| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
U.S. Geological Survey, 240 W. Prospect Rd., Ft. Collins, CO 80526
* Corresponding author (robert_stottlemyer{at}usgs.gov)
Received for publication September 18, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: DOC, dissolved organic carbon • DON, dissolved organic nitrogen • DIC, dissolved inorganic carbon • DIN, dissolved inorganic nitrogen
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the taiga–tundra ecotone at the northern extent of the boreal biome, much SOM is below the annual thaw depth and not readily available. Any factor that increases the depth to permafrost could change soil moisture and temperature, SOM decomposition, nutrient availability and cycling, and respiration rates (Chapin et al., 1995; Jonasson et al., 1999). Research in the southern boreal Wallace Lake watershed, Michigan, shows that slight gains in SOM decomposition increase available nutrients as inorganic N in amounts greater than the sum of other N sources (precipitation, fixation) (Stottlemyer and Toczydlowski, 1999). An increase in available N will alter above- or belowground C/N ratios, which, in turn, could accelerate aboveground production or belowground heterotrophic respiration (Shaver et al., 1992). Which process is greater in magnitude may determine whether the ecosystem becomes a C source or sink (Oechel et al., 1995; McKane et al., 1997).
Since 1950, temperatures along the southern slopes of the Brooks Range in Northern Alaska have increased especially in late winter (Herrmann et al., 2000). In 1990, we began study of nutrient cycles in the 800-ha Asik treeline watershed in the Noatak National Preserve, Alaska. The overall research objectives are to examine longer-term trends in C and nutrient budgets, possible responses to global change, and compare results with findings from a southern boreal ecotone watershed, Wallace Lake, Michigan (Herrmann et al., 2000). Wallace Lake receives 20 times the atmospheric inorganic N input as Asik, and retains >75% of N inputs (Stottlemyer et al., 1998).
In this paper, watershed-level data are presented, which examine relationships between topographic aspect, soil temperature and moisture, inorganic and organic N pools, C pools, soil CO2 efflux, and growing season net N mineralization rates. Seasonal changes in upstream and downstream water chemistry are related to soil conditions and possible changes in hydrologic flowpath. These early results served as background for ongoing research on the effect of soil temperature, moisture, and N availability on dissolved organic carbon (DOC) and dissolved organic nitrogen (DON) production and export to the aquatic ecosystem.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
The soil association is gravelly, hilly to steep Pergelic Cryaquepts–Pergelic Cryorthents (Rieger et al., 1979). The association consists of poorly drained to well-drained soils, most with permafrost. Soils are poorly drained along long gradual slopes, valley bottoms, and on steeper north aspects. Well-drained soils occur along ridges and on steeper south aspect slopes. Frost features are common. Flood plain soils are loamy along the Agashashok River. Along steeper slopes of the Asik watershed, soils are shallow and rocky with a clay loam texture. In the Asik watershed, the O2 bulk density averages 0.3 where the silt content is low to about 0.7 where the silt content is high. The O2 pH (CaCl2) varies from 4.6 in tussock (cotton grass, Eriophorum vaginatum L.) tundra to 6 beneath spruce, total C from 15% beneath alder to 40% in tussock tundra, and total N from 0.8% beneath alder to 1.7% below tussock tundra. Discontinuous permafrost exists in the watershed, especially where there is no forest. The soil surface (5 cm) is frozen from late September to May, except for upper reaches with southern aspects. The bedrock is sedimentary and metamorphic rock. About 5 to 7% of the watershed consists of talus slopes. The Noatak River drainage was not glaciated during the last ice age.
Upper elevation portions of the lower one-third and most of the middle half of the watershed are dominated by white spruce [Picea glauca (Moench) Voss]. Spruce basal area varies from 23 m2 ha-1 in bottom land to 4 m2 ha-1 on south aspects (Suarez et al., 1999). Forest understory consists primarily of mountain fern moss [Hylocomium splendens (Hedw.) B.S.G.], field horsetail (Equisetum arvense L.), and Richardson boykinia [Boykinia richardsonii (Hook.) Gray], with shrubs of willow (Salix spp.) and bog blueberry (Vaccinium uliginosum L.) The understory of the taiga–tundra transition zone and tundra is dominated by tussocks of cotton grass, bog blueberry, shrubby cinquefoil (Potentilla fruticosa L.), and birch (Betula nana L). The upper 20% of the watershed area is dominated by shrubs as birch and scattered alder [Alnus crispa (Ait.) Pursh] on more northern aspects, and mesic nontussock tundra. The stream flood zone is dominated by willow.
Meteorology–Hydrology
A 10-m meteorological tower with datalogger and solar panel was located in the bottom third of the watershed. Air and soil temperature (5- and 10-cm depth), relative humidity, radiation, wind speed, and direction were monitored year-round. In the alpine zone, a datalogger monitored air and soil temperature. Precipitation was sampled using bulk collectors. During winter, the sampler was a 20-cm diameter, 1.5-m long tube lined with custom-fit heavy plastic liners. During the growing season, a 10-cm diameter plastic tube with funnel was used with a prerinsed qualitative filter in place to minimize dust and particulates from entering the sample. Precipitation samples were collected weekly during the growing season and early fall.
Stream discharge was measured at a natural cross section, not a constructed weir. Stage height was monitored by standing stake and pressure transducer (Level Logger). Each year a discharge curve was developed by measuring cross sections and velocity at varying stage heights. Water temperature was monitored year-round by datalogger. Stream water was sampled weekly at the mouth and five upstream stations from late May to mid-September. Sampling was daily or more frequent at the mouth during periods of rapid hydrograph change. The stream was frozen with little or no flow from mid-October to late April or early May.
Soils and Forest Floor
Fifty 15 by 15 m plots were established in major vegetation types. Grids of 10 plots each were located on north and south aspects in spruce, in a scattered alder stand with north aspect, in birch with opposing south aspect, and in a spruce–tundra transition zone with south aspect. In each grid, plots were located 100 m apart along two elevation tiers and instrumented as follows. About one-third of the plots were permanently instrumented with air (1 m height) and forest floor (5 cm in O2) thermisters (Onset), litter traps, and three ceramic cup tension lysimeters at 30 cm depth (SoilMoisture Corp.) across the plot's lower boundary. An additional third of the plots were instrumented only with thermisters, which were moved to an additional subset of plots each month of the growing season for comparisons with the record from the permanent instrumented plots. The O2 samples for analysis and incubations were collected with plastic tubes 5 cm in diameter and 7 cm long. Moisture (gravimetric), inorganic N pools (2 M KCl), C/N ratios (Leco Model 1000 CHN Analyzer), net N mineralization rates, and O2/soil respiration (PP Systems, EGM-2) were monitored on all plots every 3 or 4 wk at three to five points along the plot boundaries tangential the contour.
The O2 net nitrate-nitrogen (NO-3–N), ammonium-nitrogen (NH+4–N), and total inorganic N (NO-3–N + NH+4–N) mineralization rates beneath each species type was estimated using the buried polyethylene bag technique (Eno, 1960). With this method, net mineralization is the sum of mineralized NH+4–N plus NO-3–N from organic N, minus immobilization of NH+4–N and NO-3–N. Net nitrification is the sum of NO-3–N from both organic N and NH+4–N minus immobilization of NO-3–N. After removing intact the surface organic layer, the upper 5 to 7 cm of O2 horizon was sampled with a 5-cm diameter soil corer. Paired cores were pulled from each sampling point. One core from each pair, representing a nonincubated sample, was placed in a Whirl Pac, brought to the field laboratory and extracted with 2 M KCl. The other core was placed in a 0.025-mm thick polyethylene bag and returned to the same hole for incubation. The surface organic matter was then replaced. After about 25+ d, the field incubated sample was removed and returned to the field laboratory for extraction.
Chemical Analyses
After collection, forest floor samples were processed at the field station. Each core was weighed. A large subsample was then sieved. A 10-g subsample was extracted with 50 mL of 2 M KCl, and NO-3–N (cadmium reduction) and NH+4–N (indophenol) determined on a Lachat autoanalyzer at the U.S. Forest Service Rocky Mountain Research Station, Ft. Collins, CO. Another subsample was frozen for C/N analyses at the Colorado laboratory. The O2 moisture was determined by oven drying (105°C for 24+ h) a subsample in Kotzebue. Bulk density was calculated from total oven dry weight/volume.
For precipitation and stream water samples, pH, specific conductance, and alkalinity (titration with 0.01 M H2SO4 to pH 4.5, stream water samples only) were determined at the field station. Separate filtered (prerinsed, 0.45 µm) subsamples were shipped in coolers to our Colorado laboratory for ion analyses (Waters Ion Chromatography). An additional filtered sample was sent to Michigan Technological University, Houghton, MI, for analyses of DOC (Sievers Model 600) and DON (Shimadzu/Antek 720C N analyzer).
Data Analysis
General meteorological data and stream water data from 1996 to 1999 were summarized and shown as monthly averages. Vegetation plot data were primarily from 1999. Plot results for CO2 efflux, O2 inorganic and organic N pools, N incubations, and soil moisture for all points within each plot were composited by date giving an n of 8 to 10 for each vegetation type and aspect. To compare O2 temperature with N incubations, the mean O2 temperature for the N incubation period was used for each plot. To correlate CO2 efflux with O2/soil temperature, instantaneous temperature was used. Variances among species types were equal (Bartletts test for homogeneity of variance). For determining differences among species types, season, and etc., Systat (Wilkinson et al., 1996) Multivariate General Linear Hypothesis (MGLH) ANOVA routines were used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
During the last 5 yr, stream flow has been largely confined to the period from late April to early November. Stream discharge peaked in late spring and late summer (Fig. 3) . Mean discharge during the flow period was 80 L s-1, and annual runoff was 13.1 cm yr-1 or 45% of precipitation.
|
Watershed O2 moisture was higher (p < 0.05) on south aspect plots. Soil moisture was greater (p < 0.01) beneath the transition zone vegetation than spruce, alder, or birch, and higher (p < 0.01) beneath spruce and tundra than alder or birch. Within spruce or between alder (north aspect) or birch (south aspect) there was no difference (p > 0.05) in soil moisture by aspect.
Forest floor O2 C/N ratios (p < 0.01) and C pools (p < 0.05) were larger on south aspects and reduced in stream bottom land. The O2 C/N ratios ranged from 18 beneath spruce to 24 in the spruce–tussock transition zone (Table 1). The C/N ratios were higher (p < 0.05) beneath the transition zone and birch than spruce. The C pools where greater (p < 0.05) beneath spruce and the transition zone than under alder and higher beneath the transition zone then birch. Within spruce, O2 C content did not differ by aspect. Forest floor O2 total N pools were higher (p < 0.05) beneath spruce and the transition vegetation than beneath alder. Beneath spruce no difference in O2 total N content was observed by aspect, but soil total N content was reduced (p < 0.05) in the stream bottom land. Soil inorganic N pools were smaller (p < 0.05) beneath alder than spruce, the transition zone, or tussock tundra.
|
|
Seasonal watershed NH+4–N pools in the O2 declined (p < 0.01) by early July relative to early June, but increased (p = 0.07) again by early August. Soil NO-3–N pools showed no change. Net NH+4–N mineralization rates were lower (p < 0.01) in July and early August than June. Moisture declined (p < 0.01) in July relative to June.
During the growing season, soil and forest floor CO2 efflux increased with temperature and moisture (Table 2). But there was no interaction between moisture and temperature on CO2 efflux. Soil and forest floor CO2 efflux was greater (p < 0.001) on south aspects. The CO2 efflux was higher (p < 0.01) beneath birch and the transition zone O2 than alder.
Beneath spruce, O2 respiration was greater (0.02 g C h-1, p < 0.05) on the south aspect. Soil respiration increased (p < 0.01, 0.02 g C h-1 °C-1) with soil temperature, and rates increased (p < 0.001) throughout the growing season into August when sampling stopped. On the north aspect beneath spruce, incorporating both soil moisture and temperature in the ANOVA model accounted for most variation in soil CO2 efflux (p < 0.01, R2 = 0.91). Beneath alder, soil respiration rates increased (p < 0.01) in late June and July relative to rates in early June, and respiration increased (p < 0.01, 0.03 g C h-1 °C-1) with soil temperature. But there was no interaction between soil moisture, temperature, and CO2 efflux. Beneath birch shrub stands, there was no seasonal change in soil CO2 efflux rate or interaction between soil moisture, temperature, and CO2 efflux.
Stream Chemistry
At the watershed mouth, stream water concentrations of Ca2+, Mg2+, NO-3–N, SO2-4, and bicarbonate declined with increased discharge (Fig. 3, Table 3). The decline in NO-3–N was slight. Conversely, stream water DOC concentration increased with stream discharge. Stream water DON concentrations were low, often at level-of-detection, and averaged 0.01 mg L-1 in both the alpine and subalpine. On average, stream water DON was 10% of DIN. Subalpine stream water Ca2+, Mg2+, SO2-4, and HCO-3 concentrations were correlated (p < 0.001), and Ca2+ concentrations were inverse with DOC (p < 0.001).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In Asik, heavy stream ice generally formed above the gauging station for a distance of 300 m or more to where spruce occupied stream bottom land. The presence of this ice made accurate stream hydrologic measurements difficult, if not impossible, through most of May. In 1999, the stream ice melt persisted up to late June. Watershed discharge responded rapidly to summer storms, a characteristic of catchments where discontinuous permafrost persists (Hilgert and Slaughter, 1983). Summer precipitation can be substantial. For example, in the last week of July 1999, >9 cm of rain fell. Unlike the annual pattern of discharge observed at Caribou–Poker Creeks or Rock Creek, where runoff peaks during May–June, the Asik watershed had two peaks that were often comparable: late spring and autumn. The autumn increase (Fig. 3) reflects the increase in seasonal precipitation, reduced evapotranspiration, and continued thawing of the soil active layer (zone of annual freeze and thaw) and permafrost the result of high late summer soil temperatures. Because of the increase in regional temperatures, especially in late spring, melting of the soil active layer and permafrost may be increasing and watershed runoff during the study period could be higher than a longer-term record would indicate. The increase in thaw depth of the soil active layer from mid-June to the end of July along transects tangential to the stream ranged from 15 to 50 cm on the south aspect and 30 to 40 cm on the north (R. Stottlemyer, unpublished data, 1997–1999).
Soils and Forest Floor
Data collected during a growing season, or even several years, must be interpreted with caution. Particular care is needed when assessing the magnitude of process response to physical factors such as forest floor and soil temperature or moisture. Present process rates often reflect prior conditions. At Asik, a change in depth of the soil active layer or permafrost melt rates would further complicate defining what may only be short-term cause–effect linkages. For this reason, such quantification was minimized in this paper, and the focus was on looking for potential relationships among forest floor processes and chemical and physical characteristics.
The higher forest floor O2 and mineral soil moisture on southern aspects likely reflected the greater melting of the soil active layer and perhaps discontinuous permafrost, especially in the transition zone plots. The greater moisture in the transition plots helped account for the absence of watershed-level differences in O2 temperature by aspect. Differences in temperature by aspect beneath spruce, where the soil active layer is deep or absent and permafrost is discontinuous, supported this hypothesis. Differences in O2 moisture also contributed to the changes in temperature by vegetation type.
The correlation between O2 moisture and C content suggests moisture may account for the higher C content and C/N ratios on southern aspects in the watershed. The large SOM pools and high moisture in the transition zone plots likely contributed disproportionately to watershed-level differences in O2 C content with aspect. Deleting the transition plots from the analysis shows no aspect difference in C content among the 40 remaining plots. Beneath spruce, the absence of O2 and soil moisture change with aspect likely accounted for the lack of difference in C and N pool size and C/N ratios.
The O2 N pools in the vegetation transition and tussock plots were comparable to other values in the Arctic, but C pools at Asik appeared lower (Cheng et al., 1998). Beneath spruce, soil N pools were 50% greater than observed at Rock Creek, Alaska (Stottlemyer, 1992), or Wallace Lake, Michigan, near the southern boreal–hardwood ecotone (Stottlemyer et al., 1998), and 20% above the average for interior Alaskan taiga (Van Cleve et al., 1983). Beneath alder, the Asik O2 contained about 70% of the N found at Rock Creek. However, Rock Creek had a much higher alder stem density.
Seasonal O2 inorganic N pools were comparable to what has been found at Wallace Lake (Stottlemyer and Toczydlowski, 1999). Mean concentrations were somewhat higher than the mean of samples collected in July 1990 adjacent to the Asik watershed (Binkley et al., 1994). This difference can be accounted for by low inorganic N pools in July relative to June and August. The low inorganic N pool size beneath alder reflected the sparse distribution of individual plants and perhaps the season sampled. An earlier study at Asik, in different plots beneath alder, spruce, willow, tussuck, and Dryas sp. using ion exchange resin bags collected after a full year, indicated alder had the highest inorganic N availability (Binkley et al., 1994).
Monthly net N mineralization rates for June beneath spruce were comparable to late spring results from a multiyear study in Wallace Lake (Stottlemyer and Toczydlowski, 1999). However, net N mineralization rates for the full growing season at Wallace Lake and Rock Creek were positive beneath spruce and alder. At Asik, immobilization exceeded mineralization rates. The O2 sequestering of inorganic N during the growing season suggests that the forest floor may be a C sink during this time. The low growing season temperatures in surface soils and organic horizons at Asik may help account for the low net N mineralization rates. In laboratory incubations of soils from the Asik watershed, net N mineralization rates increased (p < 0.001) when temperatures were raised from 5 to 12°C (Binkley et al., 1994). The warmest in situ temperatures at Asik were beneath alder (mean daily = 7.0°C). The mean daily temperature met or exceeded 12°C beneath spruce only 12 d during the growing season, and just 5 d beneath alder.
The regional temperature increases in spring (April–May) could warm surface organic and shallow soil horizons, reduce their moisture, and increase C efflux and N mineralization rates. Soil respiration in northern ecosystems may respond more to temperature and moisture change than nutrient availability (Illeris and Jonasson, 1999). Overall, O2 C and N mineralization rates in the Asik watershed appear to respond positively to temperature during the growing season. At the watershed level, forest floor and soil CO2 efflux were correlated with temperature and inorganic N pools. Beneath spruce, there was a weak correlation between CO2 efflux and temperature on the north aspect, but a stronger (p < 0.001) correlation on the south aspect. The maximum mean daily O2 temperature on any north aspect spruce plot was 10°C, and on the south aspect 18°C. Such results suggest O2 and soil temperature thresholds, above which respiration responses might be stronger to temperature increases, assuming moisture is not limiting. In laboratory incubations, Nadelhoffer et al. (1991) found soil C and N mineralization rates were insensitive to soil temperatures <9°C, but doubled or more from 9 to 15°C. In a warm desert ecosystem, Conant et al. (2000) found soil respiration rates were related to C pool size, but soil respiration rates decreased with increased temperature and decreased moisture. At present, soil moisture does not appear to limit O2 or soil C and N mineralization rates at Asik.
Stream Chemistry
Stream water base cation (CB) and HCO-3 concentrations were greater than observed in most high concentration upland watersheds in North America (Herlihy et al., 1991). Concentrations were similar to levels in Rock Creek, Denali National Park, Alaska (Stottlemyer, 1992) and semi-arid Utah (Bond, 1979). Stream water CB concentrations can provide indications of water source and age (Rice and Bricker, 1995). In early and mid-June, the low Ca2+ concentrations suggest that discharge from the watershed was dominated by "new water" from snow and ice melting (McNamara et al., 1997). Later in summer, the increase in stream water Ca2+ concentrations suggests greater contributions from "old" water deeper in the soil with longer flowpath length to the stream. The sharp decline in Ca2+, SO2-4, and HCO-3 (not plotted) concentrations with >10 cm of total rainfall in late July and early August further supports this hypothesis. Stream water Ca2+ concentrations at midelevation showed a similar trend until mid-July, when increased thawing of the soil active layer and permafrost followed by precipitation in late July and early August diluted concentrations. Alpine stream water Ca2+ concentrations suggest a combination of factors such as short flowpath, less reactive soil and more talus, and domination by new water contribute to low concentrations. In alpine stream water, the absence of seasonal change in Ca2+ concentration suggests rapid penetration of precipitation and melt waters through porous soil and talus to the stream with little contribution from substrate exchange or weathering (Stottlemyer et al., 1997). Alpine stream water showed less correlation among ion concentrations, suggesting short-term variation in source area as may result from diurnal change in discharge (Stottlemyer and Troendle, 1992). The magnitude of alpine diurnal discharge at Asik is difficult to quantify because of subsurface flow.
The correlations among subalpine stream water Ca2+ and SO2-4 concentrations suggest the processes that regulate Ca2+ also influence SO2-4 concentration. The subalpine stream water SO2-4 concentrations exceeded levels in poorly buffered and most well-buffered eastern U.S. streams with much greater atmospheric SO2-4 inputs (Herlihy et al., 1991).
Asik stream water NO-3–N concentrations were also higher than observed in most surface waters, even in regions where atmospheric NO-3 inputs are elevated. The average Asik alpine stream water NO-3 concentration (27 µeq L-1) exceeded values in many headwater streams throughout the eastern USA (Kaufmann et al., 1991), where atmospheric inputs are typically 30 to 60 times greater than at Asik. But, undisturbed Alaskan streams often have high NO-3–N concentrations year-round (Stottlemyer, 1992). The Asik concentrations likely reflect a combination of variation in depth of the soil active zone and scattered permafrost channeling more runoff through shallow soils where DIN pools are greater, and low biological sequestration, especially in the alpine.
Stream water NO-3–N concentrations at the mouth varied little during the growing season, and declined only slightly during late summer and early fall when stream water discharge increased. In-stream nitrification of NH+4–N can be an important process affecting NO-3–N concentration (Mulholland et al., 2000). However, the Asik stream NH+4–N concentrations, even in the alpine, were at detection levels suggesting this process was not important. Seasonal change in in-stream NO-3–N uptake and near-stream riparian sequestering may also be factors (Folster, 2000). Stream periphyton growth was substantial, and NO-3–N is often limiting in summer. The NO-3–N concentration pattern may also indicate that, regardless of flowpath, the ecosystem was leaking mineralized N during the growing season (Stottlemyer et al., 1997). The alpine stream NO-3–N concentrations especially suggest leakage during the growing season since precipitation inputs are small. In alpine headwaters where talus dominates, the small amount of soil and vegetation can still dominate soil- and stream water DIN concentrations (Campbell et al., 1995). The Asik alpine is more exposed to wind, which would reduce snowpack protection to the soil (Mitchell et al., 1996). In early spring there may be more frequent freeze–thaw cycling in the soil active layer. Freeze–thaw cycles in soils have a pronounced effect on biotic activity and nutrient loss (Biederbeck and Campbell, 1971). Soil freeze–thaw cycles release N and C from soil microbial communities, and have been found to significantly contribute to soil NO-3 (Campbell et al., 1971; Boutin and Robitaille, 1995). Such processes have not yet been studied in the Asik alpine. In general, the trend in subalpine soil DIN pools during the growing season matches the changes in stream water NO-3–N concentration at the mouth. However, to date there are too few data to quantify this relation.
The similar pattern in seasonal alpine and subalpine stream water DOC concentrations suggests the sources were alike throughout the watershed. The increase in stream water discharge and DOC concentrations during spring melt suggest flushing of large soil and forest floor reservoirs accumulated from the previous fall and winter. Concentrations increased again in late July as melting soils and permafrost, and greater precipitation amounts increased stream water discharge. Stream water DOC was the only solute for which the concentrations were similar between the alpine and mouth of the watershed during June and into July. The concentration similarity suggests that, despite smaller alpine SOM pools, substantial portions of SOM are mobile before and during the growing season. The small change in stream water concentrations from the alpine to the watershed mouth also suggests relatively low in-stream biological use (Sun et al., 1997). The divergence in alpine and subalpine stream water DOC concentrations in July and early August indicates a substantial increase in DOC contributions from the subalpine. Alpine stream water was not sampled later than early August.
Nutrient Budgets
Caution must be used in making estimates of watershed input and output chemical budgets when only a few years of data are available. This is particularly true for watersheds where seasonal change in the depth of the soil active layer and discontinuous permafrost are present and annual temperatures have been warming. This combination could increase annual runoff, and the change in meltwater amount may alter seasonal flowpaths to the stream. Any factor that increases the amount of annual runoff may also increase the relative percentage of runoff that moves laterally through shallower soils. This can increase nutrient export, especially for relatively mobile ions such as K+ and NO-3–N (Stottlemyer and Troendle, 1999). The decline in stream water ion concentrations with discharge was less than a 1:1 ratio, and NO-3–N declined the least with increased discharge. We have estimated that watershed DIN output in stream water exceeds precipitation inputs by 70% (Herrmann et al., 2000). The DON output was estimated at <1% the DOC output.
The output of stream water DOC averaged 0.015 mg m-2 d-1, and dissolved inorganic carbon (DIC) averaged 0.3 mg m-2 d-1. The amount of C exported in stream water was small compared with CO2 losses to the atmosphere from the soil surface in the Asik watershed and other sites in northwest Alaska (Oechel et al., 1995; McKane et al., 1997). The Asik watershed DOC output was <10% of long-term outputs at the southern boreal Wallace Lake watershed in Michigan (Stottlemyer et al., 1998). With warming temperatures, the potential for increased DOC export to the aquatic ecosystem appears high.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
