow [O'Donnell et al., 2010], our findings indicate that proportion of protein-like compounds is highly variable across sites, ranging from less than five percent to more than 25%. Protein-like fluorescence has been linked to high rates of DOM biodegradation [Balcarczyk et al., 2009; Fellman et al., 2008]. In the Porcupine River, CDOM is enriched in protein-like compounds during winter relative to summer. The absence of a protein-like peak in summer-autumn flow (Figure 3a) may indicate that protein-like compounds are rapidly mineralized in stream during summer-autumn flow, when temperatures are warmer and microbial activity is greater than during under-ice conditions. This hypothesis is supported by the negative correlation observed between the percent of protein-like compounds and DOC:DON ratio (Figure 5), which can be a strong predictor of DOC biodegradability [Hunt et al., 2000]. Alternately, this pattern may represent a seasonal shift in the source-water contribution to stream flow, from a protein-rich to a protein-depleted source. This alternative hypothesis is supported by recent work byKraemer and Brabets [2012], who document the contribution of subpermafrost groundwater to the Porcupine River during winter, when the suprapermafrost aquifer is frozen. As permafrost thaw progresses and subpermafrost groundwater discharge increases, we might expect the proportion of protein-like compounds to increase during summer-autumn flow, likely resulting in a larger fraction of labile DOM. The response of riverine DOM to permafrost thaw and increased groundwater discharge will likely vary over space and time as a function of permafrost vulnerability and watershed hydrology.

4.2. DOM Composition as a Reflection of Suprapermafrost Groundwater Discharge

[34] Spatial variability in DOC concentrations and DOM composition of winter flow can be used to highlight differences in watershed hydrology as influenced by permafrost configuration. In a small subset of watersheds, DOM composition of winter flow closely resembles the DOM composition observed during summer-autumn flow. For example, at Hess Creek, the summer-autumn and winter fluorescence EEMs are nearly identical, indicating a consistent source water and organic matter input across seasons (Figures 3e and 3d). Winter flow at Hess Creek is also characterized as having high DOC concentrations, DOC:DON ratio, and SUVA254 values (Table 1 and Figure 5a), which closely approximate values observed during summer-autumn flow. Similar patterns in DOM are also observed at small streams in the Forty-Mile and Birch Creek watersheds, and in several blackwater streams of the Tanana River basin (e.g., Tolovana River). Based on these observations and other field data, we hypothesize that Hess Creek is primarily receiving groundwater inputs from a suprapermafrost aquifer and minimal inputs from a subpermafrost aquifer. Soils of the Hess Creek watershed are underlain by thick deposits of ice-rich loess silt (i.e., yedoma) [O'Donnell et al., 2011b; Kanevskiy et al., 2012], frozen material which is characterized as having low hydraulic conductivity [Burt and Williams, 1976]. This feature of yedoma likely restricts subpermafrost groundwater flow to stream flow. While the spatial extent of yedoma in Alaska is highly uncertain, Kanevskiy et al. [2011]have published a preliminary map showing the possible range and distribution of yedoma in Alaska. A recent synthesis of permafrost temperatures revealed that ice-rich permafrost, such as the Hess Creek yedoma deposits, has been slow to warm and thaw in recent decades due to the high latent heat content [Romanovsky et al., 2010]. Recent wildfires in the Hess Creek watershed have resulted in only modest increases in soil temperature and active layer thickness [O'Donnell et al., 2011a, 2011b], highlighting the resilience of yedoma to warming and disturbance.

[35] The relative stability of ice-rich permafrost in the Hess Creek watershed indicates that changes in groundwater discharge and DOM composition at this and similar sites will likely occur through the formation of a thickening suprapermafrost aquifer associated with increasing thaw depths. Groundwater inputs from the active layer typically occur from May through October [Ge et al., 2011]. During summer, the active layer is unfrozen and the potential for lateral flow through the suprapermafrost aquifer is optimal. During autumn, a freezing front descends from the ground surface through the active layer, lowering the hydraulic conductivity of soils in the suprapermafrost aquifer and reducing lateral hydrologic inputs to stream flow. Upon the complete refreezing of the active layer, hydrologic inputs from the suprapermafrost aquifer effectively cease. In this instance, under-ice sampling of stream water will likely reflect autumnal inputs from active layer soils. Present-day DOM characterization at Hess Creek captures this hydrologic condition, with high DOC concentrations, SUVA254, and HPOA values. Under projected warming, thaw depth may increase, resulting in the formation of a closed talik above the permafrost table, and a growing suprapermafrost aquifer [Bense et al., 2009; Ge et al., 2011]. In this instance, the generation of winter flow would likely originate from unfrozen zones beneath seasonally frozen surface soils [Shur et al., 2005]. DOC leaching from deep, newly unfrozen zones would likely result in a shift in radiocarbon age (toward an older signal) and SUVA254 of DOM, which are well correlated in some arctic rivers [Neff et al., 2006]. However, some field evidence indicates that even unfrozen silt may limit soil water infiltration, due to its low hydraulic conductivity, and constrain the establishment of deep suprapermafrost flow paths (J. Koch, personal communication, 2012). Together, radiocarbon and analytical characterization of DOM in winter flow will help to identify these post-thaw shifts in watershed hydrology and C mobilization.

4.3. DOM Composition as a Reflection of Subpermafrost Groundwater Discharge

[36] Subpermafrost groundwater discharge is likely reflected by the DOM composition of some rivers in the Tanana River basin, such as the groundwater-fed Clearwater Creek and the glacially fed Tanana River at Nenana and Delta River. DOC concentrations in these basins are generally low (<2 mgC L−1), and DOM composition is characterized by low aromaticity and a high proportion of HPI; virtually the opposite of DOM composition in Hess Creek. In the Tanana River, winter flow also has a distinct composition with respect to CDOM fluorescence. However, more analytical work is necessary to identify the compounds associated with EEMs peaks (Figure 3d), both in terms of chemical composition and reactivity. CDOM in the Tanana River basin appears to be primarily of terrestrial origin (FI values range from 1.45 to 1.51). Based on these observations, it is likely that DOM composition of winter flow reflects organic matter originating from older, subpermafrost groundwater. The Tanana River originates in the Alaska Range southeast of Fairbanks, where summer-autumn flow is generated primarily by glacial meltwater and clearwater mountain streams [Striegl et al., 2007; Kraemer and Brabets, 2012]. Sediments of the glacial outwash plains north of the Alaska Range are composed of coarse sand and gravel [Hopkins et al., 1955], which, together with steep hydraulic gradients, promote high groundwater discharge [Williams and van Everdingen, 1973; Walvoord and Striegl, 2007]. The importance of groundwater in the Tanana basin has also been documented on the Tanana Flats, a large low-lying area situated on an alluvial fan where numerous fens have formed in response to groundwater discharge [Racine and Walters, 1994; Jorgenson et al., 2001]. Interactions between permafrost and groundwater in the Tanana basin are complex: while the presence of thick ice-rich permafrost restricts groundwater discharge in some areas, the movement of groundwater in other areas has contributed to widespread permafrost thaw [Jorgenson et al., 2001]. The absence of permafrost on some south-facing hillslopes [seeMacLean et al., 1999] may also influence DOM composition, allowing for deeper infiltration of soil water and deeper groundwater flow paths. Our mixing model analysis indicates that continued permafrost thaw and increased groundwater discharge in the Tanana River basin will result in substantial increases in the HPI fraction of DOC, with only modest changes in DOC concentration and DOM aromaticity (Figure 7).

4.4. DOC Concentration in Relation to CDOM Properties

[37] In the Yukon River basin, about 12% of the annual DOC flux presently occurs during the winter season [Striegl et al., 2007]. Obtaining accurate DOC flux estimates for winter flow is challenging, given the high degree of uncertainty in measuring flow under ice relative to open water [Moore et al., 2002] and also due to the paucity of winter chemistry samples. Recent advances in in situ monitoring using optical sensors will likely improve flux estimates of riverine DOC [Saraceno et al., 2009; Pellerin et al., 2012]. Optical properties have proven to provide useful proxies, not only for DOC concentration, but also for lignin phenol concentration, and other chemical fractions of DOC [Hernes et al., 2009; Spencer et al., 2008, 2009]. In the present study, we observed strong correlations between CDOM absorbance and DOC concentration at three different wavelengths across all winter flow study sites. Similar results have been observed across the hydrograph for fourteen sampling locations in the YRB across the hydrograph [Spencer et al., 2008]. These observations suggest that in situ monitoring of winter-flow optical properties holds considerable promise for improving DOC flux estimates from subcatchments of the YRB.Spencer et al. [2012]also observed strong correlations between CDOM and DOC and HPOA, and their findings argue for watershed-specific ratings for CDOM-DOC relationships. We contend that CDOM-DOC observations across basins are critical for improving DOC flux estimates, particularly given the lack of winter flow samples from remote arctic and subarctic rivers.

4.5. Projected Changes in Permafrost Thaw, Groundwater Discharge, and DOM

[38] Changes in riverine DOM following permafrost thaw will likely influence the amount of DOM transported from rivers to the Arctic Ocean and also the amount of C released from surface waters to the atmosphere [Cole et al., 2007; Aufdenkampe et al., 2011]. Walvoord and Striegl [2007]estimated a 9–11% reduction in DOC export in the YRB between 1960 and 2050 based on past and projected changes in groundwater discharge to the Yukon River. Findings from our simple mixing model are supportive of this analysis, demonstrating that as the percentage of groundwater discharge increases during summer-autumn flow, DOC concentration in rivers of the YRB will likely decrease. Moreover, our findings illustrate important shifts, previously undocumented, in the composition of riverine DOM with increasing groundwater discharge (Figure 7). Most notably, we might expect decreases in DOM aromaticity and increases in the HPI of DOM as groundwater discharge increases. Prior studies have identified hydrophilic compounds as a labile fraction driving high rates of DOM mineralization in rivers [Qualls and Haines, 1992; Michaelson et al., 1998]. Findings from this study also indicate a positive relationship between DOC:DON and SUVA254 in winter flow (Figure 5a), which suggests that as DOM aromaticity decreases, biodegradability may increase.

[39] Our mixing analysis represents a simplified approach for understanding system behavior and DOC responses to shifting source-water contributions. Though these simplifications are commensurate with the available data and current understanding of system change, the analysis does not account for all complexities inherent in hydrologic, chemical, and biologic processes that govern DOC dynamics in watersheds underlain by permafrost. Thus, uncertainty exists with respect to these projected changes in DOC concentration and DOM composition. While field- and laboratory-based studies have documented some key factors controlling the release of DOC from high-latitude soils [e.g.,Neff and Hooper, 2002; Dutta et al., 2006], other factors remain poorly constrained. For instance, regional warming at high latitudes may alter controls on DOC export, through changes in DOC production pathways, post-thaw hydrology, and disturbance regimes. Moreover, climate-driven changes in these factors will likely drive both positive and negative feedbacks to DOC export. Results from a process model analysis of the Arctic basin [McGuire et al., 2010] indicate that DOC export to the Arctic Ocean has increased by 0.003 TgC yr−1 over the 20th century, primarily in response to warming effects on organic matter decomposition and DOC production. Increasing active layer thickness will transfer organic matter from frozen to unfrozen soil layers [O'Donnell et al., 2011a], potentially increasing the production and leaching potential of DOC. However, increasing active layer thickness may also increase sorption of DOM onto thawed mineral soils [Kawahigashi et al., 2006], reducing the lateral transfer of DOC from terrestrial to riverine ecosystems. Increasing active layer thickness may also reduce the transit time of soil water through an aquifer [Lyon et al., 2010], allowing more opportunity for microbially driven mineralization and transformation of the DOC pool. Riverine export of DOC from Arctic peatlands, which account for ∼30% of the land area in the YRB, can comprise up to 15% of peatland net primary production [Waddington and Roulet, 1997]. However, recent modeling analyses indicate that permafrost thaw may reduce the areal extent of peatlands [Avis et al., 2011], which may reduce the flux of riverine DOC. Consumption of soil organic matter by wildfire in arctic and subarctic regions will reduce soil organic carbon stocks [Turetsky et al., 2011; Mack et al., 2011], and consequently, may reduce DOC production and leaching [Kicklighter et al., 2012]. In recent years, considerable advances have been made in modeling groundwater [McKenzie et al., 2007] and DOC dynamics [Fan et al., 2010] in watersheds affected by frozen ground. A critical next step will be to embed some of these local complexities into regional- and global-scale process models.

5. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[40] The Yukon River Basin is unique among major arctic rivers, in that it flows from east to west through the discontinuous permafrost zone (∼68% of watershed area [Holmes et al., 2012]), whereas other major arctic rivers flow from permafrost-free headwaters north through the continuous permafrost zone, where thaw is minimal and localized [Romanovsky et al., 2010]. Permafrost temperatures in the discontinuous zone are near 0°C, and thus, permafrost in the region is more vulnerable to warming and thaw than the comparatively cooler permafrost in the continuous zone [Jorgenson et al., 2010]. In this sense, the YRB is ideal for studies of permafrost hydrology and biogeochemistry, and may serve as an analog for future conditions in other arctic watersheds underlain by continuous permafrost, where more extensive thaw is projected over the next century [Lawrence et al., 2008]. Our findings highlight the need for improved DOM characterization in arctic rivers for detection of changes in terrestrial ecosystems and permafrost-groundwater interactions. In the YRB, DOM variability in catchments appears to be driven in part by spatial variation in suprapermafrost and subpermafrost groundwater contributions to stream flow. Increases in groundwater discharge expected to occur with permafrost thaw will likely result in reduced DOC concentration, increased DOM aromaticity, and a larger fraction of hydrophilic compounds in summer-autumn flow. These responses of riverine DOM will vary across space and time, depending upon rates of permafrost thaw and hydraulic properties of local aquifers. Considerable uncertainty remains regarding the processes underlying these post-thaw shifts in riverine DOC, given the complexities and spatial heterogeneity of subsurface hydrology in permafrost settings. Future experimental work should aim to quantify various DOC loss pathways in thawed soils as determined by changing flow paths and groundwater transit times.


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[41] The authors would like to thank scientists affiliated with the U.S. Geological Survey Yukon River project, who collected and processed samples between 2001 and 2011. In particular, we thank Rob Striegl, Kimberly Wickland, Paul Schuster, Doug Halm, and Mark Dornblaser. We also thank Richard Smith, Deb Repert, and Ben Kamark for analytical support in the lab, and Cory McDonald for assisting with Figure 1. Richard Smith, Rob Spencer, and two anonymous reviewers provided valuable comments on an earlier version of this manuscript. Funding for Jon O'Donnell was provided by the U.S. Geological Survey Climate Effects Network. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information
gbc1942-sup-0001-t01.txtplain text document4KTab-delimited Table 1.
gbc1942-sup-0002-t02.txtplain text document2KTab-delimited Table 2.
gbc1942-sup-0003-t03.txtplain text document3KTab-delimited Table 3.
gbc1942-sup-0004-t04.txtplain text document2KTab-delimited Table 4.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.