Introduction
The export of total dissolved N (TDN) (including dissolved organic and inorganic N (DON, DIN)) from Arctic watersheds depends on the interplay between watershed morphology, hydrology, and N cycling (Bowden et al. 2008; McNamara et al. 2008). In continuous permafrost catchments, these factors are tightly linked: permafrost confines runoff and biological activity to the active layer and hydrologic processes exert a strong control on nutrient export dynamics (MacLean et al. 1999; Stieglitz et al. 2003; Townsend-Small et al. 2011; Yano et al. 2010; Harms and Jones 2012; Koch et al. 2013; Harms and Ludwig 2016). The timing of TDN flux depends largely on runoff volume and the response to hydrological inputs in a catchment. Runoff, which is needed to induce N transport (Perakis 2002), is influenced by catchment morphology, as the latter constitutes an underlying control on snow accumulation, hydrological flow paths and runoff response (Beven et al. 1988; Ferguson 1999; Woo 2012). The composition of TDN exported also varies according to catchment morphology, as this regulates hydrological flow paths, soil moisture regimes, the reduction/oxidation state of soils, and hence N cycling processes (Pinay et al. 2002; Petrone et al. 2007; McNamara et al. 2008; Yano et al. 2010; Harms and Jones 2012; Abbott et al. 2015; Harms and Ludwig 2016).
The theoretical understanding of watershed N cycling and retention suggests that the export of DIN in surface runoff from N-deficient ecosystems should only result from spatial or temporal separation of sources (e.g., atmospheric deposition and mineralization in soils) and sinks (e.g., microbial assimilation and uptake by vegetation) of N (Vitousek and Reiner 1978; Shaver et al. 1992; Perakis 2002; Lafrenière and Lamoureux 2008; Yano et al. 2010). This strong coupling between hydrology and N cycling processes means that fluvial N export in Arctic watersheds is highly sensitive to permafrost disturbance (Bowden et al. 2008; Kokelj et al. 2009; Harms and Jones 2012; Harms et al. 2013; Abbott et al. 2015). Disturbances, including thermokarst features such as retrogressive thaw slumps, thermokarst gullies, and active layer detachment slides (ALDs), can release pools of N that are usually isolated from biogeochemical cycling and hydrological export (Shaver et al. 1992; Jones et al. 2005; McClelland et al. 2007; Harms and Jones 2012; Keuper et al. 2012; Harms et al. 2013). Disturbances may result in new surface flow paths (e.g., through easily erodible mineral soils and between compression folds or factures created by ALDs) or activate subsurface pathways that are otherwise dormant at certain times of the thaw season (Petrone et al. 2006; Koch et al. 2013; Lafrenière and Lamoureux 2013). The disturbance impacts on vegetation cover, soil temperature, moisture, oxygen content, and nutrient availability may also stimulate inorganic N production but dampen N consumption relative to undisturbed areas (Paulter et al. 2010; Woods et al. 2010; Harms et al. 2013). Alaskan studies documented greatly increased DIN concentration immediately downstream of thermokarst gullies, thaw slumps, and slides (Bowden et al. 2008; Harms et al. 2013; Abbott et al. 2015). Although it is generally expected that permafrost disturbance will affect N export from Arctic rivers (Frey et al. 2007; Frey and McClelland 2009; Harms et al. 2013; Abbott et al. 2015), the seasonal characteristics and duration of impact of ALDs on fluvial N dynamics are poorly constrained High Arctic ecosystems (Lewis et al. 2012; Louiseize et al. 2014).
ALDs are widespread in some Arctic and High Arctic regions. These mass movement events and the resulting geomorphic features are expected to increase in frequency and areal extent with a warming climate (Lewkowicz 1990; Gooseff et al. 2009; Rowland et al. 2010; Vincent et al. 2013). ALDs result from the downslope movement of the vegetation cover and soils over the ice-rich permafrost table and cause exposure of the underlying mineral soils in the scar zone as well as folding of soils at the base of slopes (Lewkowicz and Harris 2005). The impact of ALDs on N export is of concern because Arctic rivers provide an important source of bioavailable N to downstream aquatic ecosystems that can be N limited (Levine and Whalen 2001; Symons et al. 2012; Hogan et al. 2014).
This research evaluates the impact of ALDs on the composition, magnitude, and timing of streamwater TDN export in continuous permafrost watersheds in the Canadian High Arctic. It is hypothesized that through the removal of the organic soil horizons (both a source of DON and a sink for DIN) and exposure of subsurface mineral soils (a potential source of DIN), ALDs will reduce the export of DON and enhance the export of DIN from Arctic permafrost catchments. To test these hypotheses, we compared pre-disturbance (2007) and post-disturbance (2012) concentrations and fluxes of dissolved N components in surface runoff from two headwater catchments at the Cape Bounty Arctic Watershed Observatory (CBAWO) on Melville Island, Nunavut, Canada. One of the catchments remains undisturbed and thus serves as a reference site for the impacted catchment, which was disturbed by several relatively large ALDs that formed at the end of the 2007 season. To our knowledge, this is the first study to examine seasonal and long-term impacts of ALDs on fluvial N in a High Arctic setting.
Methods
Site description
This research was conducted at the CBAWO (74°54ʹN, 109°35ʹW) on Melville Island, Nunavut (Fig. 1). The study area is underlain by continuous permafrost, with an active layer depth typically varying between 0.75 and 1 m. Soil moisture conditions vary from dry to saturated, which gives rise to various vegetation communities (Atkinson and Treitz 2012). Low-moisture sites primarily consist of shattered rock, fine-grained marine sediment, or till but accommodate polar desert vegetation community types that have poor vascular plant cover but a notable cover of lichens, such as Thamnolia subliformis spp. (worm lichen) and Cetraria nivalis spp. (snow lichen). Moderately wet areas host mesic-type vegetation, including the common deciduous shrub Salix arctica and Nostoc commune, a N-fixing cyanobacteria, while wet areas are typically completely vegetated by a wet sedge community type that includes species such as Eriophorium spp. and Sphagnum spp. (Atkinson and Treitz 2012).
[Image omitted: See PDF]
The disturbed catchment, Ptarmigan (PT), hosts three ALDs that formed in late July 2007 after a large rainfall event and an extended period of warmer temperatures (Lamoureux and Lafrenière 2009). These disturbances include two elongate ALDs (19 587 and 676 m2) that are physically coupled to the stream channel and an actively retrogressing isolated, compact ALD (3774 m2) (Rudy et al. 2013), which collectively define approximately 12% of the total catchment area (213 000 m2) (Lafrenière and Lamoureux 2013). The scar zones of each ALD remain free of vegetation and consist of clay-rich mineral soils (Figs. 2a and 2b). In undisturbed zones of the catchment, vegetation largely consists of polar semidesert and mesic heath communities. The runoff in the catchment is predominantly focused through a well-cut channel that carves into the mineral soils of the two elongate disturbances (Figs. 2a and 2b). The reference catchment, Goose (GS), spans 179 000 m2 and was not impacted by physical disturbance. Drainage in GS occurs via a poorly defined, vegetated channel or water tracks (Figs. 2c and 2d). The presence of a large hill on the northern edge of the GS catchment enables greater snow accumulation and therefore a prolonged runoff period compared to PT. Snow surveys in the two catchments from 2007, 2008, and 2009 illustrate that the mean end season snow water equivalence for the GS catchment was 25%–60% higher than in PT (Lafrenière unpublished data). The majority of the GS catchment retains sufficient moisture so that the dominant vegetation community in this catchment is mesic heath (Atkinson and Treitz 2012).
[Image omitted: See PDF]
The regional climate, as obtained from the closest long-term (1971–2000) meteorological record at Mould Bay, Northwest Territories (~300 km west of CBAWO), is characterized by mean winter (December–February) and summer (June–August) temperatures of −32.6 °C and 1.7 °C, respectively (Environment Canada 2013). The mean summer air temperature over the past 10 years (2003–2012) was 4.1 °C, or 2.4 °C above the climate normal at Mould Bay for 1971–2000. Mean annual precipitation at Mould Bay from 1971 to 2000 was 111 mm and mainly occurred as snowfall. Rainfall has generally been infrequent and of low magnitude according to the Mould Bay record, but July totals at CBAWO have reached 66.8 mm in recent years (Lewis et al. 2012; Favaro and Lamoureux 2014). The melt season runoff in the subcatchments at CBAWO generally lasts from early to mid-June until mid-July (Lafrenière and Lamoureux 2013).
Hydrological and environmental measurements
Discharge was measured at the outlet of each catchment in 8 in. cutthroat flumes by recording stage with Onset U20 level loggers (±2 mm) at 10 min intervals. Water level was obtained by correcting the stage measurement for barometric pressure variations, as recorded by an Onset U30 logger (±0.15 kPa). Level was converted to average hourly discharge (Q) (m3 s−1) by using the standard equations for the flumes and by taking the average discharge measured over an hour period (American Society of Civil Engineers 1974). Shielded air temperature and precipitation were measured at the WestMet weather station (109°37ʹ58.3ʺW, 74°56ʹ3.1ʺN), approximately 2.5 km northwest of MainMet (Fig. 1). Air temperature was recorded using an Onset HOBO H08 temperature sensor (±0.72 °C) in 2007 and an Onset HOBO UA-003-64 logger (±0.53 °C), each placed 1.5 m aboveground. Precipitation was measured with a Davis Industrial tipping bucket (0.2 mm tip–1) equipped with a HOBO H07 event logger in 2007 and an Onset UA-003 logger in 2012.
The seasonal discharge hydrographs were partitioned into nival melt (nival), baseflow, and stormflow periods based on large changes in seasonal discharge and the occurrence of rainfall. In this context, the terms nival, baseflow, and stormflow do not reflect a quantitative separation of source waters contributing to runoff but rather are used qualitatively to describe the general changes in runoff regime during the season. Note also that the term “seasonal” refers to the entire period of record, as our period of observation does not always encompass the full “annual” period of flow.
Hydrochemistry
Streamwater sampling in 2012 occurred from the first day of discharge (June 5) to the end of streamflow (June 24 in PT and July 1 in GS) and during stormflow (July 9–10, July 18–25). Samples were retrieved at stream outlets twice daily during peak and low discharge (approximately 1000 and 1800) throughout the nival melt and following rainfall events (stormflow periods) and once daily at approximately 1800 during baseflow conditions. Samples were collected in 1 L Nalgene bottles specific to each stream, which were tripled-rinsed with distilled water before a sampling campaign and with streamwater immediately before collection. The bottles were filled to minimize headspace and stored cool and in the dark until filtration (typically within 30–120 min). Stream sampling in 2007 was carried out in the same manner, with samples collected from the start of flow on June 11 to July 1 and 3 in PT and GS, respectively.
In 2012, rainfall samples were collected with passive collectors that were made by securing 20 L polycarbonate funnels (24 cm diameter) within a PVC frame so as to elevate the funnel openings 50 cm aboveground. A funnel placed at the base of the collector drained the rainfall into a 45 mL glass amber EPA vial through a 10 cm long Tygon R-3603 tube that penetrated the teflon septum of the vial cap. The EPA vial was suspended within a removable 1 L Nalgene bottle that was wrapped with black-out material and fastened to the neck of the collector. The collectors were opened manually during precipitation events and covered by plastic lids at all other times. Upon collection, the EPA vials were removed, closed with new caps, and replaced with a new clean vial. All vials and caps were triple-rinsed with distilled water before use.
Stream samples were filtered for dissolved inorganic N (NO2−, NO3−, and NH4+) and TDN, while precipitation samples were prepared only for DIN analyses, as the use of plastics in collectors precluded the accurate quantification of DON. Aliquots of TDN were filtered through pre-combusted 0.7 mm (GF/F) glass fiber filters using a glass filtration apparatus. The filtration unit was soaked in 3% hydrochloric acid overnight every day and triple-rinsed with distilled water before use and between samples. TDN samples were acidified to pH 2 and stored in 45 mL amber EPA vials with Teflon-lined septa that were rinsed with filtered sample and completely filled to eliminate headspace. Samples of DIN were filtered through 0.22 μm PVDF membrane filters using disposable syringes. The syringes were triple-rinsed with distilled water and sample water before and after use, while filters were rinsed with 10 mL of sample before filling the vials and discarded after each use. Aliquots of DIN were stored in plastic scintillation vials, which were pre-cleaned (triple soaked and rinsed with Milli-Q water at Queen’s University), rinsed with filtered sample, and filled with no headspace. All samples were refrigerated until analysis at Queen’s University.
Analytical methods
TDN was analyzed by NDIR and chemiluminescent detection using a Shimadzu TOC-VPCH/TNM system equipped with a high-sensitivity catalyst. Analytical error was 2.3% based on replicate analyses of standards, and the detection limit (defined as three times the standard deviation of repeats of the lowest level standard) was 0.015 ppm. Concentrations of NO2− and NO3− were measured by liquid ion chromatography using a Dionex ICS-3000. The system employed a gradient elution of 11–40 mM KOH (using an EG II KOH) flowing at 1.0 mL min–1 and was equipped with an ASRS 300 suppressor as well as an AS18 analytical column. The analytical error for NO3− was 2% based on replicate analyses of standards, and the detection limit was 0.003 ppm (0.0007 ppm N). The analytical error for NO2− was 2.5% and the detection limit was 0.0001 ppm.
NH4+ concentration for the 2012 samples was measured via colorimetry using an Astoria Pacific FASPac II flow analyzer. The detection limit was 0.01 ppm N and the analytical error was 0.9%. In 2007, NH4+ was determined by ion chromatography, and precision was 0.003 ppm N (or 5%) (Lewis et al. 2012). For NH4+ analyzed by ion chromatography, the detection limit was dependent on Na+ concentration, such that higher concentration of Na+ resulted in an increase in the detection limit of NH4+. Due to the elevated concentration of Na+ in PT, the NH4+ concentration in PT towards the end of the runoff period (5 days of record) in 2007 may have been underestimated. We conservatively estimate that the magnitude of this underestimate may be up to a factor of 2 or 3 for these samples. Note that although the concentrations of NH4+ may have been higher than reported in the last few days of flow in PT, even significantly higher concentrations of NH4+ (e.g., 2 or 3 times higher) at this time (when discharge was extremely low) would not affect the findings or interpretations presented here, neither for the differences in the two catchments prior to disturbance nor for the interannual differences in NH4+ concentration. DIN was calculated as the sum of the N from NH4+, NO3−, and NO2− (when detectable), while DON is the difference between TDN and DIN.
Mass flux calculations
Fluvial N mass export (g) for each hydrological period (nival melt, baseflow, stormflow, and seasonal) in 2012 was calculated as the sum of daily N mass flux (Md (g/ d1)). For days when two samples were collected, the daily mass flux was calculated as the product of the mean concentration (ppm N) and the total mean daily discharge (Qd (m3 s−1)). For days when only a single sample was collected, the daily mass flux was calculated as the concentration of the one sample multiplied by the total mean daily discharge. Specific fluxes of N (g N mm−1) were obtained by dividing N mass export (g N) by catchment discharge (mm). The total rainfall inputs of DIN (g N) for each subcatchment were estimated as the product of measured concentration of DIN (sum of N from NH4+ and NO3−) in rain samples and rainfall volume in the catchment (rainfall depth times catchment area) over the rainfall period. Rainfall N inputs (g N) were divided by the amount of rainfall (mm) in order to obtain the specific DIN input per unit of rain (g N mm−1).
The propagation of error for a species flux is determined using the following:
[Formula omitted: See PDF]
where pFLe is the propagated flux error for a particular hydrological period, such as nival melt or the entire hydrological year, µDFL is the mean daily flux, Ae is the analytical error, C is the mean daily concentration of a particular species, and Qe is the fractional discharge error. The daily compounded error — sum of the daily fractional error in concentration (Ae/C) and discharge error (Qe) — is multiplied by the mean daily flux (µDFL) to determine the mean daily flux error. The flux error for a given hydrological period is the sum of the daily flux error for the days within that hydrological period.
Results
Climate
Details of weather pre- and post-disturbance are presented in Favaro and Lamoureux (2014). The pre-disturbance year 2012 was the second warmest summer on record at CBAWO (after 2011), while the summer of 2007 was the third warmest. The mean daily air temperature (MDAT) in June and July 2007 were 0.23 and 8.6 °C, respectively, and maximum daily temperatures peaked at 16.8 °C on July 13 (Fig. 3d). In 2012, the MDATs for June and July were 4.1 and 8.4 °C, respectively, and the most pronounced increase in MDAT occurred between June 28 and 30, when MDAT reached a maximum of 16.8 °C (Fig. 3h). The cumulative thawing degree-days (the sum of the MDATs for days when the mean temperatures are above zero) was 388 °C in 2007 and 485 °C in 2012, hence approximately 20% lower than in 2012. Monthly rainfall totals for June and July 2007 were very similar (16.8 and 16.0 mm, respectively). The rainfall total for June 2012 (3.6 mm) was much lower than the first notable rainfall event in 2012, which registered 8.6 mm on July 9. In 2012, three smaller rainfall events occurred on July 18, 22, and 24, measuring 1.6, 0.8, and 5.8 mm, respectively (Fig. 3e). The cumulative rainfall between June 1 and July 26, 2012, was 19.2 mm.
[Image omitted: See PDF]
Discharge
Nival runoff accounts for the majority of the seasonal runoff from both of these headwater catchments in both years (Fig. 4). However, stormflow makes a significant proportion of seasonal runoff (Fig. 4).
[Image omitted: See PDF]
The 2007 hydrological record for PT and GS spanned June 11 to July 1 and June 11 to July 3, respectively (Fig. 3a). Melt runoff in 2007 peaked on June 13 in GS and June 16 in PT. In this year, however, there was no persistent discernable low-flow period (baseflow) following the nival melt because rainfall occurred consistently from June 18 to 30 (Fig. 3a). Hence, for the purposes of calculating fluxes, the 2007 hydrologic season for both catchments was partitioned into the nival and stormflow periods only (Fig. 3a). The nival period in each catchment was determined to begin on June 1 and end on June 27, while the stormflow period occurred from thereon until July 1 in PT and July 3 in GS. Note that although the runoff season for these catchments in 2007 was more than 3 weeks shorter compared to 2012, the number of days during which discharge occurred was similar between years for both catchments (Figs. 3a and 3e).
In 2012, the total seasonal discharge for both catchments was much higher than in 2007 (Fig. 4; Table 1). The 2012 seasonal discharge was approximately 5 times higher in GS and 1.5 times higher in PT, relative to 2007 (Fig. 4; Table 1). These interannual differences are largely a function of the much higher nival runoff captured in 2012, which was approximately 6 times higher in GS and 1.7 times higher in PT, relative to 2007 (Table 1).
[Image omitted: See PDF]
In 2012, the nival melt period began on June 5 and runoff peaked on June 6 and June 8 in GS and PT, respectively (Fig. 3e). June 15 was determined to be the end of the nival runoff period, as runoff decreased notably in both catchments on this date. Hence, June 16 was determined to mark the beginning of the baseflow period in both catchments. Baseflow ended on June 24 in PT, while baseflow runoff in GS continued until July 1, 2012. Total discharges during the nival (June 5–15) and baseflow (June 16 to the end of streamflow) periods were higher in GS (68 and 11 mm, respectively) than in PT (41 and 8 mm, respectively) (Fig. 4). The stormflow period in 2012 (July 9–25) included two distinct rainfall periods that reinitiated discharge. Stormflow runoff first occurred from July 9 to July 10 due to a 9.4 mm rainfall event, but hydrograph responses were subdued by dry antecedent moisture conditions (Fig. 3e). The runoff response ratio of this first rainfall event was only 0.19 in both PT and GS. Discharge was again reinitiated in both catchments from July 18 to 25 following a subsequent series of smaller magnitude rainfall events (1.6 mm on July 18, 0.8 mm on July 22). The wetter antecedent moisture conditions prior to this second series of rainfall events lead to much higher runoff response relative to the first (and larger) rainfall event (Fig. 3b). The runoff response ratios during this second rainfall event (18–25 July) were 1.62 and 0.97 for PT and GS, respectively. Cumulative discharge during stormflow was higher in PT (14 mm) than in GS (10 mm) (Fig. 4).
Stream N concentrations
NH4+ concentration
The mean seasonal NH4+ concentration was similar in the two catchments in both 2007 and 2012. In 2007, NH4+ concentration was persistently low in both catchments, with no discernable patterns over the sampling period (Fig. 3b). The only notable differences in NH4+ concentration between the two catchments in 2007 were that the mean nival period NH4+ concentration in GS was approximately 3 times higher than that for PT and also that the mean stormflow NH4+ concentration was 9 times higher in PT than for GS. Although these differences may seem substantial, it is important to note that the concentration of NH4+ was near the detection limit for most of the season in both catchments (Fig. 3b; Table 1).
In 2012, the NH4+ concentrations in both catchments were substantially higher compared to 2007 (Table 1; Figs. 3b and 3f). The 2012 nival and seasonal NH4+ concentrations for the undisturbed GS catchment were 5.8 and 5.6 times higher than in 2007, respectively (Table 1). The increases in NH4+ concentrations in GS between years (5.8 and 5.6) were very similar to the relative increase in the discharge (factor of 5.6) in this catchment between the two years (Table 1). In PT however, the 2012 nival and seasonal NH4+ concentrations were 14 and 10 times higher than in 2007, respectively, which were many times greater than the increases in nival and seasonal discharge in PT between years (1.7 and 1.5 times, respectively) (Table 1). These increases in concentration across the two catchments are much greater than could be explained by the change in the methodology used for determining NH4+ in 2012 (see the “Analytical methods” section).
Despite these large changes in concentration in both catchments from 2007 to 2012, there were no substantial differences in NH4+ concentration between the two catchments in 2012 (Table 1). However, similar to 2007, the mean nival NH4+ concentration in 2012 was slightly higher in GS than in PT, while the stormflow NH4+ concentration was slightly higher on average in PT than in GS.
NO3− concentration
In 2007, NO3− concentration in both catchments was consistently low (0.001 ppm N in GS and 0.002 ppm N in PT); nonetheless, the seasonal mean NO3– concentration in PT was twice that of GS (Fig. 3c; Table 1). There were no discernable seasonal patterns to the NO3− concentration in either stream in 2007 (Fig. 3c).
In 2012, the mean NO3− concentration in the undisturbed catchment (GS) was similar to the 2007 concentration (Table 1). However, the 2012 NO3− concentration in PT was substantially higher relative both to GS and to pre-disturbance conditions for every hydrological period (Table 1; Fig. 3g). During the 2012 nival and stormflow periods, mean NO3− concentrations in PT were between 22 and 75 times higher compared to the same periods in 2007. Stormflow during the summer of 2012 resulted in the highest measured NO3− concentration in PT. A maximum NO3− concentration of 0.348 ppm N was recorded in PT on July 18, 2012, and concentration in excess of 0.128 ppm N persisted until July 21 (Fig. 3g). Stormflow concentration of NO3− in PT almost doubled the concentration of NO3− in rainfall, whereas NO3− concentration in GS was much lower than that for rainfall, remaining near or below the detection limit throughout stormflow (Fig. 3g; Table 1).
DON concentration
Both the concentration and the seasonal variations in DON in the two catchments were nearly indistinguishable in 2007 (Fig. 3d; Table 1). In the undisturbed site (GS), the mean seasonal DON concentration was approximately 30% higher in 2012 relative to 2007, which was largely a result of the higher concentration during stormflow (Table 1). In the disturbed PT catchment, the 2012 mean seasonal DON concentration was only slightly higher than in 2007. This was also a function of the much higher stormflow DON concentration in PT in 2012 relative to the previous year. There was a slight decrease in the nival DON concentration in the PT site in 2012 (Table 1).
The 2012 DON concentration trended similarly between catchments until the end of baseflow in PT (June 24). In GS, DON concentration increased by 220% from June 28 to July 1 in conjunction with a 10 °C increase in MDAT (Fig. 3h). We found a statistically significant correlation (n = 16, p < 0.01 for all cases) between MDATs and the average daily DON (r = 0.66, p = 0.006) and NH4+ concentrations (r = 0.69, p = 0.003). The highest DON concentration in both catchments occurred during stormflow in 2012. In GS, DON peaked at 0.789 ppm N on July 10, whereas in PT, DON increased more than twofold to 0.445 ppm in response to the July 9 rainfall event. The mean stormflow DON concentrations were similar between catchments (0.441 ppm N in GS and 0.377 ppm N in PT) but were more than 2 times higher than for the other hydrological periods.
Note that despite substantial differences in the various components of N between the two catchments in both years, the mean TDN concentration was similar between catchments pre- and post-disturbance for all hydrological periods, with the exception of baseflow in 2012, when mean TDN was much higher in GS (0.217 ppm N) than in PT (0.135 ppm N) (Table 1). The late-season rainfall runoff in 2012 also had significant impacts on TDN. TDN concentration nearly doubled in GS and more than quadrupled in PT relative to baseflow levels in response to the July 9 rainfall event, mainly as a result of high DON. The seasonal mean TDN concentration in 2012 was 1.7 times higher relative to 2007 in both catchments (Table 1).
Stream N fluxes
NH4+ fluxes
NH4+ accounted for a relatively small proportion (<1% up to 26%) of the specific TDN flux (g N mm–1) at any time (Fig. 5; Table 2). Note that we report specific or discharge-normalized fluxes (g N mm–1) to facilitate comparisons between catchments and years.
[Image omitted: See PDF][Image omitted: See PDF]
The 2007 results show that seasonal NH4+ flux was nearly twice as high in GS relative to PT prior to disturbance; however, the overall NH4+ flux accounted for only approximately 3%–6% of the total seasonal N flux. The vast majority of the NH4+ flux occurred during the nival melt period in both catchments (Fig. 5; Table 2).
In 2012, the seasonal NH4+ fluxes and proportions were significantly higher relative to 2007 as a result of the increases in NH4+ concentration and discharge in 2012 (Fig. 5; Table 2). The total seasonal NH4+ fluxes were similar between catchments in 2012, yet the distribution of the NH4+ flux over the season differed between the disturbed PT and undisturbed GS sites. First, the baseflow period mass flux of NH4+ (g mm–1) in GS exceeded that of PT by more than a factor of 2 (Fig. 5; Table 2). This appears to be due to the increase in NH4+ concentration in GS during the prolonged runoff period in this catchment between June 25 and July 1 (Fig. 3f; Table 2). Second, the stormflow NH4+ flux for the disturbed PT catchment was more than 1.5 times higher than for the undisturbed GS (Table 2).
NO3− fluxes
Nitrogen from NO3− accounted for less than 1% to the total N flux in both catchments in 2007, and approximately 70%–80% was exported during the nival period (Fig. 5; Table 2). Although the 2007 NO3− fluxes were small, it is notable that the flux in PT (9.63 ± 0.69 g N mm–1) was more than 3 times that of GS (2.43 ± 0.50 g N mm–1) prior to disturbance (Table 2).
In 2012, the NO3− fluxes in both catchments increased relative to the 2007 season, with an approximately fourfold increase in total seasonal NO3− flux in GS and a 30-fold increase in the disturbed PT catchment (Fig. 5; Table 2). The changes in the contributions of NO3− to the TDN flux in the disturbed PT catchment were by far the most notable changes to the N budgets in the catchments post-disturbance. In 2012, the contribution of NO3− to the total seasonal N flux in PT was 20% (up from only 1% in 2007), while in GS, NO3− accounted for approximately 0.7% of the seasonal TDN flux (up from 0.3% in 2007) (Table 2). In GS, NO3− fluxes increased in the nival and baseflow periods, as NO3− was slightly more concentrated relative to 2007, but the 2012 stormflow NO3− flux was similar to that in 2007. In contrast, the NO3− flux in the disturbed PT catchment during the nival period was 10 times higher than 2007, and the stormflow NO3− flux was 100 times higher than in 2007 (Fig. 5; Table 2).
DON fluxes
DON accounted for the vast majority of TDN flux in both catchments for each hydrological period pre- and post-disturbance (Fig. 5; Table 2). The seasonal distribution and total fluxes of DON were very similar in PT and GS prior to disturbance in 2007, with 70%–80% of the DON being exported during the nival period (Fig. 5; Table 2). DON accounted for 94% and 96% of the seasonal TDN flux in GS and PT, respectively, in 2007.
In 2012, the total DON flux in GS increased by about 50%, but the seasonal DON flux in PT post-disturbance was the same as for 2007 (Table 2). There were substantial changes in the seasonal patterns of the DON flux in 2012 relative to 2007. In 2012, the early-season DON flux (i.e., the sum of nival and baseflow periods) in the disturbed PT catchment was approximately half of the flux observed in this catchment prior to disturbance, while the early-season flux in GS was similar to the 2007 season (Table 2). Additionally, in contrast with 2007 when most of the DON was exported during the nival period, the stormflow period in 2012 contributed significantly to the seasonal DON flux, with 50% and 63% of the seasonal DON being exported from GS and PT during stormflow, respectively (Table 2).
Retention/export of atmospheric N
Rain samples were not collected for chemistry in 2007, but in 2012, rainfall was sampled to better understand the significance of rainfall as a source of N in these watersheds. In 2012, rainfall contributed approximately 1132 ± 126 g N as NH4+ (NH4+-N) to PT and 951 ± 107 g NH4+-N to GS (Table 3). The mass of N exported from GS and PT as NH4+ during stormflow was approximately 9% and 17%, respectively, of the rainfall input. Although this might be expected given that PT had higher stormflow discharge than GS, even on a per-unit discharge basis, PT exported a much greater proportion of N as NH4+ from rainfall (61% or 130 ± 17.3 g N mm−1) than was exported in GS (47% or 84.1 ± 16.2 g N mm−1) (Table 3).
[Image omitted: See PDF]
With respect to NO3−, rainfall inputs of N to the PT watershed were 706 ± 48.4 g NO3−-N (Table 3). Approximately 45% of this N input (315 ± 9.19 g NO3−-N) was exported during stormflow; however, the specific stormflow N flux as NO3− in PT (237 ± 5.87 g N mm−1) surpassed rainfall inputs to the catchment (128 ± 10.7 g NO3−-N mm-1) (Table 3). In GS, rainfall added approximately 594 ± 40.7 g NO3−-N to the catchment, but stormflow NO3− export in GS measured only 1.46 ± 5.26 g NO3−-N, or approximately 0.25% of the rainfall NO3− mass input.
Therefore, for both NH4+ and NO3−, the GS catchment retained a larger proportion of the atmospheric N inputs relative to PT, where the mass of NO3− exported via runoff was approximately half the mass of wet deposition of NO3− from rainfall.
Discussion
Predisturbance conditions
The bulk of hydrological N fluxes in High Arctic watersheds usually occurs during the nival melt period (Lafrenière and Lamoureux 2008; McNamara et al. 2008; Lewis et al. 2012), as this has historically been the time of greatest runoff and hydrological connectivity across a larger surface area of the watershed (Hinzman et al. 2003; Woo 2012). This was the case for the catchments at our study sites in 2007, when stormflow and total discharge were low.
In 2007, prior to disturbance, the two catchments exhibited very similar concentrations and seasonal patterns in the different N components (NH4+, NO3−, DON), with DON accounting for more than 90% of the TDN flux, which is similar to that reported for the larger rivers at Cape Bounty (Lafrenière and Lamoureux 2008; Lewis et al. 2012) as well as for other Arctic rivers (McNamara et al. 2008; Townsend-Small et al. 2011). The only notable differences between the two catchments prior to disturbance were the higher mean nival period NH4+ concentration (~3 times higher) and lower seasonal mean NO3− in GS. The fact that the GS catchment typically has deeper snow cover, is on average wetter, and is more densely vegetated than the PT catchment likely explains the higher NH4+ concentration during the nival period and slightly lower stormflow NH4+ concentration in GS relative to PT. Overwinter mineralization rates in Arctic tundra soils have been shown to increase with greater snow depth (Schimel et al. 2004), which could yield higher rates of NH4+ flushing in GS relative to PT during snowmelt when immobilization or uptake by plants is limited. Furthermore, as NH4+ is strongly retained by the plants and microbes in the rooting zone during thaw and the growing season in Arctic soils (Schimel et al. 2004; Petrone et al. 2006; Yano et al. 2010), the greater extent and abundance of vegetation cover in GS relative to PT could explain the higher retention of NH4+ in this catchment during the later part of the runoff season in 2007.
The more elevated mean seasonal NO3− concentration in PT relative to GS prior to disturbance likely reflects the greater degree of channelization and thus runoff contact with mineral horizons in this catchment. Previous work from CBAWO demonstrates that prior to disturbance, total dissolved solute concentrations (Lafrenière and Lamoureux 2013) and suspended sediment loads (Lamoureux et al. 2014) in PT were higher than in GS, which both support that runoff PT has greater contact with mineral soil layers compared to GS. Several recent studies in other permafrost watersheds highlight that flow through deeper soils increases the NO3− concentrations in runoff, which likely results from the reduced retention or removal of NO3− with depth (Harms and Ludwig 2016), the higher bulk density and ion-exchange capacity of mineral soils, and (or) from evapo-concentration of solutes near the base of the freezing front from the previous year (Koch et al. 2013).
Interannual changes unrelated to disturbance
The overall hydrology and N dynamics were significantly different in 2012 relative to 2007. The notable differences in the concentrations and relative abundances of the different N species between years, even for the undisturbed catchment, indicate that interannual differences in hydrology played a role in the changes in N composition in runoff between these years.
Increases in NH4+
Both catchments exhibited significant increases in the nival and mean seasonal concentrations of NH4+ in 2012 relative to 2007, and the NH4+ concentrations were still similar in the two catchments. The increases in NH4+ concentrations in 2012 may be linked to increases in discharge volume. Studies by Yano et al. 2010 indicated that NH4+ retention decreases with increases in water flow volume, and Harms and Ludwig (2016) also showed that net NH4+ production increases along flow paths where residences times are shorter, due to a reduction in the potential assimilation or nitrification of NH4+ with higher flow rates and water volumes. The increase in the seasonal NH4+ concentrations in the undisturbed GS catchment between years was on the same order as the increases in discharge; thus, increases in discharge could explain the increase in the NH4+ concentrations in the undisturbed watershed between years. On the other hand, the increases in NH4+ concentration in the PT catchment between years were many times greater than the increases in discharge, which indicates that discharge changes alone cannot explain the higher NH4+ in the disturbed catchment. The additional increase in the NH4+ concentration in the PT runoff post-disturbance is likely explained by the removal of vegetation and organic-rich surface soils across 12% of the watershed, where the vast majority of NH4+ would likely have been retained (Yano et al. 2010). The higher proportion of atmospheric inputs of NH4+ exported in PT relative to GS in 2012 supports a reduced NH4+ sink in PT relative to GS.
Increases in late-season DON
In continuous permafrost watersheds, DON usually dominates TDN because much of the seasonal runoff interacts with organic-rich horizons, where DON is high due to the decomposition and subsequent leaching of N-bearing compounds from soil organic matter (SOM) (Petrone et al. 2006; Lafrenière and Lamoureux 2008; Yano et al. 2010; Lewis et al. 2012; Woo 2012). The only other key change in N composition in the undisturbed catchment between years was an increase in the mean seasonal DON concentration, which appears to be due to elevated DON concentration in late season rainfall runoff events in 2012 that were not observed as part of the study season in 2007.
Late-season runoff can make substantial contributions to seasonal dissolved loads including dissolved organic matter (DOM) and DON in permafrost watersheds (Petrone et al. 2006; McNamara et al. 2008; Lewis et al. 2012). Studies have shown that DON concentrations typically increase with higher flow volumes (Townsend-Small et al. 2010) and in the late-season especially following rainfall events (Townsend-Small et al. 2010; Lewis et al. 2012). The increases in the late-season concentration of DON and other solutes associated with summer rainfall have been attributed to the increases in rates of decomposition and weathering with increasing temperatures, and the resulting accumulation of solutes in the active layer followed by their subsequent export, as rainfall runoff reestablishes hydrological connectivity in the catchment (Weintraub and Schimel 2003; Koch et al. 2007; Melle et al. 2015). Therefore, the increase in DON concentration and fluxes in GS and in PT in 2012 are likely explained by the higher stormflow discharge volumes and rainfall runoff occurring later in the season in 2012 relative to 2007. Large stormflow-related TDN fluxes have been observed within larger watersheds at the CBAWO and elsewhere in the Arctic (MacIntyre et al. 2006; Petrone et al. 2006; McNamara et al. 2008; Townsend-Small et al. 2010; Lewis et al. 2012), suggesting that stormflow may become increasingly important for N delivery if projected increases in temperature and summer rainfall hold true (Kirtman et al. 2013).
Disturbance impacts on N export
Decrease in DON concentration
It is commonly thought that where organic layers are relatively thin, permafrost disturbance will result in lower concentrations and export of DON (and therefore lower DON/TDN ratios) because of diminished runoff contact with organic soil horizons (MacLean et al. 1999; Walvoord and Striegl 2007). The similarity in DON concentration between the two catchments prior to disturbance but significantly lower nival and low baseflow DON concentration in PT post-disturbance indicates that, as hypothesized, the removal of a significant portion of the source area of DON (e.g., organic soils) by the ALD resulted in a decrease in the DON export. The nival period DON concentration in PT was 19% lower in 2012 relative to 2007, which was similar to the area affected by the removal or organic soils from ALDs (12%). The large areas of exposed mineral soils within ALD scar zones may have promoted additional DON removal by adsorption of DOM onto mineral soils (Kawahagashi et al. 2006, Lavoie et al. 2011) or by mineralization (e.g., ammonification of organic N including DON to NH4+). Mineralization in the unvegetated ALD scar zones could be enhanced by C) limitation or the fulfillment of microbial N demand, which may result from lack of a labile C-rich substrate (e.g., fresh litter inputs) and lower competition/demand for N (Weintraub and Schimel 2003; Schimel et al. 2004; Wild et al. 2013).
On the other hand, the removal of the organic soils did not appear to have an effect on the DON concentration during the late-season stormflow period in 2012. Similar to GS, the PT catchment had a relatively large increase in the DON concentration during the 2012 stormflow, which appears to have driven an increase (~12%) in the seasonal DON concentration (Table 1). Although we know that late-season runoff has the potential to increase DON concentration and flux, as outlined in the case of GS in the “Increases in late-season DON” section, we would not expect this impact to be discernable given the removal of soil in the ALD. The result indicates that rainfall runoff has greater connectivity with the undisturbed areas of the watershed relative to runoff during the nival period. Although we do not have detailed mapping of the snow accumulation across the PT catchment, we know that snow is redistributed and preferentially accumulates in depressions on the landscape. The ALD in PT is a large depression in the watershed, and the snow accumulation is typically concentrated within the ALD. The concentration of nival flow through the disturbed area could explain the reduction in DON at this time of year, relative to late season rainfall, which appears to access a more extensive area rich in DOM.
Increases in NO3−
The similar mean NO3− concentration between catchments prior to disturbance (2007) but substantially higher NO3− concentrations in PT relative to GS in 2012 indicates that ALDs led to higher NO3− in PT and this played a large role in the difference in the composition of dissolved N between the catchments in 2012. Nutrient retention theories and studies of watershed N fluxes suggest that hydrologic NO3− loss from N-deficient ecosystems, such as those in High Arctic regions, should only result from spatial or temporal separation of sources and sinks of NO3− (Shaver et al. 1992; Perakis 2002; Lafrenière and Lamoureux 2008; Yano et al. 2010). Nitrification, the oxidation of NH4+ to NO3−, is a dominant source of NO3− in pristine watersheds. Nitrification in Arctic environments is usually low during the growing season because NH4+ availability, which is a primary control of nitrification, is strongly limited by the high demand and preferred uptake of NH4+ by plants and heterotrophic microrganisms (Giblin et al. 1991; Nordin et al. 2004; Booth et al. 2005; Weintraub and Schimel 2005; Yano et al. 2010). In ALD scar zones, however, NH4+ availability is likely higher than in undisturbed soils due to lower plant/microbial demand in the absence of organic soil horizon and vegetation (Keuper et al. 2012; Harms et al. 2013; Wild et al. 2013).
In the disturbed PT catchment, the elevated 2012 NO3− concentrations throughout the season and the elevated NO3− export ratios during stormflow runoff indicate that there was likely enhanced production and (or) an additional source of NO3− in the catchment that was spatially disconnected from sink areas as a result of the ALD. Stormflow data highlight that NO3− sinks were efficient in GS but limited in PT. The specific flux of NO3− during stormflow in GS was approximately 130 times lower than in PT and concentrations in runoff remained low despite substantial inputs from rainfall. Hence, NO3− was mostly retained (hydrologically or biologically) within the GS catchment or lost via denitrification, which is consistent with increased N demand during summer months (Lipson and Monson 1998).
In addition to providing hydrologic pathways through areas with potentially less N demand, disturbance features also likely have larger inorganic N pools at the surface due to the exposure of inorganic N-rich mineral soils (Keuper et al. 2012) and facilitate enhanced mineralization and degradation of SOM by providing favorable moisture, oxygen, and nutrient conditions (Paulter et al. 2010; Keuper et al. 2012). Paulter et al. (2010) found evidence that additional nutrients introduced by ALDs stimulated microbial activity and led to increased degradation of previously frozen SOM, which is consistent with the release of older particulate organic carbon from PT (Lamoureux and Lafrenière 2014). Keuper et al. (2012) demonstrated that the available N and the rates of mineralization of N in the near-surface permafrost soils can be up to seven 7 and 3 orders of magnitude higher, respectively, than for soils in the rooting zone in subarctic peatland sites. Harms et al. (2013) found approximately 3 times higher rates of nitrification but lower denitrification in thermokarst gully soils compared to adjacent undisturbed soils. Isotopic evidence presented in Louiseize et al. (2014) also supports that the ALDs in the PT catchment promoted enhanced export of nitrified NO3− that likely originated from mineral-rich scar zones. These findings support that the 1–2 order of magnitude increase in NO3− concentrations observed in PT following disturbance could be explained by the release of new inorganic N pools previously inaccessible to bigeochemical cycling and export, thus resulting in relatively higher nitrification and NO3− leaching loss (Jones et al. 2005)
Hence, seasonal NO3− export form the disturbed PT catchment is likely enhanced because (1) the removal of the organic soil horizons can result in reduced NO3− retention, (2) by exposing N-rich permafrost soils to the surface, the ALD-affected area is subject to enhanced inorganic N pool and likely supports higher net mineralization and higher nitrification rates in surface soils, and (3) the direct coupling between ALDs and the stream channel facilitates hydrological export of NO3− from the disturbed areas.
Impact on seasonal patterns of dissolved N
The results from this study support the findings of other studies that demonstrate that physical permafrost disturbance will likely lead to increased DIN export from Arctic catchments (Petrone et al. 2006; Walvoord and Striegl 2007; Bowden et al. 2008; Harms and Jones 2012; Harms et al. 2013; Abbott et al. 2015). Our observations provide new insight into the potential impact this increase in DIN might have on downstream ecosystems by illustrating the seasonal pattern of the changes in N export. Additionally, our results indicate that the impact of the ALD is still evident even 5 years after the disturbance event.
In this study, the significantly higher NO3− concentrations in PT in 2012 bolstered TDN flux during the stormflow period and resulted in a seasonal DIN flux that was 95% greater in PT than in GS (Table 3). The higher DIN (hence lower DON) contribution to TDN in PT post-disturbance is consistent with other Arctic studies that report diminished concentrations of organic constituents (Kokelj et al. 2002) and elevated concentrations of DIN due to disturbance (Bowden et al. 2008; Lantz et al. 2009).
The seasonal TDN export from GS in 2012 was higher than from PT because of the extended baseflow period in GS (June 16 to July 1), which was more important than the nival period in delivering TDN. Low runoff volumes during the PT baseflow period resulted in limited N export; however, in GS, a large snowbank maintained runoff while organic soils likely helped retain moisture as the active layer continued to develop (Woo 2012). Rising temperatures during baseflow likely stimulated increases in decomposition and mineralization (Weintraub and Schimel 2003), thus enabling DOC, DON, and NH4+ to accumulate in the active layer (Nadelhoffer et al. 1991; Schmidt et al. 1999; McNamara et al. 2008; Schaeffer et al. 2013). The statistically significant correlations between mean daily air temperatures and the mean DON and NH4+ concentrations found in this study also support that these species accumulate in soils with warming temperatures. Hence, the mobilization of high concentrations of DON and NH4+ led to a relatively large TDN flux during baseflow in GS despite minimal runoff at this time.
Late-season rainfall was an important source of inorganic N inputs to both watersheds, and rainfall runoff was a key mechanism for mobilizing the N accumulating in soils in the late season. In both catchments, the dominance of stormflow in generating TDN flux in 2012 highlights that hydrological reconnectivity (due to precipitation) after periods of warming can mobilize large quantities of N from undisturbed High Arctic watersheds, and even more so from disturbed watersheds. These results show that hydrologic connectivity following the nival melt, whether maintained by melt runoff or slow release from storage during the baseflow period or rainfall later in the season, is an important factor in controlling nutrient export.
Conclusions
Permafrost disturbances have the potential to impact seasonal N transport by perturbing linkages between morphology, hydrology, and N cycling processes. As hypothesized, our results show that NO3− and DON are especially sensitive to the disturbance in this permafrost setting. Increases in DIN fluxes from soils and watersheds subject to permafrost disturbances have been documented in other studies (Bowden et al. 2008; Harms et al. 2013; Abbott et al. 2015). However, this study represents the first investigation of both the seasonal and longer term (5 years post-disturbance) impacts of ALDs on TDN export in High Arctic watersheds. As such, the results presented here significantly advance our understanding of the potential duration and seasonal response of N to disturbance and provide a basis for modeling N response to permafrost and climate change in the region.
This study demonstrates that seasonal patterns in concentration and mass flux of dissolved N from High Arctic watersheds are influenced by both hydrological changes, such as total runoff volumes and late-season rainfall, and physical disturbance. Late-season rainfall runoff increased the concentration and flux of all dissolved N, and higher total runoff resulted in increases in dissolved NH4+ export and, in the absence of disturbance, increases in DON export. Although the total dissolved seasonal N flux is largely a function of the duration of the runoff season and sustained runoff during late summer in particular, the disturbance of catchment surficial soils by ALDs significantly increased the total DIN flux and altered the relative proportions of the N species in runoff in the disturbed catchment. Furthermore, this study demonstrates that although disturbances do not appear to greatly affect the total fluxes of N from the landscape, ALDs can have a persistent (~5 years) impact on the magnitude of DIN export from small High Arctic watersheds.
The enhanced delivery of DIN observed in our study has implications for the C cycle in aquatic systems in permafrost settings. Several studies suggest that the lability (or biodegradability) of DOM is related to addition of inorganic N, which stimulates the mineralization of DOM (Holmes et al. 2008; Mann et al. 2012). Hence, the increases in DIN may lead to an increase of the lability of DOM and enhance atmospheric CO2 emission from High Arctic catchments affected by permafrost degradation, thereby promoting a positive feedback on the C cycle. Furthermore, given that disturbances often occur adjacent to streams and lakes, as well as coastal areas that are often N-limited systems (Levine and Whalen 2001; Vancoppenolle et al. 2013), projected changes in rainfall and disturbance (Lewkowicz and Harris 2005; Lewis and Lamoureux 2010; Rowland et al. 2010; Kirtman et al. 2013) stand to significantly enhance the delivery of nutrients to downstream aquatic ecosystems, and likely later in the season when aquatic productivity and nutrient requirements are high.
Acknowledgements
This research was funded by the ArcticNet NCE and the Government of Canada International Polar Year program. We are also grateful to the Polar Continental Shelf Program (PCSP), Natural Resources Canada, who provided excellent logistical support. Valuable field, laboratory, and GIS assistance was provided by K. Rutherford, G. Montross, S. Montross, D. Lamhonwah, and A. Rudy. We also acknowledge the two anonymous reviewers; their input significantly improved the quality of the manuscript.
Abbott B.W., Jones J.B., Godsey S.E., Larouche J.R., and Bowden W.B. 2015 Patterns and persistence of hydrologic carbon and nitrogen export from collapsing upland permafrost Biogeosci. Discuss 12 2063 -2100
American Society of Civil Engineers 1974 Generalized discharge relations for cutthroat Flumes J. Irrigat. Drain. Div 98 569 -583
Atkinson D.M. and Treitz P.M. 2012 Arctic ecological classifications derived from vegetation community and satellite spectral data Rem. Sens 4 3948 -3971
Beven K.J., Wood E.F., and Sivapalan M. 1988 On hydrological heterogeneity – catchment morphology and catchment response J. Hydrol 100 353 -375
Booth M.S., Stark J.M., and Rastetter E. 2005 Controls on nitrogen cycling in terrestrial ecosystems: a synthetic analysis of literature data Ecol. Monogr 75 139 -157
Bowden W.B., Gooseff M.N., Balser A., Green A., Peterson B.J., and Bradford J. 2008 Sediment and nutrient delivery from thermokarst features in the foothills of the North Slope, Alaska: potential impacts on headwater stream ecosystems J. Geophys. Res 113 G02026
Environment Canada. 2013. http://climate.weather.gc.ca/climate_normals/results_e.html [accessed 25 May 2013].
Favaro E.A. and Lamoureux S.F. 2014 Antecedent controls on rainfall runoff response and sediment transport in a High Arctic catchment Geogr. Ann. A Phys. Geogr 96 433 -446
Ferguson R.I. 1999 Snowmelt runoff models Prog. Phys. Geogr 23 2 205 -227
Frey K.E. and McClelland J.W. 2009 Impacts of permafrost degradation on arctic river biogeochemistry Hydrol. Process 23 169 -182
Frey K.E., McClelland J.W., Holmes R.M., and Smith L.C. 2007 Impacts of climate warming and permafrost thaw on the riverine transport of nitrogen and phosphorus to the Kara Sea J. Geophys. Res 112 G04S58
Giblin A.E., Nadelhoffer K.J., Shaver G.R., Laundre J.A., and McKerrow A.J. 1991 Biogeochemical diversity along a riverside toposequence Arct. Alaska Ecol. Monogr 61 415 -435
Gooseff M.N., Balser A., Bowden W.B., and Jones J.B. 2009 Effects of hillslope thermokarst in northern Alaska Eos, Trans. Am. Geophys. Union 90 4 29 -30
Harms T.K. and Jones J.B. 2012 Thaw depth determine reaction and transport of inorganic nitrogen in valley bottom permafrost soils Global Change Biol 18 2958 -2968
Harms T.K. and Ludwig S.M. 2016 Retention and removal of nitrogen and phosphorus in saturated soils of arctic hillslopes Biogeochemistry 127 291 -304
Harms T.K., Abbott B.W., and Jones J.B. 2013 Thermo-erosion gullies increase nitrogen available for hydrologic export Biogeochemistry 117 299 -311
Hinzman, L.D., Kane, D.L., Yoshikawa, K., Carr, A., Bolton, W.R., and Fraver, M. 2003. Hydrological variations among watersheds with varying degrees of permafrost. Proceedings of the Eighth International Conference on Permafrost, Zurich, Switzerland, pp. 407–411.
Hogan E.J., McGowan S., and Anderson N.J. 2014 Nutrient limitation of periphyton growth in arctic lakes in south-west Greenland Polar Biol 37 1331 -1342
Holmes R.M., McClelland J.W., Raymond P.A., Peterson B.J., and Stieglitz M. 2008 Lability of DOC transported by Alaskan rivers to the Arctic Ocean Geophys. Res. Lett 35 L03402
Jones J.B., Petrone K.C., Finlay J.C., Hinzman L.D., and Bolton R.W. 2005 Nitrogen loss from watersheds of interior Alaska underlain with discontinuous permafrost Geophys. Res. Lett 32 L02401
Kawahagashi M., Kaiser K., Rodionov K., and Guggenberger G. 2006 Sorption of dissolved organic matter by mineral soils of the Siberian forest tundra Global Change Biol 12 1868 -1877
Keuper F., van Bodegom P.M., Dorrepaal E., Weedon J.T., van Hal J.R., van Logtestijn R.S.P., and Aerts R. 2012 A frozen feast: thawing permafrost increases plant-available nitrogen in subarctic peatlands Global Change Biol 18 1998 -2007
Kirtman, B., Power, S.B., Adedoyin, J.A., Boer, G.J., Bojariu, R., Camilloni, I., Doblas-Reyes, F.J., Fiore, A.M., Kimoto, M., Meehl, G.A. et al. 2013. Near-term climate change: projections and rredictability. In Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge.
Koch O., Tscherko D., and Kandeler E. 2007 Temperature sensitivity of microbial respiration, nitrogen mineralization, and potential soil enzyme activities in organic alpine soils Global Biogeochem. Cycles 21 GB4017
Koch J.C., Runkel R.L., Striegl R.G., and McKnight D.M. 2013 Hydrologic controls on the transport and cycling of carbon and nitrogen in a boreal catchment underlain by continuous permafrost J. Geophys. Res. Biogeosci 118 698 -712
Kokelj S.V., Smith C.A.S., and Burn C.R. 2002 Physical and chemical characteristics of the active layer and permafrost, Herschel Island, Western Arctic Coast, Canada Permafrost Periglac. Process 13 171 -185
Kokelj S.V., Zajdlik B., and Thompson M.S. 2009 The impacts of thawing permafrost on the chemistry of lakes across the Subarctic Boreal–Tundra transition, Mackenzie Delta Region, Canada Permafrost Periglac. Process 20 185 -199
Lafrenière, M.J. Unpublished data. 2005–2009 snow water equivalence data, Cape Bounty. http://www.polardata.ca [accessed 25 May 2016].
Lafrenière M.J. and Lamoureux S.F. 2008 Seasonal dynamics of dissolved nitrogen exports from two High Arctic watersheds, Melville Island, Canada Hydrol. Res 39 4 323 -335
Lafrenière M.J. and Lamoureux S.F. 2013 Thermal perturbation and rainfall runoff have greater impact on seasonal solute loads than physical disturbance of the active layer Permafrost Periglac. Process 24 3 241 -251
Lamoureux S.F. and Lafrenière M.J. 2009 Fluvial impact of extensive active layer detachments, Cape Bounty, Melville Island, Canada Arct. Antarct. Alp. Res 41 1 59 -68
Lamoureux S.F. and Lafrenière M.J. 2014 Seasonal fluxes and age of particulate organic carbon exported from Arctic catchments impacted by localized permafrost slope disturbances Environ. Res. Lett 9 4 045002
Lamoureux S.F., Lafrenière M.J., and Favaro E. 2014 Erosion dynamics following localized permafrost slope disturbances Geophys. Res. Lett 41
Lantz T.C., Kokelj S.V., Gergel S.E., and Henry G.H.R. 2009 Relative impacts of disturbance and temperature: persistent changes in microenvironment and vegetation in retrogressive thaw slumps Global Change Biol 15 1664 -1675
Lavoie M., Mack M.C., and Schuur E.A.G. 2011 Effects of elevated nitrogen and temperature on carbon and nitrogen dynamics in Alaskan arctic and boreal soils J. Geophys. Res. Biogeosci 116 G03013
Levine M.A. and Whalen S.C. 2001 Nutrient limitation of phytoplankton production in Alaskan foothill lakes Hydrobiologia 455 189 -201
Lewis T. and Lamoureux S.F. 2010 Twenty-first century discharge and sediment yield predictions in a small high Arctic watershed Global Planetary Change 71 27 -41
Lewis T., Lafrenière M.J., and Lamoureux S.F. 2012 Hydrochemical and sedimentary responses of paired High Arctic watersheds to unusual climate and permafrost disturbance, Cape Bounty, Melville Island, Canada Hydrol. Process 26 13 2003 -2018
Lewkowicz, A.G. 1990. Morphology, frequency and magnitude of active-layer detachment slides, Fosheim Peninsula, Ellesmere Island, NWT. Proceedings of the 5th Canadian Permafrost Conference. Vol. 54.
Lewkowicz A.G. and Harris C. 2005 Frequency and magnitude of active-layer detachment failures in discontinuous and continuous permafrost, northern Canada Permafrost Periglac. Process 16 115 -130
Lipson D.A. and Monson R.K. 1998 Plant–microbe competition for soil amino acids in the alpine tundra: effects of freeze–thaw and dry–rewet events Oecologia 113 406 -414
Louiseize N.L., Lafrenière M.J., and Hastings M.G. 2014 Stable isotopic evidence of enhanced export of microbially derived NO3− following active layer slope disturbance in the Canadian High Arctic Biogeochemistry 121 565 -580
MacIntyre S., Sickman J.O., Goldthwait S.A., and Kling G.W. 2006 Physical pathways of nutrient supply in a small, oligotrophic Arctic lake during summer stratification Limnol. Oceanogr 51 1107 -1124
MacLean R., Oswood M.W., Irons J.G.I., and McDowell W.H. 1999 The effect of permafrost on stream biogeochemistry: a case study of two streams in the Alaskan (U.S.A.) taiga Biogeochemistry 47 239 -267
Mann P.J., Davydova A., Zimov N., Spencer R.G.M., Davydov S., Bulygina E., Zimov S., and Holmes R.M. 2012 Controls on the composition and lability of dissolved organic matter in Siberia’s Kolyma River basin J. Geophys. Res. Biogeosci 117 G01028
McClelland J.W., Stieglitz M., Pan F., Holmes R.M., and Peterson B.J. 2007 Recent changes in nitrate and dissolved organic carbon export from the upper Kuparuk River, North Slope, Alaska J. Geophys. Res 112 G04S60
McNamara J.P., Kane D.L., Hobbie J.E., and Kling G.W. 2008 Hydrologic and biogeochemical controls on the spatial and temporal patterns of nitrogen and phosphorus in the Kuparuk River, arctic Alaska Hydrol. Process 22 3294 -3309
Melle C., Wallenstein M., Darrouzet-Nardi A., and Weintraub M.N. 2015 Microbial activity is not always limited by nitrogen in Arctic tundra soils Soil Biol. Biogeochem 90 52 -61
Nadelhoffer K.J., Giblin A.E., Shaver G.R., and Laundre J.A. 1991 Effects of temperature and substrate quality on element mineralization in six Arctic soils Ecology 72 1 242 -253
Nordin A., Schmidt K.S., and Shaver G.R. 2004 Nitrogen uptake by soil microbes and plants in relation to soil nitrogen supply Ecology 85 4 955 -962
Paulter B.G., Simpson A.J., McNally D.J., Lamoureux S.F., and Simpson M.J. 2010 Arctic permafrost active layer detachments stimulate microbial activity and degradation of soil organic matter Environ. Sci. Technol 44 4076 -4082
Perakis S. 2002 Nutrient limitation, hydrology and watershed nitrogen loss Hydrol. Process 16 3507 -3511
Petrone K.C., Jones J.B., Hinzman L.D., and Boone R.D. 2006 Seasonal export of carbon, nitrogen, and major solutes from Alaskan catchments with discontinuous permafrost J. Geophys. Res 111 G02020
Petrone K., Fuffam I., and Laudon H. 2007 Hydrologic and biologic control of nitrogen export during snowmelt: a combined conservative and reactive tracer approach, Water Resour Res 43 W06420
Pinay G., Clement J.C., and Naiman R.J. 2002 Basic principles and ecological consequences of changing water regimes on nitrogen cycling in fluvial systems Environ. Manage 30 4 481 -491
Rowland J.C., Jones E.C., Altmann G., Bryan R., Crosby B.T., Geernaert G.L., Hinzman L.D., Kane D.L., Lawrence D.M., et al. 2010 Arctic landscapes in transition: response to thawing permafrost EOS Trans 91 26 229 -230
Rudy A.C.A., Lamoureux S.F., Treitz P., and Collingwood A. 2013 Identifying permafrost slope disturbance using multi-temporal optical satellite images and change detection techniques Cold Regions Sci. Technol 88 37 -49
Schaeffer S.M., Sharp E., Schimel J.P., and Welker J.M. 2013 Soil–plant N processes in a High Arctic ecosystem, NW Greenland are altered by long-term experimental warming and higher rainfall Global Change Biol 19 3529 -3539
Schimel J.P., Bilbrough C.B., and Welker J.M. 2004 Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic tundra communities Soil Biol. Biochem 36 2 217 -227
Schmidt I.K., Jonasson S., and Michelsen A. 1999 Mineralization and microbial immobilization of N and P in arctic soils in relation to season, temperature, and nutrient amendment Appl. Soil Ecol 11 147 -160
Shaver G.R., Billings W.D., Chapin F.S., Giblin A.E., Nadelhoffer K.J., Oechel W.C., and Rastetter E.B. 1992 Change and the carbon balance of arctic ecosystems Bioscience 42 433 -441
Stieglitz M., Shaman J., McNamara J., Engle V., Shanley J., and Kling G.J. 2003 An approach to understandig hydrologic connectivity on the hillslope and the implications for nutrient transport Biogeochem. Cycles 17 1105
Symons C.C., Arnott S.E., and Sweetman J.N. 2012 Nutrient limitation of phytoplankton communities in Subarctic lakes and ponds in Wapusk National Park, Canada Polar Biol 35 481 -489
Townsend-Small A., McClelland J.W., Holmes R.M., and Peterson B.J. 2011 Seasonal and hydrologic drivers of dissolved organic matter and nutrients in the upper Kuparuk River, Alaskan Arctic Biogeochemistry 103 109 -124
Vancoppenolle M., Bopp L., Madec G., Dunne J., Ilyina T., Halloran P.R., and Steiner N. 2013 Future Arctic Ocean primary productivity from CMIP5 simulations: uncertain outcome, but consistent mechanisms Glob. Biogeochem. Cycles 27 605 -619
Vincent W.F., Lemay M., Allard M., and Wolfe B. 2013 Adapting to permafrost change: a science framework EOS 94 42 373 -375
Vitousek P.M. and Reiners W.A. 1978 Ecosystem succession and nutrient retention: a hypothesis Bioscience 25 6 376 -381
Walvoord M.A. and Striegl R.G. 2007 Increased groundwater to stream discharge from permafrost thawing in the Yukon River basIn: potential impacts on lateral export of carbon and nitrogen Geophys. Res. Lett 34 L12402
Weintraub M.N. and Schimel J.P. 2003 Interactions between carbon and nitrogen mineralization and soil organic matter chemistry in Arctic tundra soils Ecosystems 6 129 -143
Weintraub M.N. and Schimel J.P. 2005 Nitrogen cycling and the spread of shrub control changes in the carbon balance of Arctic tundra ecosystems Bioscience 55 408 -415
Wild B., Schnecker J., Bárta J., Capek P., Guggenberger G., Hofhansl F., Kaiser C., Lashchinsky N., Mikutta R., et al. 2013 Nitrogen dynamics in Turbic Cryosols from Siberia and Greenland Soil Biol. Biochem 67 84 -93
Woo, M.K. 2012. Permafrost hydrology. Springer, New York.
Woods G.C., Simpson M.R., Paulter B.G., Lamoureux S.F., Lafrenière M.J., and Simpson A.J. 2010 Evidence for the enhanced liability of dissolved organic matter following permafrost slope disturbance in the Canadian High Arctic Geochem. Cosmochim. Acta 25 7226 -7241
Yano Y., Shaver G.R., Giblin A.E., Rastetter E.B., and Nadelhoffer K.N. 2010 Nitrogen dynamics in a small arctic watershed: retention and downhill movement of 15N Ecol. Monogr 80 2 331 -351
Melissa J. Lafrenière [email protected]
Department of Geography and Planning, Queen’s University, Kingston, ON K7L 3N6, Canada
Nicole L. Louiseize
Department of Geography and Planning, Queen’s University, Kingston, ON K7L 3N6, Canada
Scott F. Lamoureux
Department of Geography and Planning, Queen’s University, Kingston, ON K7L 3N6, Canada
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2017. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
This study investigates the impacts of active layer detachments (ALDs) on nitrogen in seasonal runoff from High Arctic hillslope catchments. We examined dissolved nitrogen in runoff from an undisturbed catchment (Goose (GS)) and one that was disturbed (Ptarmigan (PT)) by ALDs, prior to disturbance (2007) and 5 years after disturbance (2012). The seasonal dynamics of nitrogen species concentrations and fluxes were similar in both catchments in 2007, but the mean seasonal nitrate concentration and mass flux from the disturbed catchment were on the order of 30 times higher relative to the undisturbed catchment in 2012. Stormflow yielded 45% and 60% of the 2012 total dissolved nitrogen flux in GS and PT, respectively, although rainfall runoff provided less than 25% of seasonal discharge. Results support that through the combined effects of increased disturbance and rainfall, climate change stands to significantly enhance the export of nitrate from High Arctic watersheds. This study highlights that the increase in the delivery of nitrate from disturbance is especially pronounced late in the season when downstream productivity and the biological demand for this often limiting nutrient are high. Our results also demonstrate that the impact of ALDs on nitrate export can persist more than 5 years following disturbance.