SEARCH

SEARCH BY CITATION

Keywords:

  • Arctic tundra ecosystem;
  • spatial and temporal variability;
  • net ecosystem CO2 exchange;
  • ecosystem respiration;
  • controlling factors;
  • climate change

Abstract

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

[1] Temporal and spatial variability in the Arctic introduces considerable uncertainty in the estimation of the current carbon budget and Arctic ecosystem response to climate change. Few representative measurements are available for land-surface parameterization of the Arctic tundra in regional and global climate models. In this study, the eddy covariance technique was used to measure net ecosystem CO2 exchange (NEE) of Alaskan wet sedge tundra and moist tussock tundra ecosystems during the summer (i.e., 1 June to 31 August) from 1999 to 2003 in order to quantify the seasonal and spatial variability in NEE and to determine controlling factors on NEE in these tundra ecosystems. Warmer and drier conditions prevailed for the moist tussock tundra compared with that of the wet sedge tundra. Over the 5-year period, the wet sedge tundra was a sink for carbon of 46.4 to 70.0 gC m−2 season−1, while the moist tussock tundra either lost carbon of up to 60.8 gC m−2 season−1 or was in balance. The contrasting patterns of carbon balance at the two sites demonstrate that ecosystem difference can be more important in determining landscape NEE than intraseasonal and interseasonal variability due to environmental factors with respect to NEE. The wet sedge tundra showed an acclimation (e.g., over days) to temperature, while the moist tussock tundra illustrated a strong temperature dependence. Warming and drying accentuated ecosystem respiration in the moist tussock tundra, causing a net loss of carbon. Better characterization of spatial variability in NEE and associated environmental controls is required to improve current and future estimates of the Arctic terrestrial carbon balance.

1. Introduction

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

[2] A better understanding of the carbon budget in the Arctic region is essential for accurate assessment and prediction of regional responses and associated feedbacks on climate change [Chapin et al., 2005]. The Arctic has an important role in the global carbon budget due to the presence of up to 190 Gt of carbon currently stored in cold tundra soils [Post et al., 1982; Schlesinger, 1991; Hobbie et al., 2000] and the potential for substantial positive feedbacks to global climate change through enhanced soil respiration and carbon release to the atmosphere under a projected warmer and drier climate [Maxwell, 1992; Oechel et al., 1993; Oechel and Vourlitis, 1994; Ping et al., 1997; Vourlitis and Oechel, 1997; Grogan and Chapin, 2000]. Recent observations report that thermal profiles of permafrost and northern latitude temperature records have risen 2°–4°C across northern Alaska and Canada during the last few decades [Lachenbruch and Marshall, 1986; Oechel et al., 1993; Osterkamp and Romanovsky, 1999]. This change in temperature will eventually affect the hydrology and thermal regime as well as feedback processes of Arctic tundra ecosystems which will ultimately influence the direction and magnitude of the carbon budget of the Arctic tundra region [Tieszen et al., 1980; Kane et al., 1991; Chapin and Shaver, 1996; Johnson et al., 1996, 2000; Oechel et al., 2000b; Chapin et al., 2005].

[3] Modeling efforts have been conducted in an attempt to understand the regional carbon balance over the Arctic [Clein et al., 2000; McGuire et al., 2000; Oechel et al., 2000a; Vourlitis et al., 2003]. Owing to the lack of representative measurements, a single land-surface parameterization for the Arctic tundra is used in regional and global climate models [McGuire et al., 1992; Bonan, 1995; Lynch et al., 1995; Clein et al., 2000]. Model simulation studies of terrestrial carbon dynamics give a wide range of predictions ranging from small to large carbon sinks and sources for high latitude regions [Wang and Polglase, 1995; King et al., 1997; Xiao et al., 1998; Clein et al., 2000]. These differing carbon flux estimates arise in part from a lack of information and understanding regarding the spatial and temporal variability of Arctic carbon dynamics and associated physical and environmental controls on these processes [Walker et al., 1989; McFadden et al., 1998; Vourlitis et al., 2000b].

[4] Synthesis studies on factors controlling net ecosystem CO2 exchange (NEE) of a diversity of ecosystems have been published as part of the FLUXNET effort [Janssens et al., 2001; Law et al., 2002; Baldocchi, 2003]. Desert ecosystems are primarily limited by water with the timing and frequency of precipitation events a critical factor in determining seasonal and annual productivity [Hastings et al., 2005]. Grasslands are typically found in areas of moderate summer rainfall along with cold winters and often found to be in balance with respect to NEE on an annual basis [Frank and Dugas, 2001; Suyker and Verma, 2001]. Temperate hardwood and deciduous forests have some of the highest annual NEE values (often equal to tropical forests) limited by summer temperatures [Barford et al., 2001; Curtis et al., 2002]. Tropical forests NEE values are often found to have little seasonal variation but large interannual differences ranging from a slight source during an El Niño event changing to a strong annual sink as the warm phase dissipates [Loescher et al., 2003].

[5] Measurements of NEE have been conducted in the Alaskan Arctic using chamber (e.g., leaf, soil, and ecosystem level) and eddy covariance methods [Chapin et al., 1980; Oberbauer et al., 1991; Oechel et al., 1993; Chapin and Shaver, 1996; Vourlitis and Oechel, 1997; Oechel et al., 2000b; Harazono et al., 2003]. Although these studies have provided a quantitative understanding of the carbon budget and a mechanistic understanding of how tundra ecosystems respond to direct and indirect environmental changes, a limitation of most of the previous studies conducted over the short term (e.g., one or two seasons) is they may overlook responses (e.g., homeostatic adjustment and interaction among environmental factors) that are essential in determining carbon balance over the long term [Oechel and Strain, 1985; Leadley and Reynolds, 1992]. Measurements from plot and local scale cannot be representative of the Arctic region owing to spatial heterogeneity [McFadden et al., 1998, 2003; Vourlitis et al., 2003] and make it difficult to extrapolate results from instantaneous measurements of carbon and nutrient gain to long-term ecosystem responses to climatic change [Chapin and Shaver, 1996] owing to species interactions in a community which may mediate ecosystem processes, and different time lags between physiological, demographic, and ecosystem processes.

[6] One of the characteristics of the Arctic landscape is spatial heterogeneity in tundra vegetation types, topography, hydrology, and climate at numerous scales (i.e., from meters to hundreds of kilometers) [McFadden et al., 1998; Mullier et al., 1999; Hinkel et al., 2001; Vourlitis et al., 2003]. For instance, distinctive spatial variability in land cover (vegetation type) over the Arctic region is apparent at different resolution as shown in Figure 1. The key processes and responses of the Arctic tundra are determined by different controlling factors at a local scale and can result in potential variations in the pattern and magnitude of NEE [Fan et al., 1992]. Walker et al. [1998] reported the difference in the magnitude of gross photosynthesis, carbon gain, and respiration between a moist acidic tundra (pH < 5.5) and a moist nonacidic tundra (pH > 6.5) over the Kuparuk River region despite very similar environmental conditions other than pH. When CO2 flux measurements obtained only from the moist acidic tundra are extrapolated over the same region, there is an overestimation in gross photosynthesis by at least 35%, respiration by 140%, and NEE by about 15% [Walker et al., 1998]. These results show the potential error that is introduced in generalizing a local-scale measurement to a larger scale and the uncertainty in estimating regional carbon budgets [McFadden et al., 1998; Walker et al., 1998; Vourlitis et al., 2000b].

image

Figure 1. Land cover maps with three different resolutions over the location of the measurements sites. (a) Land cover map with the resolution of 0.03 × 0.03 km obtained from a satellite image (Landsat and SPOT) over the Barrow region. (b) NDVI map with the resolution of 1 × 1 km obtained from a satellite image (NOAA 14) over the same region in Figure 1a. (c) NDVI map with the same resolution as Figure 1b over a larger regional scale including the two measurements sites. Figures 1a and 1b are enlarged views of the square area in Figure 1c. (d) Land cover map with the resolution of 40 × 60 km. Each color represents different land cover types. Barrow represents the wet coastal sedge tundra, while Atqasuk represents the moist tussock tundra.

Download figure to PowerPoint

[7] Although the importance of understanding the spatial variability has been emphasized in conjunction with the regional carbon balance, there are gaps in flux measurements with respect to the spatial range of variability in NEE as well. In particular, there are insufficient extensive measurements conducted geographically which can be used to characterize the Arctic region over a longer time period [Chapin et al., 2000]. In order to begin to understand the temporal and spatial variability in carbon balance and to alleviate the uncertainty in the carbon budget over the Arctic tundra region, this study was conducted in two major tundra types in the Alaskan Arctic North Slope region (i.e., a wet sedge tundra and a moist tussock tundra) [Joint Federal State Land Use Planning Commission for Alaska, 1973; Mullier et al., 1999] using the eddy covariance method from 1999 to 2003. The data used for the analysis in this study were limited to the summer period (1 June to 31 August), which is the main period of plant growth. These two sites are separated by 100 km and are representative of about 62% of the Alaskan Arctic North Slope region using land-cover classification according to major ecosystems of Alaska [Joint Federal State Land Use Planning Commission for Alaska, 1973; Mullier et al., 1999]. One of the objectives of this study was to characterize the temporal and spatial patterns of NEE for the Arctic region using the carbon flux measurements from two different tundra ecosystems. Even though these locations have inherent differences (coastal versus inland), we wanted to determine similar patterns of environmental controls on NEE could be identified. The measurements made over the five years in this study allow us to determine how responsive these ecosystems are with respect to NEE to climate variation. Specific questions asked are: (1) to what degree are spatially different tundra ecosystems similar in their response to environmental factors with respect to carbon balance and (2) what is more important with respect to annual carbon balance of Arctic tundra ecosystems, ecosystem difference or temporal variance?

2. Materials and Methods

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

2.1. Site Description

[8] The study was conducted in the wet sedge tundra and the moist tussock tundra during the summer season from 1999 to 2003. The wet sedge tundra site (71°19′21.09″N:156°36′33.17″W) is located in the vicinity of Barrow, Alaska, while the moist tussock tundra site (70°28′10.6″N:157°24′32.2″W) is located approximately 100 km south of Barrow in the vicinity of Atqasuk, Alaska (Figure 1). The wet sedge tundra site is about 2 km south of the Arctic Ocean and is adjacent to the National Oceanic and Atmospheric Administration (NOAA) Climate Monitoring and Diagnostics Laboratory (CMDL). The site is composed of low- and high-centered polygons, ice wedges (polygon troughs), and drained thaw lakes [Sellmann et al., 1972]. The site is characterized by low species diversity, dominance of grasses and sedges, rare occurrences of tussock tundra, and an absence of shrubs [Brown et al., 1980]. About 80% of plant vegetation is dominated by herbaceous sedges including Carex aquatilis, Eriophorum russeolum, and Eriophorum angustifolium, and the shrub Salix rotundifolia [Webber, 1978]. Mosses (e.g., Calliergon richardsonii and Cinclidium subrotundum) and lichens (e.g., Peltigera sp.) are also abundant. A moss layer under the vascular plants provides almost 100% cover. The wet sedge tundra is dominated by wet, acidic soils (soil pH 5.2) [Walker et al., 2003]. Owing to the presence of permafrost, soil drainage is poor throughout the summer season. Soils at the measurement site have a tripartite morphology: an organic-rich surface horizon, a horizon of silty clays to silt loam textured mineral material, and an underlying perennially frozen organic-rich horizon [Brown et al., 1980]. The depth of the soil organic matter (SOM) from the top surface varies from 0 to 10 cm. The mineral substratum consists of silty clays, which are generally layered from about 10 to 30 cm in depth. Total SOM content is 29 kg C m−3 down to a depth of 1 m (C. L. Ping and G. J. Michaelson, personal communication, 2001).

[9] The moist tussock tundra site near Atqasuk is composed of a variety of moist-wet coastal sedge tundra with moist-tussock vegetation predominating in well-drained upland areas [Batzli, 1980] and is characterized as acidic tundra (soil pH 4.8) [Walker et al., 2003]. Vegetation is dominated by the tussock forming sedge, Eriophorum vaginatum as well as other evergreen and deciduous forbs and shrubs [Komárková and Webber, 1980; Walker et al., 1989]. Soils are developed on aeolian sands of Quaternary age [Everett, 1980] and consist of approximately 95% sand and 5% clay and silt to a depth of 1 m [Walker et al., 2003]. The depth of SOM ranges from 0 to 18 cm. Total organic carbon content to a depth of 1 m is 38 kg C m−3, approximately 1.3 times greater than the wet sedge tundra (C. L. Ping and G. J. Michaelson, personal communication, 2001).

2.2. Eddy Covariance and Environmental Measurements

[10] Measurements of NEE were made using the eddy covariance technique [Baldocchi et al., 1988] from a tower height of 5 m for the wet sedge tundra and 3 m for the moist tussock tundra. Wind velocity (i.e., vertical, streamwise, and crosswind speed) and temperature were measured with a three-dimensional sonic anemometer (Model R3, Gill Instruments, Lymington, UK) at 10 Hz for both sites. To measure CO2 and water vapor fluctuations, an open-path infrared gas analyzer (IRGA) designed by NOAA Atmospheric Turbulence and Diffusion Division (ATDD) was used in 1999 [Auble and Meyers, 1992] and an open-path IRGA (Model LI-7500, LI-COR, Inc., Lincoln, Nebraska) from 2000 to 2003 at the wet sedge tundra site, while a NOAA/ATDD IRGA was used from 1999 to 2000 and a LI-7500 IRGA from 2001 to 2003 at the moist tussock tundra site. Half-hour eddy covariances and associated statistics were calculated online from 10-Hz raw data and stored on a personal computer. Fluctuating components for the flux calculations were obtained using a 400-s running mean and digital recursive filter [McMillen, 1986, 1988]. No linear detrending was used in the flux calculations in this study. The sonic anemometer data set was rotated to force the mean crosswind and vertical wind speeds to zero and to align the streamwise wind with the mean wind vector [McMillen, 1986]. The CO2 and water vapor fluxes were corrected for the variation in air density due to simultaneous transfers of water vapor and sensible heat according to Webb et al. [1980].

[11] Environmental measurements were sampled every minute, averaged over 30 min, and logged using a data logger (Model 21X, Campbell Scientific Inc., Logan, Utah). Net radiation (Rn) and photosynthetically active radiation (PAR) were measured with a net radiometer (Model Q7, REBS, Seattle, Washington) and a quantum sensor (Model LI-190SB, LI-COR, Inc.) respectively at a height of approximately 1.2 m for both sites. Air temperature (Tair) and relative humidity (RH) were measured using a PRT (platinum resistance thermometers) and capacitive polymer RH chip (Model HMP45C-L, Vaisala Inc., Helsinki, Finland) at a height of 3.0 m for the wet sedge tundra and at a height of 1.5 m for the moist tussock tundra. Soil temperature (Tsoil) was measured using type-T thermocouples at 0, 5, and 10 cm below the ground at the wet sedge tundra site and 0 and 5 cm at the moist tussock tundra site. Ground heat flux (G) was measured using two soil heat flux plates (Model HFT-1, REBS) at the wet sedge tundra and four soil heat flux plates at the moist tussock tundra. The soil heat flux plates were buried 2 cm below the moss surface at both sites. Soil moisture content was measured at different depths (e.g., 5 cm, 10 cm, 20 cm, and 30 cm from the surface) using time domain reflectometery (TDR) (Model CS610, Campbell Scientific) at both sites. Soil moisture content was calculated using an empirical equation for the organic soil [Roth et al., 1992] and a different empirical equation for the mineral soil [Topp et al., 1980].

[12] Active layer depth (the depth of thaw), defined as the vertical distance between the surface and the point where a steel rod meets resistance, was measured every week from mid June (following seasonal snowmelt) to late August for both sites. There were seven 400-m transects, which were laid out every 45° from true north (excluding 315° owing to the presence of a building), with 27 flags marking locations of measurements at the wet sedge tundra site, while there were eight 200-m transects laid out every 45° from true north with 20 flags at the moist tussock tundra site. Three measurements of the active layer depth were conducted around each flag.

2.3. Energy Balance Closure

[13] The closure of the energy balance was examined by a linear regression between the sum of latent heat (LE) and sensible heat (H) fluxes (H + LE) and the available energy (Rn − G) for both sites. Using half-hourly data collected during the five summer seasons, the energy balance closure at the wet sedge tundra showed 76% agreement (i.e., H + LE = 0.76 × (Rn − G), r2 = 0.90), while that at the moist tussock tundra was slightly better, showing H + LE = 0.81 × (Rn − G) with r2 = 0.94. Previous studies report that energy balance closure was improved when daily average fluxes are used for the assessment of the energy closure [Blanken et al., 1997; Scott et al., 2004]. However, the analysis of the energy closure using the daily average values showed no improvement of the energy closure in this study.

[14] A possible cause for the lack of energy closure may arise from measurement errors in Rn (associated with calibration factor of net radiometer) [Halldn and Lindroth, 1992; Twine et al., 2000]. Replacing domes without calibration and harsh winter conditions may result in a change in the calibration factor and generate a bias in Rn. Errors and uncertainties in the spatial characterization of Rn and G may be another reason for the energy imbalance [McFadden et al., 1998; Anthoni et al., 1999]. Source areas of Rn and G are relatively small compared to that of LE and H [Schmid, 1997]. The Arctic tundra ecosystem is heterogeneous on scales of meters in microtophography, hydrology, and microclimate from hummocks and frost action [Weller and Holmgren, 1974; McFadden et al., 1998]. Therefore, within the scale of the source areas, albedo, soil temperature, soil moisture content, and heat capacity may differ [McFadden et al., 1998]. Although the two sites are flat and have long fetches (>500 m), heterogeneity in scales may cause temperature and water vapor gradient along the fetches and generate horizontal advection of H and LE, another reason for the observed energy imbalance.

2.4. Gap-Filling Methods

[15] The percentage of data collection during the five summer measurement periods was 74% at the wet sedge tundra site and 68% at the moist tussock tundra site. The gaps in data collection were primarily due to power failures, instrument errors, and data rejection following quality assessments. In order to calculate daily and seasonal carbon balance, a look-up table method was used to fill these missing data [Falge et al., 2001]. The look-up tables were created to represent changing environmental conditions and plant physiology based on monthly periods. The interval of the look-up table was 50 μmol m−2 s−1 with a range of 0 to 100 μmol m−2 s−1 and 100 μmol m−2 s−1 with a range of 100 to 1600 μmol m−2 s−1 for PAR and 2°C with a range of −8.0 to 28.0°C for Tair. Values not present in the look-up table were linearly interpolated from the measured values [Falge et al., 2001].

3. Results

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

3.1. General Environmental Condition

[16] General climate conditions during the measurement periods are presented in Table 1. Both sites are located above the Arctic Circle (66°34′); thus the Sun remains above the horizon from 10 May to 2 August, resulting in 24 hours of daylight. We refer to early morning and late evening during this period; however, there is always measurable incoming light radiation. There was relatively little difference in the 5-year summer average PAR between the sites, while the 5-year summer average Tair and Tsoil at the wet sedge tundra were significantly lower by 4.4°C in Tair and 5.5°C in Tsoil. The 5-year summer average VPD at the wet sedge tundra was much lower owing to the influence of the Arctic Ocean, which brings higher humidity air to the coastal inland [Harazono et al., 1998]. Rain events occurred mainly in July and August for both sites with slightly less precipitation at the wet sedge tundra. In general, the climate conditions at the moist tussock tundra were noticeably warmer and drier than at the wet sedge tundra over the five summers.

Table 1. Comparison of Summer (1 June to 31 August) Mean Weather Conditions of Photosynthetically Active Radiation (PAR), Air Temperature (Tair), Surface Soil Temperature (Tsoil), Precipitation (PPT), and Vapor Pressure Deficit (VPD) at the Wet Sedge Tundra and the Moist Tussock Tundra Sites From 1999 to 2003
VariableYearSiteDifferencea
Wet SedgeMoist Tussock
  • a

    Difference was calculated by subtracting values at the moist tussock tundra from those at the wet sedge tundra.

  • b

    Summer mean values were not available due to a lack of the measurements at the moist tussock tundra in 2001.

  • c

    Five-summer mean was calculated using the values from 1999 to 2003 for each site.

PAR, μmol m−2 d−11999415.0403.111.9
2000369.4379.2−9.8
2001421.3N/Ab 
2002389.7362.327.4
2003295.3391.9−96.6
5-summer meanc378.1384.2−6.1
Tair, °C19993.07.7−4.7
20002.76.6−3.9
20011.6N/Ab 
20021.86.3−4.5
20032.16.1−4.0
5-summer mean2.26.7−4.5
Tsoil, °C19990.38.0−7.7
20001.06.6−5.6
2001−0.5N/Ab 
20021.66.8−5.2
20034.15.8−1.7
5-summer mean1.36.8−5.5
VPD, kPa19990.070.25−0.18
20000.090.20−0.11
20010.07N/Ac 
20020.080.26−0.18
20030.100.22−0.12
5-summer mean0.080.23−0.15
PPT, mm199957.866.2−8.4
200086.6117.0−30.4
200165.5N/Ab 
200256.847.59.3
200350.153.4−3.3
5-summer mean63.471.0−7.6

[17] The timing of the snow-free period, determined from surface albedo, at the wet sedge tundra varied considerably during the measurement years: day of year (DOY) 164 in 1999, DOY 165 in 2000, DOY 162 in 2001, DOY 144 in 2002, and DOY 155 in 2003 [Stone et al., 2002; R. Stone, personal communication, 2004]. The occurrence of snowmelt at the moist tussock tundra was determined using upwelling solar irradiance and was DOY 154 in 2001, DOY 137 in 2002, and DOY 151 in 2003 (source of data: Atmospheric Radiation Measurement (ARM) program, http://www.archive.arm.gov/). The earliest snowmelt during the measurement periods occurred in 2002 and was due to unusually warm temperatures in May. The early snowmelt caused by warmer temperature at the moist tussock tundra initiated plant growth 4–8 days earlier than the wet sedge tundra during the measurement years.

3.2. Seasonal Trend of Net Ecosystem CO2 Exchange and Environmental Factors

3.2.1. Diurnal Variation of NEE

[18] At the wet sedge tundra, the diurnal pattern of NEE in June exhibited a range of variation from 0.06 to −0.07 gC m−2 hr−1, depending on plant photosynthesis and temperature (Figure 2). A strong net carbon uptake occurred in June 2002, coinciding with an early snowmelt and high NDVI values (e.g., 0.3–0.4) (J. Gamon, personal communication, 2002). The diurnal patterns of NEE in July and August showed that the maximum values of NEE corresponded with the peak in PAR during a day (data not shown). The duration of carbon uptake in July was longer than in June or August owing to net carbon uptake during late evening and early morning and was the result of high PAR coupled with high photosynthetic potential (higher NDVI). The shortened time of net carbon uptake in August was mainly due to the combination of a shorter photoperiod, reduced PAR, and lower plant productivity due to plant senescence.

image

Figure 2. Average diurnal trends for each half-hourly datum by month of NEE for the summer seasons from 1999 to 2003. Each point represents a half-hour average of NEE; open circles indicate the values of NEE at the wet sedge tundra; solid lines indicate the values of NEE at the moist tussock tundra. Positive sign indicates a carbon source, while negative sign indicates a carbon sink.

Download figure to PowerPoint

[19] The moist tussock tundra was a net source of carbon during most of the diurnal period in June (the exception being 2002), with the peak carbon source occurring between 1200 and 1600 hours. The magnitude of carbon source in July and August was smaller relative to June and the ecosystem became a weak carbon sink in the morning. During late evening and early morning periods, the ecosystem remained a weak carbon source throughout the summer. Overall, the patterns and amplitudes of NEE at the moist tussock tundra showed strong intraseasonal and interseasonal variability compared to the wet sedge tundra, changing from a carbon source in June to a weak carbon sink in July and August.

3.2.2. Seasonal Variation of Daily Environmental Factors and NEE
3.2.2.1. Active Layer Depth

[20] The trend of the active layer depth at the wet sedge tundra exhibited a rapid linear increase to a depth of 30 cm around DOY 200 (Figure 3b). The thaw rate of the active layer slowed after DOY 200, reaching a maximum depth of 32–40 cm by DOY 230. At the moist tussock tundra, the active layer depth increased linearly throughout the summer (Figure 4b) and did not show the decreased thaw rate in late summer as shown in the wet sedge tundra. During the measurement periods, the maximum thaw depth for both sites occurred in 1999, reaching 40 cm at the wet sedge tundra and 43 cm at the moist tussock tundra, coinciding with maximum summer temperatures observed from 1999 to 2003 (Table 1).

image

Figure 3. Seasonal trends of (a) daily photosynthetic active radiation (PAR), air temperature (Tair), and NEE, and (b) daily soil moisture (SM), active layer depth (TD), precipitation (PPT), and NEE each year at the wet sedge tundra and the moist tussock tundra from 1999 to 2003. In order to understand the processes controlling NEE working at different timescales (e.g., PAR for instantaneous effect; active layer depth for a longer-term effect), environmental factors were separated as shown above and daily NEE is displayed in both Figures 3a and 3b. Soil moisture data were obtained at 5 cm below the ground at both sites.

Download figure to PowerPoint

image

Figure 4. Captions as shown in Figure 3 but for the moist tundra site. The soil moisture data collected in 2002 and 2003 were excluded from the analysis owing to unacceptable data quality. In order to display clearly the fluctuation of variables, the scale of the variables has been adjusted and is different than in Figure 3b.

Download figure to PowerPoint

3.2.2.2. Soil Moisture Condition

[21] The seasonal trend of soil moisture at 5 cm below the surface is presented in Figures 3 and 4. Soil moisture at both sites was less than 0.1 m3 m−3 before snowmelt in early June. The maximum soil moisture at the wet sedge tundra typically occurred between DOY 175 and 182 reflecting the initiation of the seasonal thaw cycle (Figure 3). The different timing of the maximum soil moisture appeared directly related to Tsoil, which caused an earlier thaw and increased soil permeability and infiltration of snowmelt water into the soil. For example, relatively warm Tsoil (around 4.0°C) in late June of 2000 resulted in the earliest thaw, while relatively cold Tsoil ranging from 0° to 2°C in late June for the other years slowed thaw activity. After the peak in the soil moisture each summer, the soil moisture content showed small increases coinciding with precipitation events but remained relatively constant until the end of the summer. The low variability in soil moisture was due to consistently waterlogged soils resulting from poor drainage. The highest soil moisture was observed in 2000 and was a result of increased summer precipitation in the Barrow region (Table 1).

[22] At the moist tussock tundra, the soil moisture sharply increased and reached an initial peak at around 0.53 m3 m−3 near DOY 165 (Figure 4). The earlier thaw at the moist tussock tundra was primarily due to warmer Tsoil. There was considerable variation in the soil moisture throughout the summer and increases in the soil moisture coincided with rainfall (e.g., DOY 183 to 190 in 2000 and DOY 210 to 216 in 2001). Precipitation, rapid drainage, and a higher evapotranspiration (e.g., five-summer average daily evapotranspiration of 1.24 mm d−1 at the moist tussock tundra, while 0.83 mm d−1 at the wet sedge tundra) (H. J. Kwon, unpublished data, 2005) under warm temperatures strongly influenced soil moisture content of the surface soil layer.

3.2.2.3. Daily NEE

[23] The trends of daily NEE at the two sites differed markedly in direction and magnitude during the measurement period (Figures 3 and 4). Before the snowmelt (DOY 137 to 164), the patterns of daily NEE at the wet sedge tundra was nearly neutral over most of the measurement years (Figure 3). Substantial increases in carbon loss appeared after the snowmelt. While there was a strong carbon sink in the early summer of 2002, this may be due to relatively high rates of photosynthesis and canopy development as a result of the early snowmelt in May. The timing and magnitude of the daily maximum carbon sink or source varied from season to season (Table 2 and Figure 3). The peak daily net carbon uptake generally occurred between mid July and mid August, while the peak of daily carbon loss occurred between late June and early July. As the season progressed, the rate of net carbon uptake gradually decreased and the wet sedge tundra remained in balance in late August. The daily variation in NEE was relatively independent of the environmental factors; daily average PAR, Tair, and soil moisture (the relationships between daily NEE and these factors showed low R2 (<0.2) each year). The wet sedge tundra did not respond consistently to similar environmental conditions (e.g., colder, wetter periods for daily net carbon uptake or warmer, drier periods for daily carbon loss) with respect to NEE during the measurement period. The active layer depth, which correlates with longer-term seasonal development of the ecosystem, was not associated with the seasonal patterns of daily NEE either.

Table 2. Magnitudes of Daily Maximum Carbon Gain and Loss (Unit in gC m−2 d−1) and Total Accumulation of Carbon (Unit in gC m−2 season−1) Each Year During the Measurement Periodsa
 Wet Sedge TundraMoist Tussock Tundra
Daily Max. Carbon GainbDaily Max. Carbon LossbCumulative NEEDaily Max. Carbon GainbDaily Max. Carbon LossbCumulative NEE
  • a

    Positive sign indicates a carbon source, while negative sign indicates a carbon sink.

  • b

    Numbers in parentheses represent day of year (DOY).

  • c

    The measurement period of the data was from DOY 196 to 243.

1999−5.4 (223)+2.3 (174)−70.0−1.8 (232)+1.8 (172)+30.9
2000−2.8 (200)+0.7 (196)−46.4−3.4 (152)+3.2 (194)+60.8
2001−2.9 (223)+0.7 (159)−51.7−0.9 (207)c+0.8 (230)c−2.0
2002−4.4 (152)+0.9 (175)−60.8−0.9 (204)+1.1 (226)+2.7
2003−4.3 (202)+1.9 (199)−48.8−1.3 (197)+1.4 (164)−1.1

[24] The moist tussock tundra lost carbon from the ecosystem to the atmosphere in early June (Figure 4). The source strength diminished after mid June and the ecosystem became a weak carbon sink in early July with the exception of 2000 which showed almost continuous carbon loss throughout the summer. NEE from early July to mid August was nearly in balance with a seasonal transition to a carbon source occurring around DOY 230 while the wet sedge tundra still remained a carbon sink. In contrast to the wet sedge tundra, the variation of daily NEE was coupled with changes in environmental factors at the moist tussock tundra. The loss of carbon, in particular, coincided with a reduction in soil moisture resulting from a lack of rainfall and increasing Tair under high levels of PAR (e.g., DOY 181 to 186 in 1999 and DOY 193 to 196 in 2000) (Figure 4).

[25] Cumulative NEE over the summer measurement years differed remarkably between the two study sites (Figure 5). The wet sedge tundra was a relatively strong seasonal carbon sink, while the moist tussock tundra was an equally strong seasonal carbon source (1999 and 2000) or nearly in balance (2001–2003). At the wet sedge tundra, cumulative carbon in 1999 exhibited a strong carbon loss in early July and a subsequent strong uptake in late July. The seasonal accumulation of carbon reached −70.0 gC m−2 season−1 in 1999, which was the strongest seasonal carbon gain observed in this study (Table 2). The patterns and amount of cumulative carbon in 2000, 2001, and 2003 were similar, showing a weak carbon loss from the tundra until early July and a gradual increase of net carbon uptake as the season progressed. In 2002, the early plant growth resulting from the early snowmelt caused a strong net carbon uptake in late June (around DOY 170). The magnitude of the cumulative carbon remained at −20.0 gC m−2 season−1 during a cold period between DOY 160 and 190. The carbon gain resumed around DOY 190 and the total cumulative carbon was −60.8 gC m−2 season−1 in 2002. Inter-seasonal variability of the cumulative NEE in the moist tussock tundra was more prominent than that in the wet sedge tundra, showing a dramatic change from a carbon source to a balanced carbon budget.

image

Figure 5. Cumulative NEE during the summer seasons from 1999 to 2003 at the wet sedge tundra and the moist tussock tundra. Total seasonal accumulation of NEE are indicated in parentheses.

Download figure to PowerPoint

3.3. Control of Environmental Factors on Net Ecosystem CO2 Exchange

[26] At the wet sedge tundra, ecosystem level light compensation intensities approached between 100 to 200 μmol m−2 s−1 (Figure 6), which is higher than reported for individual leaves of the principal grass species in the Barrow region [Tieszen, 1975]. Light intensity for saturation (i.e., light intensity for maximum net carbon uptake) at the ecosystem level differed from month to month, reaching 600 μmol m−2 s−1 in late June (DOY 168–181) and 1000 μmol m−2 s−1 in late July (DOY 198–212). The moist tussock tundra in early June (DOY 152–167) illustrated a weak carbon sink at low light intensity and became a carbon source at light intensities above 400 μmol m−2 s−1. The ecosystem lost significant amounts of carbon to the atmosphere, reaching a maximum value of 0.11 gC m−2 hr−1 at 1600 μmol m−2 s−1. From DOY 182 to 244, the response of NEE to changing PAR was similar, demonstrating that ecosystem level light compensation occurred at 100 μmol m−2 s−1, while light saturation was between 600 and 700 μmol m−2 s−1. During these periods, the magnitude of net carbon uptake was significantly lower than the wet sedge tundra over the same 100 to 700 μmol m−2 s−1 PAR range. At light levels above the saturation point, there was a carbon loss as high as 0.12 gC m−2 hr−1. While some uptake of carbon occurred during the summer, the predominant direction of NEE was a net loss of carbon to the atmosphere. The source activity may have resulted from increased rates of ecosystem respiration (ER) with an increase in temperature as PAR increased.

image

Figure 6. Response of NEE to variation in PAR at the wet sedge tundra during the summer seasons from 1999 to 2003. Half-hourly data of NEE were categorized in light intervals of 0–50 μmol m−2 s−1, 50–100 μmol m−2 s−1, and every 100 μmol m−2 s−1 from 100 to 1600 μmol m−2 s−1 and averaged for each interval for each half month during the 5-year measurement period. Error bars indicate standard deviations.

Download figure to PowerPoint

image

Figure 6. (continued)

Download figure to PowerPoint

[27] The effect of Tair on NEE in June at the wet sedge tundra was variable, illustrating repeated periods of a source and a sink throughout the whole Tair range (Figure 7). In July, the magnitude of NEE at lower Tair (e.g., 2° to 4°C) was as high as at higher Tair (e.g., 16° to 18°C). The magnitude of NEE tended to gradually decrease as Tair increased from 4° to 16°C. The wet sedge tundra in August was a net carbon sink varying from −0.2 to −0.5 gC m−2 hr−1 throughout most of the Tair range. In contrast, at the moist tussock tundra, there were a few occurrences of a small amount of carbon gain at different Tair in different months (e.g., −4° to 0°C in June, 10° to 18°C in July, and 10° to 14°C in August). In general, there was a consistent carbon loss throughout the range of Tair and the degree of carbon loss increased at a higher Tair at the moist tussock tundra. In order to understand how soil temperature influences NEE (via ER), the relationship between NEE and soil temperature was examined at both sites and showed similar results with the relationship between NEE and Tair (data not shown).

image

Figure 7. Relationship of NEE and air temperature (Tair) during the summer seasons from 1999 to 2003. Half-hourly data of NEE were categorized by temperature intervals of 2°C and averaged for each interval for each month during the 5-year measurement period. Error bars indicate standard deviations.

Download figure to PowerPoint

[28] Figure 8 illustrates the relationship between NEE and VPD at both sites showing how the two different tundra ecosystems are affected by VPD. During the measurements, the range of variation in VPD was greater as a result of a higher maximum value of VPD in the moist tussock tundra. The variability in NEE at both sites was considerably higher when VPD was low, and it is assumed that the values of NEE were influenced by other environmental factors such as PAR. The wet sedge tundra tended to gain more carbon at low VPD, while at high VPD, the ecosystem response varied. At the moist tussock tundra, the average trend of NEE showed a gradual increase in carbon loss as VPD increased. As VPD increased, increasing the lower bound (i.e., minimum value) of NEE structured the linear relationship. This provides a good example of how an environmental factor can affect variation in NEE by defining a lower limit.

image

Figure 8. Influence of vapor pressure deficit (VPD) on NEE at the wet sedge tundra and the moist tussock tundra. Half-hourly data collected from 1000 to 1600 hours (i.e., the main hours of photosynthetic activity) were categorized by VPD with 0.25-kPa intervals during the 5-year measurement periods. Solid circles represent average NEE at each interval of VPD, while thick soil lines represent linear regression lines of the average NEE data. Percentage of NEE data between −0.3 and −0.6 gC m−2 hr−1 was less than 3% of the total collected data at both sites, and those data are excluded in the data presentation above.

Download figure to PowerPoint

[29] Environmental factors such as PAR, Tair, and VPD can have simultaneous and confounding affects on the variation of NEE. In order to assess the relationship of NEE with simultaneous changes in PAR, Tair, and VPD, a stepwise multiple regression analysis was performed in SYSTAT (version 10, SYSTAT Software Inc., Point Richmond, California) and used the data (24 hours) collected each summer at two different sites. This analysis also included the interaction terms among PAR, Tair, and VPD to understand relative importance of the joint effect of two or three variables on NEE (Table 3). The values of R2 were small (<0.5) most of year except the year 2000 at the moist tussock tundra, indicating that the degree to which the environmental factors were related to NEE was low. However, the coefficients corresponding to the majority of the environmental factors including the interaction term were statistically significant. Magnitude and sign of the coefficients varied among years at the wet sedge tundra. PAR showed a consistently negative coefficient with different magnitude of the coefficients. At the moist tussock tundra, some environmental factors (e.g., PAR and PAR*Tair) showed inconsistency in the coefficient values and signs, while Tair, PAR*VPD, and Tair*VPD showed positive coefficients. VPD alone showed negative coefficients. Time lags were found between NEE and environmental variables and the stepwise multiple regression analysis was conducted after the data were corrected for time lags. The analysis with time lag correction showed very similar results (R2 and coefficient values) with the analysis without time lag correction (results not shown).

Table 3. Statistical Information for the Relationship of NEE With Changes in PAR, Tair, and VPD Using a Stepwise Multiple Regression Modela
YearEffectWet Sedge TundraMoist Tussock Tundra
R2 ValueCoef.Std ErrPR2 ValueCoef.Std ErrP
  • a

    Model: YNEEi = β0 + β1 XPARi + β2XTairi + β3XVPDi + β4XPARi*XTairi + β5XPARi*XVPDi + β6XTairi*XVPDi + β7XPARi*XTairi *XVPDi. Regression coefficient (Coef) and standard error (Std Err) are reported. Dashes indicate nonsignificance of the effect on NEE.

  • b

    CONS, P, T, and V indicate constant (intercept), PAR, Tair, and VPD, respectively.

  • c

    Asterisk indicates cross product of the variables.

1999CONSbR2 = 0.12−0.027160.0039P < 0.001R2 = 0.340.001470.0017P < 0.01
Pb −0.000040.0000P < 0.001 
Tb  0.001070.0003P < 0.001
Vb 0.222350.0560P < 0.001 −0.128770.0211P < 0.001
P*Tc −0.000030.0000P < 0.001 0.000000.0000P < 0.001
P*Vc 0.001190.0001P < 0.001 0.000460.0000P < 0.001
T*Vc −0.033180.0044P < 0.001 0.002420.0011P < 0.05
P*T*Vc  −0.000010.0000P < 0.001
2000CONSR2 = 0.190.015850.0029P < 0.001R2 = 0.550.016960.0027P < 0.001
P −0.000110.0000P < 0.001 0.000080.0000P < 0.001
T 0.002670.0007P < 0.001 0.001560.0005P < 0.001
V −0.309410.0507P < 0.001 −0.402060.0350P < 0.001
P*T −0.000010.0000P < 0.001 −0.000010.0000P < 0.001
P*V 0.000940.0001P < 0.001 0.000800.0001P < 0.001
T*V 0.012760.0040P < 0.01 0.017260.0026P < 0.001
P*T*V −0.000040.0000P < 0.001 −0.000030.0000P < 0.001
2001CONSR2 = 0.240.015180.0023P < 0.001R2 = 0.230.026980.0020P < 0.001
P −0.000170.0000P < 0.001 −0.000110.0000P < 0.001
T  
V  −0.163480.0278P < 0.001
P*T −0.000010.0000P < 0.001 
P*V 0.000850.0001P < 0.001 0.000360.0001P < 0.001
T*V  0.004410.0016P < 0.01
P*T*V −0.000050.0001P < 0.001 −0.000010.0000P < 0.001
2002CONSR2 = 0.27R2 = 0.170.012570.0012P < 0.001
P −0.000150.0000P < 0.001 −0.000050.0000P < 0.001
T  
V 0.205140.0531P < 0.001 −0.013180.0048P < 0.01
P*T −0.000010.0000P < 0.001 0.000000.0000P < 0.001
P*V 0.001120.0001P < 0.001 0.000220.0000P < 0.001
T*V −0.021980.0049P < 0.001 
P*T*V −0.000040.0000P < 0.001 0.000000.0000P < 0.001
2003CONSR2 = 0.06R2 = 0.07
P −0.000220.0000P < 0.001 0.000020.0000P < 0.001
T −0.003830.0006P < 0.001 0.002420.0004P < 0.001
V  −0.155380.0283P < 0.001
P*T 0.001350.0003P < 0.001 −0.000010.0000P < 0.001
P*V  0.000270.0000P < 0.001
T*V 0.018480.0059P < 0.01 0.004540.0015P < 0.01
P*T*V −0.000090.0000P < 0.001 −0.000010.0000P < 0.001

4. Discussion

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

4.1. Patterns of Net Ecosystem CO2 Exchange

[30] The seasonal pattern of daily NEE at both sites exhibited substantial increases in carbon release after the snowmelt, reaching a maximum carbon loss of ∼3.0 gC m−2 d−1 for the moist tussock tundra. In part, this is attributed to substantial microbial activity with the rise in Tsoil [Brooks et al., 1997; Bilbrough et al., 2000; Olsson et al., 2003] and the release of trapped CO2 during the phase change from ice to soil water [Coyne and Kelly, 1971; Friborg et al., 1997]. A similar increase in CO2 efflux during the initial thawing period has been recorded for the Arctic tundra [Fahnestock et al., 1998; Harazono et al., 2003]. At the wet sedge tundra, the daily maximum carbon sink occurred from mid July to mid August, which coincided with the maximum canopy development as indicated by ground-based measurements of NDVI ranging from 0.49 to 0.58 (data collected from 2000–2002) (J. Gamon, personal communication, 2002). The decreased rates of carbon gain in the late summer season resulted from the combination of reduced PAR, and a decrease in plant productivity due to the onset of plant senescence [Tieszen, 1975, 1978].

[31] The moist tussock tundra was frequently a small source of carbon to the atmosphere or a weak sink. While we were unable to partition the net flux into GPP and ER, it is thought that the relative contribution of ER was tightly linked with the magnitude of carbon loss at the moist tussock tundra. Warmer and drier conditions at the moist tussock tundra would intensify ER [Johnson et al., 1996] and either lessen carbon gain to the ecosystem or enhance carbon loss to the atmosphere. In a study where moist tussock and wet sedge tundra at the same latitude were measured, the latter incorporated 1.5 times more carbon despite a twofold lower estimated GPP [Vourlitis et al., 2000a]. The higher carbon uptake by the wet sedge was attributed in part to a lower ER. Warmer, drier conditions and a lower water table have consistently resulted in less carbon uptake, or a greater loss of carbon in Arctic ecosystems [Oechel et al., 1993, 1998], primarily by an increase in aerobic respiration.

[32] The strong carbon gain occurring in early June 2002 (unlike other measurement years) at the wet sedge tundra may be a result of a relatively high photosynthetic capacity due to an early onset of plant development following the earlier disappearance of snow [Oechel and Sveinbjörnsson, 1978; Tieszen, 1978; Semikhatova et al., 1992] and inhibition of ER from low temperatures and the frozen soils at the time [Grant and Rochette, 1994]. This result suggests that early snowmelt in spring, which extends the length of the growing season, will affect seasonal carbon balance by increasing carbon sequestration to the ecosystem [Aurela et al., 2004].

4.2. Environmental Controls on Net Ecosystem CO2 Exchange

[33] Previous results from individual leaf studies, photosynthesis of the Arctic plants show seldom light saturated [Tieszen, 1973, 1978; Grulke et al., 1990]. In this study, however, light saturation at the whole ecosystem level was observed between 600 and 1000 μmol m−2 s−1 at both sites. At the wet sedge tundra site, inhibition of carbon gain or eventually carbon loss (only observed early in the summer), after light saturation, may be due to an increase in ER, photo-inhibition or photo-oxidation of the photosynthetic apparatus [Oechel and Collins, 1976] and/or water stress in the tundra plants [Stoner and Miller, 1975; Caldwell et al., 1978; Miller et al., 1978; Grulke and Bliss, 1988] under high radiation and temperature. At the moist tussock tundra, the pattern of an exponential increase in NEE with increasing PAR suggests that light was indirectly controlling NEE by increasing temperature and ER resulting in carbon loss from the tundra ecosystem. The contrast in the light response of NEE at the two different sites suggests differences in physiology and the seasonal development of the ecosystems, and in part to difference in response to environmental conditions. These results suggest that PAR has a direct and immediate effect on controlling NEE before light saturation occurs whereas at higher levels of PAR, other factors such as temperature and water stress may have a greater influence on increasing ER. ER appeared to play a larger role in determining NEE at the moist tussock tundra site than the wet sedge tundra site.

[34] The wet sedge tundra showed relative independence from temperature with respect to NEE, while for the moist tussock tundra, a considerable increase in carbon loss was observed as temperature increased. On the basis of leaf-level gas exchange studies at Barrow [Tieszen, 1975, 1978; Limbach et al., 1982], the optimal air temperature for photosynthesis ranges between 10° and 20°C. In this study, the optimum temperature for the whole ecosystem level at the wet sedge tundra was less distinct, illustrating high rates of net carbon uptake over a range of temperatures from 0° to 24°C. This demonstrates that the wet sedge tundra is adapted to a wide range of air temperature [Tieszen, 1978; Chapin et al., 1995] and that the photosynthetic ability of the ecosystem may not be constrained by low air temperatures [Billings, 1975]. Other possible reasons for the lack of influence of temperature on NEE may be (1) a stronger effect of PAR than temperature on NEE over the short-term period (i.e., 30 min), (2) a time lag of the system's response to temperature change, and (3) the effect of secondary limitations such as water stress from increased temperatures.

[35] On the other hand, the increase in carbon loss at the moist tussock tundra site may be due to a greater increase in ER than in photosynthesis at high temperature and low soil moisture conditions [Shaver et al., 1992; Johnson et al., 1996; Sommerkorn et al., 1999] even with adequate radiation. Owing to a lack of stomatal control, mosses and lichens can be highly responsive to high temperature and low soil moisture. Desiccation of the moss tissues can enhance the rates of respiration [Oechel and Sveinbjörnsson, 1978; Skre and Oechel, 1981]. However, above ground respiration including vascular plants and mosses is relatively small contributing about 10% to the total ER [Green and Lange, 1994; Sommerkorn et al., 1999], though it accounts for more than 83% of total biomass (g m−2) at the moist tussock tundra site [Walker et al., 2003]. Soil respiration is highly responsive to temperature changes with increasing rates of microbial respiration when the soil changes from saturated to mesic condition [Heal et al., 1981; Peterson et al., 1984]. Soil respiration consists of about 90% of the total ER [Sommerkorn et al., 1999] and can exceed gains from gross photosynthesis resulting in the moist tussock tundra becoming a net carbon source. A higher ER is attributed to the well-drained soil at the moist tussock tundra resulting in an increase in soil temperature and better soil aeration, and thus an increase in aerobic microbial activity and decomposition rates [Billings et al., 1983; Johnson et al., 1996].

[36] Summer mean soil temperatures were consistently warmer in the moist tussock than the wet sedge tundra (Table 1). The effect of air temperature on NEE was similar to that of soil temperature, indicating that NEE at the wet sedge tundra was mostly determined by the photosynthetic capacity, while NEE at the moist tussock tundra was influenced by ER. Therefore the different responses of the two ecosystems to temperature change may be that soil temperature (near freezing except in 2003) was never high enough to have a major influence on NEE (via ER) in the wet sedge tundra, while the moist tussock tundra was more responsive to temperature changes under warmer and drier conditions.

[37] The wet sedge tundra tended to gain more carbon than it lost carbon at low VPD suggesting that the ecosystem is well adapted to low Tair and high RH conditions. Midday depression of NEE was found when the Tair and VPD were high (H. J. Kwon, unpublished data, 2005). The effect of VPD on NEE is associated not only with environmental conditions but also different physiological stages of the plants during each growing season [Tieszen, 1978]. Thus the lack of consistency in the pattern of NEE at high VPD, when all 5-year summer data were used in this analysis, may reflect the varying degree of influence of VPD on NEE depending on the sensitivity of physiological responses, which can change within and between years [Chapin and Shaver, 1996].

[38] The moist tussock tundra became a strong carbon source to the atmosphere as VPD increased. These changes in the magnitude and sign of NEE may be attributed to increased water stress between the atmosphere and the ecosystem [Stoner and Miller, 1975; Caldwell et al., 1978; Miller et al., 1978; Grulke and Bliss, 1988; Williams et al., 2000] and/or enhanced ecosystem respiration rates. As the VPD increases, the plant reduces the stomatal aperture restricting water loss [Miller et al., 1978] and limits the photosynthetic rate [Stoner and Miller, 1975; Caldwell et al., 1978]. Drier air coupled with lower soil moisture levels tends to promote increased ER and eventually increase CO2 efflux [Caldwell et al., 1978; Oberbauer et al., 1991]. The different degree of influence of water stress on NEE suggests that moist tussock tundra is more responsive to changes in both Tair and VPD than at the wet sedge tundra [Shaver et al., 1992].

[39] In order to understand how the seasonal carbon budget was affected by interseasonal variation in average climate for the summer, the relationship between cumulative NEE and seasonal average environmental factors was assessed (data not shown). Seasonal average PAR did not have an impact on cumulative NEE at either ecosystems. Warmer summers had virtually no influence on carbon loss in the wet sedge tundra, whereas increasing temperature in the moist tussock tundra had a substantial effect on carbon loss. This result implies that the moist tussock tundra is more responsive to warmer temperature than the wet sedge tundra. Although the interseasonal comparison can indicate a relationship between seasonal carbon budget and climate variation and a possible transition of the ecosystems, there are not sufficient data to report and draw a concrete conclusion in this study.

4.3. Influence of Spatial Variability of Net Ecosystem CO2 Exchange on Carbon Budget

[40] The high degree of spatial heterogeneity in Arctic tundra ecosystems at a variety of scales complicates the process of estimating a regional-scale carbon budget [McFadden et al., 1998; Vourlitis et al., 2003] and introduces considerable uncertainty into simulations of the Arctic regional response to climate change. Notably, the spatial resolutions used in parameterizations of regional-scale models are often coarse, and the large difference in carbon flux reported for adjacent ecosystems in this study is neglected.

[41] Distinctive spatial variability is presented in the land cover map with the resolution of 0.03 × 0.03 km (Figure 1a) obtained from digital analysis of Landsat multispectral scanner (MSS), Landsat thematic mapper (TM), and the Satellite Pour l'Observation de La Terre (SPOT) multispectral scanner data. Most of the spatial variability in classification is lost when the resolution of the land cover is 1 × 1 km as obtained from NOAA 14 satellite. This resolution (Figures 1b and 1c) is similar to that used in the globally applied MOD17 model estimates of net primary production using imagery from the Moderate-Resolution Imaging Spectroradiometer (MODIS) sensor [Turner et al., 2005]. Regionally applied carbon balance models such as Terrestrial Ecosystem Model (TEM) and General Ecosystem Model (GEM) simulate carbon dynamics over pan-Arctic tundra using two tundra types, polar desert/alpine tundra and wet/moist tundra, with 0.5° × 0.5° resolution [Clein et al., 2000; McGuire et al., 2000]. Figure 1c shows a land cover map created with a similar resolution used in TEM.

[42] Coarse spatial resolutions usually result in a coarsening of the land cover classification scheme. Although the MODIS-based carbon flux estimates use a 1 × 1 km resolution vegetation map, the model algorithm contains broad vegetation classes and the two sites used in this study in the Arctic region are treated as open shrublands (F. Heinsch, personal communication, 2004). TEM and GEM also treat the two measurement sites as the same vegetation type (i.e., wet/moist tundra) [Clein et al., 2000]. This lack of detailed representation of the spatial heterogeneity in classification with the regional model applications contributes to uncertainty of carbon budget estimates over the Arctic region. For example, Clein et al. [2000] estimated net ecosystem productivity (NEP) over the pan-Arctic using TEM and GEM from 1995 to 2100. The simulations spanned with actual climate input data from 1921 to 1994 and projected climates from 1995 to 2100. The magnitude of NEP simulated by the models for summer of 2000 (June∼August) for both of the sites in this study ranged from a sink of 53.5 to 60.7 gC m−2 summer−1 (J. S. Clein and A. D. McGuire, personal communication, 2004). The contrasting pattern of carbon exchange from the measurements at both sites was not represented in the model results.

[43] The discrepancy between the measurements and the predictions of the models may be associated with a lack of understanding of ecosystem function and environmental controls on different tundra ecosystems as well as problems with classification. Although there are increasing efforts to incorporate the spatial variability at the regional scale, the models still cannot capture the finer spatial scale variability which controls and determines the magnitude and direction of carbon balance. The results in this study suggest that generalization and simplification in the large-scale models lose important information such as ecosystem function, and ecosystem response and sensitivity to a changing climate. The discrepancy between the measurements and the prediction of the models highlights the significance of understanding and incorporating the spatial variation in climate and vegetation in order to reduce uncertainties in estimating carbon budget over the Arctic region.

5. Conclusion

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

[44] The results in this study show the spatial and temporal patterns and controls on NEE of two different tundra ecosystems over five summer seasons. The contrasting patterns and magnitudes of NEE at the two sites demonstrate that spatial variability over fairly short distances can be more important in landscape NEE than in intra-annual and interannual variability owing to the two ecosystem's response to environmental factors. Wet sedge tundra shows the potential for substantial carbon gain across a wide range of temperatures, suggesting rapid and effective acclimation (e.g., over days) to temperature. In contrast, the moist tussock tundra illustrates a strong temperature dependence showing a strong carbon loss as temperature increases. The difference in the ecosystem response to temperature may be from the result of influence of soil temperature on NEE (via ER). Soil temperatures were consistently warmer in moist tussock tundra than wet sedge tundra, which may be why NEE was more responsive to temperature in the moist tussock tundra. The relative effects of water stress on moist tussock tundra are stronger than on wet sedge tundra resulting in greater carbon loss at high temperature and VPD. The importance of ecosystem respiration is highlighted by the increase in carbon loss which exceeded gains from gross photosynthesis under high temperatures when upper soil layer is relatively dry.

[45] Fine-scale variability (i.e., geomorphology, low- and high-centered polygons, soil type, and hydrology) is a characteristic of the Arctic tundra region. This variability may be largely missed in current experiments and models over the larger spatial scales [Oechel and Vourlitis, 1994; McGuire et al., 2000; Vourlitis et al., 2003]. Although the eddy covariance technique can detect fine-scale variability (∼1 km) and provide information on the patterns and primary controls on carbon exchange, this approach cannot cover larger-scale variability (∼100 km) unless multiple tower measurements are applied to the regional scale. These two measurement sites therefore cannot be generalized to represent the heterogeneous characteristic of the entire Arctic region. It is difficult to extrapolate conclusions attained from one kind of tundra to another tundra region [Billings, 1973]. Extrapolations of the carbon balance from the local level to the region will require careful consideration. Currently, the regional models present function and structure of tundra ecosystems in a simpler way compared with most measurements and plot-scale ecosystem models [Clein et al., 2000; Williams et al., 2000]. Simplification of model parameters, and differences in spatial and temporal scales between measurements and models, limit the ability of the models to capture regional-scale dynamics with the accuracy and precision possible with plot-based measurements. For improved model assessment of the Arctic regional carbon balance, better characterization of spatial variability and associated environmental controls will be required. This will improve current calculation and predictions of the Arctic terrestrial carbon balance.

Acknowledgments

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

[46] This research was supported by the National Science Foundations, Arctic Systems Sciences, the Land-Atmosphere-Ice-Interactions (LAII) Program (OPP-9732105), the Terrestrial Ecological Research Initiative (TECO) Program (DEB-973004), and the Study of Environmental Arctic Change (SEARCH) Program (OPP-0119060). Logistical support was provided by personnel from the Barrow Arctic Science Consortium (BASC) and VECO Polar Resources (VPR). We also gratefully acknowledge Faith Ann Heinsch, David Turner, and Fred Huemmrich for their constructive comments as well as field assistance provided by Glen Kinoshita, Kimberly Davis, and Phillip Lambert. We thank reviewers for reviewing and improving this manuscript.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusion
  8. Acknowledgments
  9. References
  10. Supporting Information
  • Anthoni, P. M., B. E. Law, and M. H. Unsworth (1999), Carbon and water vapor exchange of an open-canopied ponderosa pine ecosystem, Agric. For. Meteorol., 95, 151168.
  • Auble, D. L., and T. P. Meyers (1992), An open path, fast response infrared absorption gas analyzer for H2O and CO2, Boundary Layer Meteorol., 59, 243256.
  • Aurela, M., T. Laurila, and J. P. Tuovinen (2004), The timing of snow melt controls the annual CO2 balance in a subarctic fen, Geophys. Res. Lett., 31, L16119, doi:10.1029/2004GL020315.
  • Baldocchi, D. D. (2003), Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of ecosystems: Past, present and future, Global Change Biol., 9, 479492.
  • Baldocchi, D. D., B. B. Hicks, and T. P. Meyers (1988), Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods, Ecology, 69, 13311340.
  • Barford, C. C., et al. (2001), Factors controlling long- and short-term sequestration of atmospheric CO2 in a mid-latitude forest, Science, 294, 16881691.
  • Batzli, G. O. (1980), Patterns of vegetation and herbivory in arctic tundra: Results from the Research on Arctic Tundra Environments (RATE) Program, Arct. Alp. Res., 12, 401518.
  • Bilbrough, C. J., J. M. Welker, and W. D. Bowman (2000), Early spring nitrogen uptake by snow-covered plants: A comparison of arctic and alpine plant function under the snowpack, Arct. Antarct. Alp. Res., 32, 404411.
  • Billings, W. D. (1973), Arctic and alpine vegetations: Similarities, differences, and susceptibility to disturbance, BioScience, 23, 697704.
  • Billings, W. D. (1975), Arctic and alpine vegetation: Plants adaptations to cold summer climates, in Arctic and Alpine Environments, edited by J. D. Ives, and R. G. Barry, pp. 403443, Methuen, New York.
  • Billings, W. D., J. O. Luken, D. A. Mortensen, and K. M. Peterson (1983), Increasing atmospheric carbon dioxide and soil nitrogen on the carbon balance of tundra microcosms, Oecologia, 65, 2629.
  • Blanken, P. D., T. A. Black, P. C. Yang, H. H. Neumann, Z. Nesic, R. Staebler, G. den Hartog, M. D. Novak, and X. Lee (1997), Energy balance and canopy conductance of a boreal aspen forest: Partitioning overstory and understory components, J. Geophys. Res., 102, 28,91528,927.
  • Bonan, G. B. (1995), Land-atmosphere CO2 exchange simulated by a land surface process model coupled to an atmospheric general circulation model, J. Geophys. Res., 100, 28172831.
  • Brooks, P. D., S. K. Schmidt, and M. W. Williams (1997), Winter production of CO2 and N2O from alpine tundra: Environmental controls and relationship to inter-system C and N fluxes, Oecologia, 110, 403413.
  • Brown, J., K. R. Everett, P. J. Webber, S. F. MacLean Jr., and D. F. Murray (1980), The coastal tundra at Barrow, in An Arctic Ecosystem: The Coastal Tundra at Barrow, Alaska, edited by J. Brown et al., pp. 129, Dowden, Hutchinson and Ross, Stroudsburg, Pa.
  • Caldwell, M. M., D. A. Johnson, and M. Fareed (1978), Constraints on tundra productivity: Photosynthetic capacity in relation to solar radiation utilization and water stress in Arctic and Alpine tundras, in Vegetation and Production Ecology of an Alaskan Arctic Tundra, edited by L. L. Tieszen, pp. 324342, Springer, New York.
  • Chapin, F. S.III, and G. R. Shaver (1996), Physiological and growth responses of Arctic plants to a field experiment simulating climatic change, Ecology, 77, 822840.
  • Chapin, F. S.III, P. C. Miller, W. D. Billings, and P. I. Coyne (1980), Carbon and nutrient budgets and their control in coastal tundra, in An Arctic Ecosystem: The Coastal Tundra at Barrow, Alaska, edited by J. Brown et al., pp. 458482, Dowden, Hutchinson and Ross, Stroudsburg, Pa.
  • Chapin, F. S.III, G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Launder (1995), Responses of Arctic tundra to experimental and observed changes in climate, Ecology, 76, 694711.
  • Chapin, F. S.III, et al. (2000), Arctic and boreal ecosystems of western North America as components of the climate system, Global Change Biol., 6, 211223.
  • Chapin, F. S.III, et al. (2005), Role of land-surface changes in Arctic summer warming, Science, 310, 657660.
  • Clein, J. S., B. L. Kwiatkowski, A. D. McGuire, J. E. Hobbie, E. B. Rastetter, J. M. Melillo, and D. W. Kicklighter (2000), Modeling carbon responses of tundra ecosystems to historical and projected climate: A comparison of a plot-and a global-scale ecosystem model to identify process-based uncertainties, Global Change Biol., 6, 127140.
  • Coyne, P. I., and J. J. Kelly (1971), Release of carbon dioxide from frozen soil to the arctic atmosphere, Nature, 234, 407408.
  • Curtis, P. S., P. J. Hanson, P. Bolstad, C. Barford, J. C. Randolph, H. P. Schmid, and K. B. Wilson (2002), Biometric and eddy-covariance based estimates of annual carbon storage in five eastern North American deciduous forests, Agric. For. Meteorol., 113, 319.
  • Everett, K. R. (1980), Distribution and variability of soils near Atkassok, Alaska, Arct. Alp. Res., 12, 433446.
  • Fahnestock, J. T., M. H. Jones, P. D. Brooks, D. A. Walker, and J. M. Welker (1998), Winter and early spring CO2 efflux from tundra communities of northern Alaska, J. Geophys. Res., 103, 29,02329,027.
  • Falge, E. D., et al. (2001), Gap filling strategies for defensible annual sums of net ecosystem, Agric. For. Meteorol., 107, 4369.
  • Fan, S. M., S. C. Wofsy, P. S. Bakwin, D. J. Jacob, S. M. Anderson, P. L. Kebabian, J. B. McManus, C. E. Kolb, and D. R. Fitzjarrald (1992), Micrometeorological measurements of CH4 and CO2 exchange between the atmosphere and subarctic tundra, J. Geophys. Res., 97, 16,62716,643.
  • Frank, A. B., and W. A. Dugas (2001), Carbon dioxide fluxes over a northern, semiarid, mixed-grass prairie, Agric. For. Meteorol., 108, 317326.
  • Friborg, T., T. R. Christensen, and H. Soegarrd (1997), Rapid response of greenhouse gas emission to early spring thaw in a subarctic mire as shown by micro-meteorological techniques, Geophys. Res. Lett., 24(23), 30613066.
  • Grant, R. F., and P. Rochette (1994), Soil microbial respiration at different temperatures and water potentials: Theory and mathematical modeling, Soil Sci. Soc. Am. J., 58, 16811690.
  • Green, T. G. A., and O. L. Lange (1994), Photosynthesis in poikilohydric plants: A comparison of lichens and bryophytes, in Ecophysiology of Photosynthesis, edited by E. D. Schulze, and M. M. Caldwell, pp. 391342, Springer, New York.
  • Grogan, P., and F. S. Chapin III (2000), Initial effects of experimental warming on above- and below-ground components of net ecosystem CO2 exchange in Arctic tundra, Oecologia, 125, 512520.
  • Grulke, N. E., and L. C. Bliss (1988), Comparative life history characteristics of two high arctic grasses, northwest territories, Ecology, 69, 484496.
  • Grulke, N. E., G. H. Reichers, W. C. Oechel, U. Hjelm, and C. Jaeger (1990), Carbon balance in tussock tundra under ambient and elevated CO2, Oecologia, 83, 483494.
  • Halldn, S., and A. Lindroth (1992), Errors in net radiometry: Comparison and evaluation of sic radiometer designs, J. Atmos. Oceanic Technol., 9, 762783.
  • Harazono, Y., M. Yoshimoto, M. Mano, G. L. Vourlitis, and W. C. Oechel (1998), Characteristics of energy and water budgets over wet sedge and tussock tundra ecosystems at North Slope in Alaska, Hydrol. Processes, 12, 21632183.
  • Harazono, Y., M. Mano, A. Miyata, R. C. Zulueta, and W. C. Oechel (2003), Inter-annual carbon dioxide uptake of a wet sedge tundra ecosystem in the Arctic, Tellus, Ser. B, 55, 215231.
  • Hastings, S. J., W. C. Oechel, and A. Muhlia-Melo (2005), Diurnal, seasonal and annual variation in the net ecosystem CO2 exchange of a desert shrub community (Sarcocaulescent) in Baja California, Mexico, Global Change Biol., 11, 927936.
  • Heal, O. W., P. W. Flanagan, D. D. French, and S. F. Maclean (1981), Decomposition and accumulation of organic matter, in Tundra Ecosystems, edited by L. C. Bliss, O. W. Heal, and J. J. Moore, pp. 587633, Cambridge Univ. Press, New York.
  • Hinkel, K. M., F. Paetzold, F. E. Nelson, and J. G. Bockheim (2001), Patterns of soil temperature and moisture in the active layer and upper permafrost at Barrow, Alaska: 1993–1999, Global Planet. Change, 29(3–4), 293309.
  • Hobbie, S. E., J. P. Schimel, S. E. Trumbore, and J. Randerson (2000), Controls over carbon storage and turnover in high latitude soils, Global Change Biology, 6, 196210.
  • Janssens, I. A., et al. (2001), Productivity overshadows temperature in determining soil and ecosystem respiration across European forests, Global Change Biol., 7, 269278.
  • Johnson, L. C., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, E. R. Rastetter, J. A. Laundre, and G. L. Murray (1996), Effects of drainage and temperature on carbon balance of tussock tundra microcosms, Oecologia, 108, 737748.
  • Johnson, L. C., G. R. Shaver, D. H. Cades, E. Rastetter, K. Nadelhoffer, A. Giblin, J. Laundre, and A. Stanley (2000), Plant carbon-nutrient interactions control CO2 exchange in Alaskan wet sedge tundra ecosystems, Ecology, 81, 453469.
  • Joint Federal State Land Use Planning Commission for Alaska (1973), Major Ecosystems of Alaska, U.S. Geol. Surv., Fairbanks, Alaska.
  • Kane, K. D., L. Hinzman, M. Woo, and K. R. Everett (1991), Arctic hydrology and climate change, in Arctic Ecosystems in a Changing Climate, edited by F. S. Chapin III et al., pp. 3557, Elsevier, New York.
  • King, A. W., W. M. Post, and S. D. Wullschleger (1997), The potential response of terrestrial carbon storage to changes in climate and atmospheric CO2, Clim. Change, 35, 199227.
  • Komárková, V., and P. J. Webber (1980), Two low Arctic vegetation maps near Atkasook, Alaska, Arct. Alp. Res., 12, 447472.
  • Lachenbruch, A. H., and B. V. Marshall (1986), Changing climate: Geothermal evidence from permafrost in the Alaskan Arctic, Science, 234, 689696.
  • Law, B. E., et al. (2002), Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation, Agric. For. Meteorol., 113, 97120.
  • Leadley, P. W., and J. F. Reynolds (1992), Long-term response of an Arctic sedge to climate change: A simulation study, Ecol. Appl., 2(4), 323340.
  • Limbach, W. E., W. C. Oechel, and W. Lowell (1982), Photosynthetic and respiratory responses to temperature and light of three Alaskan tundra growth forms, Holarctic Ecol., 5(2), 150157.
  • Loescher, H. W., S. F. Oberbauer, H. L. Gholz, and D. B. Clark (2003), Environmental controls on net ecosystem-level carbon exchange and productivity in a Central American tropical wet forest, Global Change Biol., 9, 396412.
  • Lynch, A. H., W. L. Chapman, J. E. Walsh, and G. Weller (1995), Development of a regional climate model of the western Arctic, J. Clim., 8, 15551570.
  • Maxwell, J. B. (1992), Arctic climate: Potential for change under global warming, in Arctic Ecosystems in a Changing Climate: An Ecophysiological Perspective, edited by F. S. Chapin III et al., pp. 1134, Elsevier, New York.
  • McFadden, J. P., F. S. Chapin III, and D. Y. Hollinger (1998), Subgrid-scale variability in the surface energy balance of Arctic tundra, J. Geophys. Res., 103, 28,94728,961.
  • McFadden, J. P., W. Eugster, and F. S. Chapin III (2003), A regional study of the controls on water vapor and CO2 exchange in Arctic tundra, Ecology, 84, 27622776.
  • McGuire, A. D., J. M. Melillo, L. A. Joyce, D. W. Kicklighter, A. L. Grace, B. Moore III, and C. J. Vorosmarty (1992), Interaction between carbon and nitrogen dynamics in estimating net primary productivity for potential vegetation in North America, Global Biogeochem. Cycles, 6, 101124.
  • McGuire, A. D., J. S. Clein, J. M. Melillo, D. W. Kicklighter, R. A. Meier, C. J. Vorosmarty, and M. C. Serreze (2000), Modelling carbon responses of tundra ecosystems to historical and projected climate: Sensitivity of pan-Arctic carbon storage to temporal and spatial variation in climate, Global Change Biol., 6, 141159.
  • McMillen, R. T. (1986), A BASIC Program for eddy correlation in non-simple terrain, Tech. Memo. ERL ARL-147, 32 pp., NOAA, Silver Spring, Md.
  • McMillen, R. T. (1988), An eddy correlation technique with extended applicability to non-simple terrain, Boundary Layer Meteorol., 43, 231245.
  • Miller, P. C., W. A. Stoner, and J. R. Ehleringer (1978), Some aspects of water relations of arctic and alpine regions, in Vegetation and Production Ecology of an Alaskan Arctic Tundra, edited by L. L. Tieszen, pp. 343356, Springer, New York.
  • Mullier, S. V., A. E. Racoviteanu, and D. A. Walker (1999), Landsat MSS-derived land-cover map of northern Alaska: Extrapolation methods and a comparison with photo-interpreted and AVHRR-derived maps, Int. J. Remote Sens., 20, 29212946.
  • Oberbauer, S. F., J. D. Tenhunen, and J. F. Reynolds (1991), Environmental effects on CO2 efflux from water track and tussock tundra in Arctic Alaska, U. S. A. Arct. Alp. Res., 23, 162169.
  • Oechel, W. C., and N. J. Collins (1976), Comparative CO2 exchange patterns in mosses from two tundra habitats at Barrow, Alaska, Can. J. Bot., 54, 13551369.
  • Oechel, W. C., and B. R. Strain (1985), Native species response to increased atmospheric CO2 concentration, in Direct Effects of Increasing Carbon Dioxide on Vegetation, edited by B. R. Strain, and J. C. Cure, pp. 117154, U.S. Dep. of Energy, Natl. Tech. Info. Serv., Springfield, Va.
  • Oechel, W. C., and B. Sveinbjörnsson (1978), Primary production processes in Arctic bryophytes at Barrow, Alaska, in Vegetation and Production Ecology of an Alaskan Arctic Tundra, edited by L. L. Tieszen, pp. 269298, Springer, New York.
  • Oechel, W. C., and G. L. Vourlitis (1994), The effects of climate change on land-atmosphere feedbacks in Arctic tundra regions, Trends Ecol. Evol., 9, 324329.
  • Oechel, W. C., S. J. Hastings, G. L. Vourlitis, M. Jenkins, G. Riechers, and N. Grulke (1993), Recent change of Arctic tundra ecosystems from a net carbon dioxide sink to a source, Nature, 361, 520523.
  • Oechel, W. C., G. L. Vourlitis, S. J. Hastings, R. Ault, and P. Bryant (1998), The effects of water table manipulation and elevated temperature on the net CO2 flux of wet-sedge tundra ecosystems, Global Change Biol., 4, 7790.
  • Oechel, W. C., et al. (2000a), A scaling approach for quantifying the net CO2 flux of the Kuparuk River Basin, Alaska, Global Change Biol., 6, 160173.
  • Oechel, W. C., G. L. Vourlitis, S. J. Hastings, R. C. Zulueta, L. Hinzman, and D. Kane (2000b), Acclimation of ecosystem CO2 exchange in the Alaskan Arctic in response to decadal climate warming, Nature, 406, 978981.
  • Olsson, P. Q., M. Sturm, C. H. Racine, V. Romanovsky, and G. E. Liston (2003), Five stages of the Alaskan arctic cold season with ecosystem implications, Arct. Antarct. Alp. Res., 25, 7481.
  • Osterkamp, T. E., and V. E. Romanovsky (1999), Evidence for warming and thawing of discontinuous permafrost in Alaska, Permafrost Periglacial Processes, 10, 1737.
  • Peterson, K. M., W. D. Billings, and D. N. Reynolds (1984), Influence of water table and atmospheric CO2 concentration on the carbon balance of Artic tundra, Arct. Alp. Res., 16, 331335.
  • Ping, C. L., G. J. Michaelson, and J. M. Kimble (1997), Carbon storage along a latitudinal transect in Alaska, Nutr. Cycl. Agroecosyst., 49, 235242.
  • Post, W. M., W. R. Emanuel, P. J. Zinke, and A. G. Stangenberger (1982), Soil carbon pools and world life zones, Nature, 298, 328338.
  • Roth, C. H., M. A. Malicki, and R. Plagge (1992), Empirical evaluation of the relationship between soil dielectric constant and volumetric water content as the basis for calibrating soil moisture measurements by TDR, J. Soil Sci., 43, 113.
  • Schlesinger, W. H. (1991), Biogeochemistry: An Analysis of Global Change, 443 pp., Elsevier, New York.
  • Schmid, H. P. (1997), Experimental design for flux measurements: Matching scales of observations and fluxes, Agric. For. Meteorol., 87, 179200.
  • Scott, R. L., E. A. Edwards, W. J. Shuttleworth, T. E. Huxman, C. Watts, and D. C. Goodrich (2004), Interannual and seasonal variation in fluxes of water and carbon dioxide from a riparian woodland ecosystem, Agric. For. Meteorol., 122, 6584.
  • Sellmann, P. V., K. L. Carey, C. Keeler, and A. D. Hartwell (1972), Terrain and coastal conditions on the arctic Alaskan coastal plain, ARPA Order 1615, 72 pp., Corps of Eng., Hanover, N. H.
  • Semikhatova, O. A., T. V. Gerasimenko, and T. I. Ivanova (1992), Photosynthesis, respiration, and growth of plants in the Soviet Arctic, in Arctic Ecosystems in a Changing Climate: An Ecophysiological Perspective, edited by S. F. Chapin III et al., pp. 169192, Elsevier, New York.
  • Shaver, G. R., W. D. Billings, F. S. Chapin III, A. E. Giblin, K. J. Nadelhoffer, W. C. Oechel, and E. B. Rastetter (1992), Global change and the carbon balance of arctic ecosystems, BioScience, 42, 433441.
  • Skre, O., and W. C. Oechel (1981), Moss functioning in different taiga ecosystems in interior Alaska: I. Seasonal, phenotypic, and drought effects on photosynthesis and response patterns, Oecologia, 48, 5059.
  • Sommerkorn, M., M. Bölter, and L. Kappen (1999), Carbon dioxide fluxes of soils and mosses in wet tundra of Taimyr Peninsular, Siberia: Controlling factors and contribution to net system fluxes, Polar Res., 18(2), 253260.
  • Stone, R. S., E. G. Dutton, J. M. Harris, and D. Longenecker (2002), Earlier spring snowmelt in northern Alaska as an indicator of climate change, J. Geophys. Res., 107(D10), 4089, doi:10.1029/2000JD000286.
  • Stoner, W. A., and P. C. Miller (1975), Water relations of plant species in the wet coastal tundra at Barrow, Alaska, Arct. Alp. Res., 7, 109124.
  • Suyker, A. E., and S. B. Verma (2001), Year-round observations of the net ecosystem exchange of carbon dioxide in a native tallgrass prairie, Global Change Biol., 7, 279289.
  • Tieszen, L. L. (1973), Photosynthesis and respiration in Arctic tundra grasses: Filed light intensity and temperature response, Arct. Alp. Res., 5, 239251.
  • Tieszen, L. L. (1975), CO2 exchange in the Alaskan arctic tundra: Seasonal changes in the rate of photosynthesis of four species, Photosynthetica, 9(4), 376390.
  • Tieszen, L. L. (1978), Photosynthesis in the principal Barrow, Alaska species: A summary of field and laboratory responses, in Vegetation and Production Ecology of an Alaskan Arctic Tundra, edited by L. L. Tieszen, pp. 241266, Springer, New York.
  • Tieszen, L. L., P. C. Miller, and W. C. Oechel (1980), Photosynthesis, in An Arctic Ecosystem: The Coastal Tundra at Barrow, Alaska, edited by J. Brown et al., pp. 102139, Dowden, Hutchinson and Ross, Stroudsburg, Pa.
  • Topp, G. C., J. L. Davis, and A. P. Annan (1980), Elecromagnetic determination of soil water content: Measurement in coaxial transmission lines, Water Resour. Res., 16, 574582.
  • Turner, D. P., et al. (2005), Site-level evaluation of satellite-based global terrestrial gross primary production and net primary production monitoring, Global Change Biol., 11, 666684.
  • Twine, T. E., W. P. Kustas, J. M. Norman, D. R. Cook, P. R. Houser, T. P. Meyers, J. H. Prueger, P. J. Starks, and M. L. Wesely (2000), Correcting eddy-covariance flux underestimates over a grassland, Agric. For. Meteorol., 103, 279300.
  • Vourlitis, G. L., and W. C. Oechel (1997), Landscape-scale CO2, H2O vapor and energy flux of moist-wet coastal tundra ecosystems over two growing seasons, J. Ecol., 85, 575590.
  • Vourlitis, G. L., W. C. Oechel, A. Hope, D. Stow, B. Boynton, J. Verfaillie Jr., R. Zulueta, and S. J. Hastings (2000a), Physiological models for scaling plot measurements of CO2 flux across an Arctic tundra landscape, Ecol. Appl., 10, 6072.
  • Vourlitis, G. L., Y. Harazono, W. C. Oechel, M. Yoshimoto, and M. Mano (2000b), Spatial and temporal variations in hectare-scale net CO2 flux, respiration and gross primary production of Arctic tundra ecosystems, Funct. Ecol., 14, 203214.
  • Vourlitis, G. L., J. Verfaillie, W. C. Oechel, A. Hope, D. Stow, and R. Engstrom (2003), Spatial variation in regional CO2 exchange for the Kuparuk River Basin, Alaska over the summer growing season, Global Change Biol., 9, 930941.
  • Walker, D. A., E. Binnian, B. M. Evans, N. D. Lederer, E. Nordstrand, and P. J. Webber (1989), Terrain, vegetation, and landscape evolution of the R4D research site, Brooks Range Foothills, Alaska, Holarctic Ecol., 12, 238261.
  • Walker, D. A., et al. (1998), Energy and trace-gas fluxes across a major arctic soil pH boundary, Nature, 394, 469472.
  • Walker, D. A., et al. (2003), Vegetation-soil-thaw-depth relationships along a low-Arctic bioclimate gradient, Alaska: Synthesis of information from the ATLAS studies, Permafrost Periglacial Processes, 14, 103123.
  • Wang, Y. P., and P. J. Polglase (1995), Carbon balance in the tundra, boreal forest and humid tropical forest during climate change: Scaling up from leaf physiology and soil carbon dynamics, Plant Cell Environ., 18, 12261244.
  • Webb, E. K., G. I. Pearman, and R. Leuning (1980), Correction of flux measurements for density effects due to heat and water vapor transfer, Q. J. R. Meteorol. Soc., 106, 85100.
  • Webber, P. J. (1978), Spatial and temporal variation of the vegetation and its production, Barrow, Alaska, in Vegetation and Production Ecology of an Alaskan Arctic Tundra, edited by L. L. Tieszen, pp. 37112, Springer, New York.
  • Weller, G., and B. Holmgren (1974), The microclimates of the Arctic tundra, J. Appl. Meteorol., 13, 854862.
  • Williams, M., W. Eugster, E. B. Rastetter, J. P. McFadden, and F. S. Chapin III (2000), The controls on net ecosystem productivity along an Arctic transect: A model comparison with flux measurements, Global Change Biol., 6, 116126.
  • Xiao, A., J. M. Melillo, and D. W. Kicklighter (1998), Transient climate change and net ecosystem production of the terrestrial biosphere, Global Biogeochem. Cycles, 12, 345360.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusion
  8. Acknowledgments
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
jgrg19-sup-0001-t01.txtplain text document1KTab-delimited Table 1.
jgrg19-sup-0002-t02.txtplain text document1KTab-delimited Table 2.
jgrg19-sup-0003-t03.txtplain text document3KTab-delimited Table 3.

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