SEARCH

SEARCH BY CITATION

Keywords:

  • continental montane larch forest;
  • eddy covariance;
  • net ecosystem CO2 exchange;
  • ecosystem respiration;
  • gross ecosystem production;
  • Mongolia

Abstract

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

[1] Mongolian boreal forest merits special attention since it is located in the transitional area between the southern Siberian boreal forest and the Asian steppe zone, a vulnerable region being potentially affected by global warming and anthropogenic activities. This paper presents the first full-year-long continuous measurements of net ecosystem CO2 flux (NEE) made over a montane larch (Larix sibirica Ledeb.) forest in Mongolia from 25 March 2003 to 24 March 2004 (366 days) using the eddy covariance technique. The hourly maximum uptake was −10.1 μmol m−2 s−1. The maximum daily uptake of −4.0 g C m−2 d−1 (negative NEE values denote net carbon uptake by the canopy from the atmosphere) occurred in July. The annual cumulative NEE was −85 g C m−2, indicating that the forest acted as a net sink of CO2. We examined the responses of NEE to environmental conditions in the growing season from May to September. Both daytime 30-min mean and daily integrated NEE responded to incident photosynthetically active radiation (PAR) in a rectangular hyperbolic fashion. Model results show that the apparent quantum yield (α) was −0.0133 ± 0.0011 μmol CO2 per μmol of photons, and the bulk light use efficiency (LUE) on the daily basis was −6.7 mmol CO2 per mole of PAR photons over the entire growing season for this forest. Additionally, daily integrated NEE was also a linear function of the normalized difference vegetation index (NDVI), a linear function of mean daily air temperature (Ta), and a quadratic polynomial function of daily means of the atmospheric water vapor pressure deficit (VPD). Among these factors, LAI (as measured by NDVI) was dominant in affecting the dynamics of NEE, followed by Ta. Lower Ta was limiting the growth rate of this montane larch forest. As daily means of VPD exceeded 1.2 kPa, net CO2 uptake by the canopy declined. Nevertheless, water stress was not observed as a problem for the forest growth.

1. Introduction

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

[2] The boreal forest (called taiga in Eurasia), the largest terrestrial biome in the world, comprises 11% of the Earth's vegetation (about 16 million km2) and 26% of the total global vegetation carbon stocks [Larsen, 1980; Bonan and Shugart, 1989; Dixon et al., 1994; Prentice et al., 2001]. It is located in the circumpolar region spanning most of interior Alaska, Canada, Scandinavia and Finland, Scotland, Russia, Kasachstan, northern Mongolia, and northeastern China [Bonan and Shugart, 1989]. Owing to its large distribution area, it plays a critical role in balancing the global carbon budget [Goulden et al., 1997; Myneni et al., 1997; Field et al., 1998; Houghton et al., 1998; Malhi et al., 1999; Schulze et al., 1999; Ciais et al., 2000; Geider et al., 2001] and mitigating global climate [Foley et al., 1994; Bonan et al., 1995; Dixon et al., 1996; Beerling, 1999; Kasischke and Stocks, 2000; Foley et al., 2003]. The boreal forest is also one of the most intensively studied ecosystem in the flux measurement community, for instance, in the Boreal Ecosystem-Atmosphere Study (BOREAS) [Sellers et al., 1997; Hall, 1999, 2001], the Northern Hemisphere Climate-Processes Land-Surface Experiment (NOPEX) [Halldin et al., 1999], and the EuroSiberian Carbonflux project [Schulze et al., 2002]. The boreal forests can act as either sources or sinks for atmospheric carbon dioxide [Malhi et al., 1999; Chapin et al., 2000; Valentini et al., 2000]. The source or sink strength of boreal forests shows large interannual variability with respect to climate [Goulden et al., 1998; Lindroth et al., 1998; Schulze et al., 1999; Baldocchi et al., 2001; Law et al., 2002], and large differences among different forest types [Apps et al., 1993; Bonan, 1993; Gower et al., 2001]. Summer drought or water stress may switch a boreal forest from a sink to a source of CO2 [Price and Black, 1990; Price and Black, 1991; Hollinger et al., 1998]. Increasing temperatures may reduce the sink strength of the boreal forests [Plochl and Cramer, 1995; Lindroth et al., 1998]. Source strength (soil respiration) of some boreal forests was shown to be a function of photosynthetic assimilation [Högberg et al., 2001]. Disturbances, for example, fire and logging, may have a considerable influence on the CO2 uptake ability of some boreal forests [Kurz and Apps, 1999; Schulze et al., 1999; Amiro, 2001; Law et al., 2002; Thornton et al., 2002; Hicke et al., 2003]. There are also many other long-term factors that must be kept in mind in quantifying the source/sink strength of the boreal forests, for example, the increase in atmospheric CO2 concentration and air temperature [Melillo et al., 1993; Dixon et al., 1994; Gifford, 1994; Amthor, 1995; Goulden et al., 1998; Chen et al., 2000; White et al., 2000; Hicke et al., 2002], nitrogen deposition [Cao and Woodward, 1998; Schulze et al., 1999; Chen et al., 2000], and the length of the growing season [Myneni et al., 1997; Goulden et al., 1998; Randerson et al., 1999; Schulze et al., 1999; Black et al., 2000; Baldocchi et al., 2001]. Clearly, until now we have accumulated much knowledge on the CO2 exchange between the boreal forests and the atmosphere in both the western and eastern hemispheres, but there is a paucity of information on NEE of the southern limit of the Siberian boreal forest, which is adjacent to the vast area of Asian steppe biome. Our study on the NEE of a marginal area of the boreal forest should have implications for quantifying the role of the entire boreal ecosystem in the global carbon cycling and its sensitivity to climate.

[3] The boreal forest in Mongolia covers about 5% of the Mongolian land area, is distributed primarily in the mountain areas such as the Khentii Mountains and the Khangai Mountains of northern Mongolia, and is dominated by Siberian larch (Larix sibirica Ledeb.) which covers more than 70% of the forested areas [Hilbig, 1995; Gunin et al., 1999]. The cold continental climate in these areas is characterized by long, cold and dry winters versus short, cool and moist summers. Low temperatures and a short growing season are the major limiting factors for plant growth [Hilbig, 1995; Gunin et al., 1999]. Although most of the Mongolian boreal forests remain pristine, the area that is adjacent to the steppe (the forest-steppe transition zone) has become seriously threatened over the past 2 decades owing to growing commercial exploitation of the forest resources, mainly by logging and mining [Erdenesaikhan and Erdenetuya, 1999; Goldammer, 2002]. Fire hazard, pest outbreak, and unsustainable excessive logging are major disturbances impacting the dynamics of Mongolian boreal forests in general [Intergovernmental Panel on Climate Change (IPCC), 1997; Goldammer, 2002]. The forest-steppe transition zone is ecologically fragile and vulnerable to environmental changes and human activities. A raising concern is thus the possible influences from the projected global warming, namely the shrinking extent of the boreal forest and the recession of the treeline to higher altitudes in Mongolia [Goldammer, 2002]. An analysis of instrumental meteorological data by Dagvadorj and Mijiddorj [1996] shows that since the 1940s, annual temperatures have increased by 1.8°C in western, 1.0°C in central, and 0.3°C in eastern Mongolia. In addition, their results also indicate that concomitantly the growing season length has increased by about 10 to 20 days, and summer precipitation, especially in August, has increased [Dagvadorj and Mijiddorj, 1996]. Recent dendrochronological research provides strong supporting evidence for these trends in climate change [Jacoby et al., 1996, 1999; D'Arrigo et al., 2000; Pederson et al., 2001]. Rising temperatures may favor carbon uptake by the Mongolian boreal forests by increasing the growing season length, but at the same time this may also lower soil moisture via increased evapotranspiration, and increase additional release of conserved plant and soil carbon via respiration. Sustained drying induced by such a temperature increase may also increase the risk of forest fires and pest outbreaks and thereby further accelerate forest area losses [IPCC, 1997]. Furthermore, the Mongolian boreal forest plays an important role in biodiversity conservation, carbon and water cycling, regional climate regulation, and sustainable social-economic development in Mongolia [Batjargal and Enkhbat, 1998; Bastian, 2000]. Therefore it is necessary to obtain a better scientific understanding of the exchange mechanisms of water, energy and CO2 over the Mongolian boreal forest and their responses to climate change and anthropogenic disturbances. The present study's goal is to close this gap in current knowledge. It was part of a 5-year international project (2002–2006), the Rangelands Atmosphere-hydrosphere-biosphere Interaction Study Experiment in Northeastern Asia (RAISE), which is currently in progress in the region along the Kherlen River, the second longest river in Mongolia. This paper reports one-full-year-long NEE measurements (from 25 March 2003 to 24 March 2004) over a montane larch forest, located in the upper reach of the Kherlen River, by means of the eddy covariance (EC) technique. The major issues covered here include (1) the dynamics of NEE across daily and seasonal scales and (2) the environmental controls on NEE.

2. Materials and Methods

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

2.1. Site Information

[4] The measurement site was located about 25 km northeast of the Mongonmorit village in the Tov province of Mongolia (48°21.112′N, 108°39.260′E, 1630 m a.s.l.) (Figure 1). The EC flux tower was erected in March 2003 in a Siberian larch (Larix sibirica Ledeb.) forest on a southwest-facing gently sloping hill of the Khentii Mountains. The site is relatively uniform, and the maximum slope at the site was about 5° to the west. The forest extended about 750 m to the south and extended over 800 m in other directions. Prevailing winds were from the northwest.

image

Figure 1. (a) Location of the eddy flux measurement site in a montane larch forest near Mongonmorit, the Tov Province of Mongolia, (b) located on a gentle slope in rolling terrain. Air photographs show (c) the tower site and (d) the character of the dense understory with shrubby birch in forest gaps after fire disturbance. The circle in Figure 1b shows the location of the eddy covariance tower.

Download figure to PowerPoint

[5] The cold continental climate is characterized by long and cold winters, short summers which provide the greatest fraction of annual precipitation, and a wide range of daily and seasonal temperature fluctuations. The meteorological records for the 10-year period between 1993 and 2002 at the Mongonmorit Weather Station, located in the foothills of the Khentii mountains (the nearest station to the measurement site) show a mean annual air temperature of the region of −2.7°C and a mean annual precipitation of 296 mm, with the bulk part falling from June through September (267 mm, 90%). January is coldest with mean daily temperatures below −23.8°C, and July is warmest with a mean air temperature exceeding 15.6°C. There are only 5 months (May to September) with mean monthly air temperatures exceeding 5°C. The number of days with mean air temperature above 5°C is 138 per year. The growing season lasts about 5 months (May to September) and is in phase with the period of abundant precipitation.

[6] The soil at the site is a spodosol (seasonal cryosol). It has a coarse texture, and is highly leached and gray in color. No calcium was found in the top 100 cm near the EC tower. Total nitrogen and carbon content of the soil over a depth of 40 cm after removal of roots by sieving averaged 0.50 ± 0.23% and 2.21 ± 0.50%, respectively (n = 3 replications, S.-G. Li et al., unpublished data).

[7] The forest was dominated by Siberian larch, which mixed with scattered or patchy white birch (Betula platyphylla Sukach.) in some places. Characteristics of the stand were determined by detailed investigation of a representative 0.26-ha area in the immediate vicinity of the tower (four 2 × 100 m plots and one 40 × 45 m plot). The mean stand height was about 20 m. The tree stem density was about 1120 ha−1, with the breast height diameter varying from 2.4 to 54.9 cm (mean 15.0 ± 8.6 cm). The projected one-sided leaf area index (LAI) of the canopy was 2.7 estimated by measuring areas of leaves collected with 23 litter traps (0.8 m in diameter and 1 m in height), which were put in place in August 2003 and collected in March 2004. This value represents the maximum LAI at the site in 2003. Conversely, the corresponding LAI estimated with the method proposed by Frazer et al. [2001] using a digital camera (Nikon Coolpix 4500, Nikon Corporation Imaging Company, Tokyo, Japan) was 2.2. This estimate was based on the 30 photos taken on 23 July 2003 (the peak growing period) and 3 March 2004 (the leafless period) at 15 fixed points located at the above 40 × 45 m plot. Annual litterfall biomass, which was measured with above littertraps, was 117 g m−2 (dry weight) per unit ground area. Disturbances to the forest stem mainly from fires, lumbering, and pest [Hilbig, 1995; Gunin et al., 1999]. The forest experienced large-scale fires in 1996 and 1997 [Valendik et al., 1998; Chuluunbaatar, 2002]. Fire scars on the tree trunks were found in many places in the stand. The stem density of dead trees due to the fire was 420 ha−1. Forest fires are closely related to spring and autumn drought in northern Mongolia [Erdenesaikhan and Erdenetuya, 1999; Goldammer, 2002]. Lumbering activities have increased drastically over the past 3 decades [Chuluunbaatar, 2002]. The tree stumps left after lumbering were 570 ha−1 in the area close to the tower. The Siberian silk moth (Dendrolimus sibiricus Tschetverikov) feeds on the leaves of trees growing in the foothills of the mountains, especially during the dry season, when insect pest outbreaks are frequent. An insect outbreak occurred in July 2003, leading to scattered leaf damage on most trees in the transitional area between forest and adjacent forest steppe. As a result of frequent fires and lumbering activities, the age structure of the larch forest spans the range from ∼70 to over 150 years old. The oldest trees are older than 300 years. The root system of larch trees is concentrated in the upper 30-cm soil layer, but fine roots were found down to a depth of 100 cm. The understory was dense and formed a distinct layer of grasses and scattered shrubs. The grasses were dominated by sedges (Carex spp.), Junegrass (Koeleria spp.), and fireweed (Chamaenerion angustifolium L.). The dominant shrub was shrubby cinquefoil (Potentilla fruticosa L.). In July 2003, when the understory plants reached their peak growth period, their leaf area index was measured by the clipping method. Quadrat size was 0.25 m2 with three replications. Live leaves were removed from the stems to determine green LAI by a scanner (CanoScan LiDE 40, Canon, Tokyo, Japan) and corresponding special software (LIA32, K. Yamamoto, Laboratory of Forest Environment and Resources, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya City, Japan). The maximum green LAI for the understory vegetation was 1.7. Similarly, measured understory green LAI was 1.5 on 11 August and 0.1 on 1 October 2003. Therefore the probable maximum LAI of the whole canopy ranged between 3.9 and 4.4, with a best estimate of 4.0. We found dense stands of white birch in two large open clear-cuts (one due to lumbering, about 540 m to the northeast of the EC tower; the other due to fire, about 400 m to the west of the tower).

[8] We did not have continuous measurements of LAI at the sites; thus we used the normalized difference vegetation index (NDVI) as a surrogate of LAI to indicate the vegetation development [Reed et al., 1994]. NDVI (until end of October) was derived by CeRES of the Chiba University of Japan from the sequence of 10-day composite VEGETATION/SPOT (VGT) images taken by the VITO center, Belgium, for a 1 × 1 km quadrant centered at the flux measurement site. Each 10-day VGT image contains clear-sky pixel observations corrected for atmospheric effects. NDVI data break between 10-day intervals was linearly interpolated.

2.2. Flux Measurements

[9] Eddy covariance measurements of carbon dioxide, water vapor, sensible heat, and momentum were carried out at a height of 30 m above the ground, about 10 m above the average canopy height over the forest, on a scaffolding tower. The tower was constructed in October 2002 and protected against entrance of grazing animals with a fence (1.5 m in height and 5 × 5 m in wideness). The tower was instrumented in late March 2003. The instrumentation included a 3-D ultrasonic anemometer–thermometer (SAT-550, Kaijo Sonic Co., Tokyo, Japan, 15-cm path length) and an open path infrared gas (CO2/H2O) analyzer (IRGA) (Li7500, LICOR Inc., Lincoln, Nebraska, 15-cm path length). The former measured instantaneous fluctuations of the vertical (w), horizontal (u), and lateral (v) winds and the virtual temperature (Tv), and the latter measured instantaneous fluctuations of water vapor (q) and carbon dioxide (c) concentrations. The separation between the midpoints of these two neighboring sensors was about 20 cm to minimize underestimation of fluxes [Lee and Black, 1994]. The two sensors were mounted on a 1-m-long boom placed on the top of the tower. The tower was also equipped with instruments to measure net radiation (CNR1, Kipp & Zonen BV, Delft, Netherlands) at 29 m, and air temperature and humidity (HMP-45D, Vaisala Inc., Helsinki, Finland) at 30 m above the ground. The air temperature/humidity sensor was shielded and aspirated. Output signals of wind speed and gas concentration were sampled at 10 Hz while output signals of radiation, temperature, and humidity were scanned at 0.2 Hz. Following the procedure in the LI-COR Instruction Manual [Li-COR Inc., 2000], we calibrated the IRGA twice in Japan against standard gases (0 ppmv CO2/N2 gas, and 303.3, 607.7, and 1000 ppmv standard CO2 gases) for CO2 signals and with a portable dew point generator (Li-610, Li-COR Inc., Lincoln, Nebraska) for water vapor signals. The calibration of zeros and spans for the IRGA showed negligible drift during the measurement period. The half-hourly mean scalar fluxes were computed online and recorded continuously by a CR23X data logger (Campbell Scientific, Logan, Utah). This data logger also recorded half-hourly means of air temperature/humidity, four components of radiation (up and down long and short wave radiation) and air pressure.

[10] We also measured the soil temperature profile at depths of 5, 10, 20, 30, 50, 70, and 100 cm with platinum resistance thermometers (C-PT, CLIMATEC Inc., Tokyo, Japan), the soil heat flux at depths of 2 and 10 cm by soil heat plates (PHF-1.1, REBS, Inc., Seattle, Washington), soil moisture profile at depths of 10, 20, 30, 70, and 100 cm by time domain reflectometry probes (CS616, Campbell Scientific, Logan, Utah), and precipitation within the canopy by a tipping bucket rain gauge (CYG-52202, RM Young Company, Traverse City, Michigan). The underground sensors were placed in late October 2002. They were sampled at 0.1 Hz, and the 30-min mean data or sum (for precipitation) were logged on a CR10X data logger (Campbell Scientific, Logan, Utah). The measurements were carried out from 25 March 2003 to 24 March 2004. Measurements continue as we write this paper.

2.3. Data Processing

[11] Net ecosystem CO2 exchange (NEE) between the forest and the atmosphere can be computed as [Baldocchi, 2003]

  • equation image

where the first term on the right-hand side is the covariance between vertical velocity fluctuations (w′) and CO2 concentration fluctuations (c′), the second term denotes the change in storage of CO2 (Ca) within the canopy, z is height above ground surface, h is the canopy height, t is time, and overbars denote a time average (30 min in this study). Negative NEE denotes a carbon flux into the ecosystem and was expressed per unit ground area, whereas positive NEE denotes the reverse. Prior to the scalar flux computation, coordination rotation was performed to force the mean vertical velocity to zero [McMillen, 1988]. Spectral and cospectral analyses were conducted periodically during intensive observation periods to check the performance of eddy flux sensors. Data post-processing included (1) the cospectral correction for the carbon dioxide and water vapor fluxes using the algorithm proposed by Eugster and Senn [1995] and (2) the correction of the scalar fluxes for the density flux effect following the algorithm described by Webb et al. [1980] and Leuning and Moncrieff [1990]. We did not measure the CO2 concentration profile and thus used the method by Hollinger et al. [1994] to estimate the change in the CO2 storage term in equation (1).

[12] Closure of the surface energy budget is often considered as a test of the overall performance of the flux measurement system [Wilson et al., 2002]. We assessed the closure via linear regression between the sum of latent heat (LE), sensible heat (H) fluxes, and the difference between net radiation (Rn) and soil heat flux (G) [Wilson et al., 2002],

  • equation image

where a1 and a2 are the intercept and the slope, respectively. Since we did not measure the storage term, we used the cumulative values of 24-hour periods here to minimize the possible effect of the energy storage upon the closure. However, the use of daily integrals may mask possible biases [Mahrt, 1998]. During the period between DOY 84 and 280, the flux measurements presented a good closure, with a1, a2, and the coefficient of determination (r2) being 5.528 W m−2, 0.750, and 0.74, respectively. Examination of the ratio of (LE + H) to (RnG) shows a slight decline (about 1% every 10 days) with time in the energy closure during this period. This is likely to be associated with the increase in the energy storage with the canopy evolution, which we neglected, and might also be due to deterioration (dirt or dust building up) of the LI-7500 sensor since we cleaned it only a few times every month. For the most of the growing season (from May 1 to September 30), a1, a2, and r2 were 0.178 W m−2, 0.781, and 0.733, respectively. Outside the growing season, relative energy budget closure was much more variable with more deviation from ideal conditions. During the entire measurement period from 25 March 2003 to 24 March 2004 (366 days), a1, a2, and r2 were 20.674 W m−2, 0.613, and 0.823, respectively. The ratio of (LE + H)/(RnG) was 0.783. In the annual total, (LE + H) accounted for about 91% of (RnG).

[13] We removed anomalous or spurious data points that were caused by sensor malfunction, rain events, sensor maintenance, IRGA calibration, power failure, etc. Roughly 48% of the data obtained from our EC system were discarded, which produced gaps in our data collection. The gap fraction was larger (>66%) in the winter season (from November 2003 to March 2004) and relatively small at other times (<35%). The gaps in the growing season were mainly due to rain events. Interpolating with fair-weather data may overestimate net CO2 uptake by the forest. For calculation of daily and annual sums of the fluxes, these gaps were filled following the strategies suggested by Falge et al. [2001]: (1) Linear interpolation was used to fill the gaps that were less than 2 hours by calculating an average of the values immediately before and after the data gap; (2) other data gaps were filled using the empirical relationships (look-up tables), such as the regressions of light versus daytime NEE (equation (3)) and temperature versus nighttime respiration with u*-correction (u*, friction velocity, ≥0.5 m s−1) (equation (4)) in cases where those relationships could be established; and (3) if these relationships could not be established owing to missing meteorological data, we used mean daily variations to fill the gaps. We used a relatively high threshold of u* because the sensitivity analysis showed that the annual NEE total leveled off when u* exceeded roughly 0.4 to 0.5 m s−1. Gaps in the precipitation data in the winter season (from October 2003 to March 2004) when the rain gauge failed were filled with data from the Mongonmorit Weather Station. Four bimonthly look-up tables were created for the period from March to October 2003, and one look-up table covered the rest of the measurement period.

[14] Incident photosynthetically active radiation (PAR) was not measured directly and was estimated from measured short-wave solar radiation (Rs) using the empirical relationship: PAR (μmol photons m−2 s−1) = 2.16 × Rs (W m−2) [Weiss and Norman, 1985].

[15] NEE can be described by a Michaelis-Menten type rectangular hyperbola [Ruimy et al., 1995; Falge et al., 2001],

  • equation image

where α is the apparent quantum yield or the initial slope of the light response curve (μmol CO2 per μmol of photons), and NEEsat is the saturation value of NEE at an infinite light level, and R is the hypothetical bulk ecosystem respiration. During the night, when turbulence was weak (u* < 0.5 m s−1), NEE was corrected using the nighttime NEE data obtained when u* ≥ 0.5 m s−1 using an exponential function [Lloyd and Taylor, 1994],

  • equation image

where NEEnight,u* is the nighttime NEE with u* correction, b1 and b2 are regression coefficients and Ta is air temperature in °C measured at 30 m above the ground.

3. Results and Discussion

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

3.1. Characteristics of Environmental Factors

[16] Time series of major environmental driving variables over the measurement year (25 March of 2003 to 24 March of 2004) are presented in Figure 2. These variables included daily integrated global solar radiation sum (Rs), daily integrated net radiation (Rn), daily averaged air temperature (Ta) and volumetric soil water content (SWC), daily precipitation sum (PPT), and the normalized difference vegetation index (NDVI). Conditions during our field experiment are compared with the mean annual NDVI cycle averaged over the period 1981–2002.

image

Figure 2. Seasonal variations in environmental variables. (a) Daily integrated global solar radiation sum (Rs) and daily integrated net radiation (Rn). (b) Daily averaged air temperature (Ta) and water vapor pressure deficit (VPD). (c) Volumetric soil water content (SWC) and daily precipitation sum (PPT). (d) Normalized difference vegetation index (NDVI) in 2003 (solid symbols) and average conditions 1981–2002 (open symbols). By courtesy of the VEGETATION program, VITO, Belgium, NDVI was computed by CeRES of the Chiba University of Japan from SPOT VGT images distributed by VITO. The vertical error bars in Figure 2d represent 1 standard deviation.

Download figure to PowerPoint

[17] The maximal and minimal Rs were 31.3 MJ m−2 d−1 (DOY 149) and 2.3 MJ m−2 d−1 (DOY 272), respectively. During winter and spring, most days were clear. In contrast, in the summer, owing to cloudiness and rain events, Rs fluctuated stringently and sometimes dropped to very low levels (Figure 2a). The maximal and minimal Rn was 18.5 MJ m−2 d−1 (DOY 172 of 2003) and −3.9 MJ m−2 d−1 (DOY 8 of 2004), respectively. The variation in Rn paralleled that of Rs (Figure 2a). Daily totals below 1 MJ m−2 d−1 were observed from DOY 309 of 2003 to DOY 41 of 2004. In contrast to summer, where the expected 1.5 months time lag [e.g., Eugster et al., 1997] between Rn and Ta is clearly seen, the winter minimum Ta was observed well before the minimum Rn. This suggests that winter Rn is strongly driven by temperature (and not the other way round) in this continental climate. On the annual basis from DOY 84 of 2003 to DOY 84 of 2004, the forest received 5200 MJ m−2 of short-wave solar energy. The annual total of Rn was about 1930 MJ m−2, which accounted for about 37% of Rs. The overall mean Ta at the site was −1.9°C. The maximal Ta was 19.0°C on 12 July (DOY 193), and the minimal Ta was −26.3°C on 2 December (DOY 336) (Figure 2b). VPD was rather low and even during fair weather days; the maximum daily mean VPD was below 1.5 kPa (Figure 2b). There were only 13 days with a maximum hourly VPD > 2 kPa. In winter, VPD was generally less than 0.5 kPa.

[18] Total precipitation from April to September of 2003 at the flux tower site was 273 mm (Figure 2c), which was about 9.6% more than the value of 249 mm for the same period at the Mongonmorit Weather Station. Owing to instrument failure, precipitation data were unavailable for the rest of the measurement period (from October 2003 to March 2004) and thus filled with the precipitation measured at the Mongonmorit Weather Station. The weather station recorded 31 mm precipitation during this period, which was about 55% more than the normal (averaged at 20 mm over 1993 to 2002) for the same period. SWC at 10 cm depth was lowest during the non-growing season (the minimum was 2.4% by volume) and highest during the rain-concentrated summer (with a maximum of 18.6%) (Figure 2c). Fluctuations in SWC were closely related to precipitation. Several summer rain spells gave rise to the increase in SWC. Declining SWC during the summer season was primarily due to canopy transpiration.

[19] The NDVI reveals the relative greenness of vegetation and thus is often used to compare seasonal changes in vegetation growth and activity. The NDVI values were below 0.25 during the leaf-out period, and began to increase from the beginning of the growing period, and reached the maximum at the peak of the growing period, with values around 0.7 in early July. Relatively lower NDVI values (e.g., DOY 121, 131, and 264) than the normal average were due to cloudiness. NDVI rapidly decreased with the onset of senescence (Figure 2d). Comparison with the long-term average 1981–2002 shows that the site had normal vegetation conditions during most of year 2003 (Figure 2d).

3.2. Diurnal Variation in NEE

[20] We used the temperature criteria, i.e., a Ta at 30 m (5°C) and a soil surface temperature at 5 cm (Ts) (0°C) to delimit the range of growing season/leaf-on period and leaf-out period. If daily mean Ta (daily mean Ts) was larger (in spring) or less (in fall) than 5°C (0°C) for three consecutive days, and thereafter there were no consecutive days on which daily mean Ta (daily mean Ts) was less (in spring) or larger (in fall) than 5°C (0°C), these temperature limits were defined as the beginning or the end of the growing season for our forest. In 2003, the dates for these temperature limits were 28–30 April for Ta (26–28 April for Ts) in spring, and 30 September to 2 October for Ta (8–10 October for Ts), respectively. Thus the growing period roughly covered the period from 1 May to 30 September and the leaf-out period covered the remaining period (from 25 March to 30 April 2003 and from 1 October 2003 to 24 March 2004). In order to describe diurnal means, we averaged NEE over the entire period 1, and in monthly intervals during period 2. The diurnal courses of NEE, together with PAR and Ta, are plotted in Figure 3. Before and after the full-leaf period, NEE varied little, ranging from 0.1 to 0.5 μmol CO2 m−2 s−1 (Figure 3a), in accordance with Ta variations from −8.3 to −13.7°C (Figure 3c). During this period, the mean Ta was −11.2°C, and on average the forest received PAR of 24.3 mol photons m−2 d−1 and released around 0.36 g C m−2 d−1 (Figure 3b). During the leaf-on period (from May to September), there was much seasonal variation in the magnitude of the average diurnal courses of NEE. NEE reached its daily maximum roughly 2 hours earlier than PAR did. Generally, NEE was stronger in the morning than in the afternoon, reflecting the variation of the balance between the photosynthesis and respiration of the canopy; that is, the light and temperature conditions in the morning were relatively favorable to photosynthesis whereas higher air temperature in the afternoon enhanced respiration (Figure 3). Nocturnal NEE in contrast showed very little short-term variation (Figure 3b). NEE was lower in May when the forest only acted as a weak carbon sink in the early morning and remained a carbon source during the rest of the day. In May the forest was still a carbon source and lost about 0.75 g C m−2 d−1 on average. In the following 4 months, the forest was a clear carbon sink (net CO2 gain from the atmosphere) with maximum uptake rates in July (Figure 3a, Table 1). The increase in sink strength from May to July was in phase with both the increase of NDVI (from 0.36 to 0.63, Figure 1d) and Ta (from 6.5 to 15.0°C, Table 1). The decrease in the sink strength thereafter was accompanied by a gradual decrease in both NDVI and Ta.

image

Figure 3. Average daily courses of (a) net ecosystem CO2 exchange (NEE), (b) incident photosynthetically active radiation (PAR), and (c) air temperature (Ta) measured at a height of 30 m above the ground over a montane larch forest in Mongolia. Solid line with open triangles, the leaf-out, senescence and winter period (averaged over the period of 25 March to 30 April 2003 and 1 October 2003 to 24 March 2004); solid line, May; dotted line, June; dotted line with solid circles, July; dashed line, August; broken line, September. Monthly curves in Figure 3a are also labeled with the number of the month (e.g., 5 denotes May). Time of day is Mongolian summer time (MST).

Download figure to PowerPoint

Table 1. Monthly Averages of Daily Integrated NEE, Daily Integrated PAR, Bulk Light Use Efficiency (LUE = GEP/PAR), and Ta During the Growing Season for a Mongolian Montane Larch Foresta
MonthNEE, g C m−2 day−1PAR, mol photons m−2 day−1LUE, mmol CO2 mol−1 photonsTa, °CPPT, mm
  • a

    Monthly sum of precipitation (PPT) is also presented.

May0.75 ± 0.3140.5 ± 16.0−2.1 ± 1.16.9 ± 4.036.5
June−1.62 ± 1.0945.7 ± 13.5−8.5 ± 3.913.7 ± 3.023.0
July−2.28 ± 0.9240.4 ± 14.2−10.9 ± 3.514.9 ± 2.4116.0
August−1.93 ± 0.7940.5 ± 11.4−9.8 ± 3.410.9 ± 2.268.5
September−0.18 ± 0.8031.2 ± 10.6−4.9 ± 3.07.2 ± 2.927.0

[21] The two extremes in the monthy mean diurnal cycles of NEE (Figure 3a) were both found in July with a 30-min mean minimum (maximum net carbon gain) of −10.1 μmol m−2 s−1 and a corresponding mean maximum (maximum net carbon loss) of 2.5 μmol m−2 s−1. The absolute maximum NEE at 30-min resolution was −20.4 μmol m−2 s−1 (DOY 196).

3.3. Seasonal Course of NEE

[22] Figure 4 shows the seasonal course and cumulative values of NEE. A large day-to-day variability can be observed. This variability correlates with the variation in daily weather conditions, namely with cloudiness and day-to-day differences in available short-wave energy gain (Figure 2a), temperature and vapor pressure deficits (Figure 2b), and water availability (Figure 2c). In the leaf-out period, the larch forest was a weak source of CO2 due to absence of strong photosynthetic activity. In most of the full-leaf period, the forest acted as a sink for CO2, and the sink strength was largely depending on the short-term fluctuations of photosynthesic uptake. Before June, the forest sometimes was a weak sink for CO2, but was more often a source of CO2 because photosynthesis was weak as a consequence of the low green leaf area (low NDVI) and respiration was the major contributor to NEE. Toward the beginning of June (DOY 152), carbon uptake increased precipitously with time until the maximum values were reached in early July. Sink strength was strongest in July, corresponding to the peak of the seasonal growth, represented by highest NDVI (Figure 2d). The CO2 uptake declined rather rapidly from late August through late September as a result of senescence and temperature drop. The forest shifted from its role as a growing-season sink toward a net source of CO2 by the end of September. It continued as a carbon source throughout winter although the magnitude was very low.

image

Figure 4. Temporal sequences of daily net ecosystem carbon exchange (NEE) and cumulative NEE. Data were taken from the measurements from 25 March 2003 to 24 March 2004 over a montane larch forest in Mongolia.

Download figure to PowerPoint

[23] The magnitude of daily integrated NEE ranged between −4.0 (DOY 190) and 1.6 g C m−2 d−1 (DOY 140) and averaged −0.23 g C m−2 d−1 over the annual course, which sums to −85 g C m−2 yr−1. Thus this Mongolian larch forest was a net sink for CO2 during our 1-year measurement period.

3.4. Responses of NEE to Environmental Conditions

[24] Responses of NEE to various environmental factors including NDVI, PAR, Ta, VPD, and SWC were examined primarily for the period from 1 May to 30 September 2003 when the forest canopy was most active in growth.

3.4.1. Response of NEE to NDVI

[25] Since NDVI is generally used as an index to represent the greenness of the canopy, it can be regarded as a surrogate for leaf area index in describing vegetation growth conditions [Reed et al., 1994]. In accordance with 10-day averaged NDVI values, NEE data were also block averaged over 10-day periods and then plotted against NDVI (Figure 5). Daily integrated NEE was found to correlate with NDVI in a linear manner over the period from May to September: NEE = (2.45 ± 0.23) − (6.68 ± 0.42) × NDVI (n = 153, adjusted r2 = 0.620, F = 249, P < 0.001). Sixty-two percent of the variance in NEE was associated with the variation in NDVI over the growing season. In addition, for the daytime NEE: NEE = (2.13 ± 0.22) − (7.53 ± 0.41) × NDVI (n = 153, adjusted r2 = 0.682, F = 326, P < 0.001).

image

Figure 5. Relationship between daily integrative net ecosystem carbon exchange (NEE) and normalized difference vegetation index (NDVI). NDVI values between the 10-day intervals were linearly interpolated. Solid triangles with standard error bars are bin averaged (NDVI bin width is 0.05). A linear regression was fitted to the data: NEE = (2.45 ± 0.23) − (6.68 ± 0.42) × NDVI (n = 153, adjusted r2 = 0.620, F = 249, P < 0.001).

Download figure to PowerPoint

3.4.2. Response of NEE to PAR

[26] The daytime 30-min mean NEE was plotted against incident PAR (Figure 6a). We used equation (3) to describe the relationship between PAR and NEE. The apparent quantum yield (α) was −0.0133 ± 0.0011 μmol CO2 per μmol of photons, the saturation value of NEE (NEEsat) was −10.1 ± 0.4 μmol m−2 s−1, and the bulk daytime ecosystem respiration (R) was 1.3 ± 0.1 μmol m−2 s−1. The coefficient of determination (r2) was 0.337. We also plotted the daily integrated NEE (in g C m−2 d−1) against incident PAR (in mol photons m−2 d−1) (Figure 6b). We also used equation (3) to build the NEE-PAR response relationship on the daily basis: α = −0.5 ± 1.1 g C mol−1 photons or −0.0418 ± 0.0947 mol CO2 mol−1 photons, NEEsat = −4.9 ± 3.2 g C m−2 d−1, R = 2.8 ± 3.8 g C m−2 d−1, n = 153, adjusted r2 = 0.09. The large scatter can partially be attributed to variations in other environmental factors such as VPD and Ta (see below). Carbon uptake (that is, NEE gets more negative) increased linearly with PAR at low to intermediate levels of PAR. At high PAR levels, NEE decreased slightly with increasing PAR and saturated up to PAR levels of around 50 mol photons m−2 d−1 (Figure 6b). Occurrence of light-saturation is due to low photosynthetic capacity of the canopy at the beginning of growth (e.g., in May), high levels of VPD (>1.0 kPa), and low soil water content (<10% for all the circled points in Figure 6b), or a combined effect of these factors (circled in Figure 6b). As compared with the daily NEE-NDVI relationship, the daily NEE-PAR relationship is rather weak. Additionally, the curvilinear shape of the NEE-PAR response curve for the larch forest is consistent with findings in the literature for forests [Wofsy et al., 1993; Baldocchi and Harley, 1995], but is different from the theoretical calculations by Leuning et al. [1995] and a literature review by Ruimy et al. [1995] in which the daily NEE-PAR response curve is linear. On the daily basis, the bulk light use efficiency (LUE) at the site was −6.7 mmol CO2 per mole of PAR photons over the entire growing season. The LUE peaked in July being −10.9 mmol CO2 per mole of PAR photons (Table 1).

image

Figure 6. Relationship between net ecosystem CO2 exchange (NEE) and incident photosynthetically active radiation (PAR) (a) on the daytime 30-min basis and (b) on the daily basis during the growing season from 1 May to 30 September 2003. Solid triangles with standard error bars are bin averaged by PAR intervals of (Figure 6a) 100 μmol photons m−2 s−1 and (Figure 6b) 5 mol photons m−2 d−1. The curves in both panels are best fits by equation (3) and described in detail in the text. Also see text for encircled points.

Download figure to PowerPoint

3.4.3. Response of NEE to Ta

[27] NEE was significantly related to Ta during the growing season (Figure 7). This relationship can be described by the linear regression: NEE = (0.841 ± 0.254) − (0.177 ± 0.022) × Ta (n = 153, adjusted r2 = 0.296, F = 65, P < 0.001). The linear increase of carbon uptake (NEE gets more negative) with increasing Ta suggests that lower Ta up to 15°–18°C (see Figure 7) is limiting growth whereas an exponentially increasing carbon loss appears to reduce NEE at temperatures >15°–18°C in this montane larch forest. Close examination of the data distribution shows that over 44% of NEE values were below −1.5 g C m−2 d−1, corresponding to a Ta range from 7° to 18°C (Figure 7), which therefore can be considered optimal for net photosynthesis. The bin-averaged NEE data (Figure 7) indicate that the forest switches from source to sink for CO2 around a threshold Ta of approximately 7°C.

image

Figure 7. Response of net ecosystem carbon exchange (NEE) to air temperature (Ta). Data were compiled from the growing period from 1 May to 31 September 2003. The line depicts the linear regression: NEE = (0.841 ± 0.254) − (0.177 ± 0.022) × Ta (n = 153, adjusted r2 = 0.296, F = 65, P < 0.001). Solid triangles with standard error bars are Ta bin averaged data. Bin size is 2°C.

Download figure to PowerPoint

3.4.4. Response of NEE to VPD

[28] The response of NEE to VPD is presented in Figure 8. A weak relation with a quadratic function can be seen. The best fits were: NEE = −(0.656 ± 0.150) − (5.283 ± 0.364) × VPD + (2.126 ± 0.184) × VPD2 (n = 4527, adjusted r2 = 0.053, F = 127, P < 0.001) for the 30-min mean data, and NEE = (0.278 ± 0.435) − (3.825 ± 1.303) × VPD + (2.170 ± 0.874) × VPD2 (n = 153, adjusted r2 = 0.051, F = 5.08, P < 0.05) for the daily sums. Although scatter is large, it is evident that NEE was reduced when VPD was high (over 1.8 kPa for 30-min mean NEE and 1.2 kPa for the daily NEE). The reduction at higher levels of VPD occurred on the days (circled in Figure 8) with high daily average Ta (>15°C), low soil water content (<10%), and relatively low NDVI (<0.4) (e.g., DOY 150–151, 163–165, and 174–175).

image

Figure 8. Response of net ecosystem CO2 exchange (NEE) to atmospheric water vapor pressure deficit (VPD) on (a) the daytime 30-min and (b) daily bases during the growing period from 1 May to 30 September 2003. Solid circles with standard error bars are VPD bin averaged data. Bin size is 0.25 kPa. The curves are the quadratic best fits: (Figure 8a) NEE = −(0.656 ± 0.150) − (5.283 ± 0.364) × VPD + (2.126 ± 0.184) × VPD2 (n = 4527, adjusted r2 = 0.053, F = 127, P < 0.001); (Figure 8b) NEE = (0.278 ± 0.435) − (3.825 ± 1.303) × VPD + (2.170 ± 0.874) × VPD2 (n = 153, adjusted r2 = 0.051, F = 5.08, P < 0.05). See text for encircled points.

Download figure to PowerPoint

3.4.5. Response of NEE to SWC

[29] Similarly, we used a quadratic function to describe the response of daily NEE to SWC as shown in Figure 9: NEE = −(4.682 ± 0.951) + (0.641 ± 0.190) × SWC − (0.025 ± 0.009) × SWC2 (n = 153, adjusted r2 = 0.10, F = 9.45, P < 0.005). It appeared that the NEE responded to SWC in a complex manner, as indicated by the large scatter in Figure 9. Apart from the encircled data points (circles I and II) in Figure 9, we found to our surprise that drought stress caused by low soil water content was not revealed to be an important factor during most of the growing season (Figure 9). Most likely, this is explained by the fact that the growing season was in phase with the rainy season, and that surface water shortage could be balanced by loading water from deeper soil layers. For circle I, which represents a 34-day period continuously from May 1 (DOY 121) through June 3 (DOY 154), although SWC was generally larger than 12% (ensemble daily averaged at 13.2 ± 0.9%), NEE was consistently positive varying from 0.1 to 1.6 g C m−2 d−1 with the ensemble average at 0.7 ± 0.3 g C m−2 d−1, indicating net carbon release from the canopy. During this period, ensemble daily averaged values were 0.31 ± 0.13 for NDVI, 41.4 ± 15.6 mol m−2 s−1 for PAR, 7.6 ± 4.5°C for Ta, and 0.6 ± 0.3 kPa for VPD, respectively. For the circle II, which represents an entire 17-day period from September 14 to 30 (DOY 257–273), ensemble daily averaged SWC was at 7.8 ± 1.3%, NEE varied from −0.4 to 1.4 g C m−2 d−1 with the ensemble average at 0.3 ± 0.5 g C m−2 d−1. During this period, ensemble daily averaged values were 0.36 ± 0.10 for NDVI, 27.5 ± 9.5 mol m−2 s−1 for PAR, 6.0 ± 2.7°C for Ta, and 0.4 ± 0.2 kPa for VPD, respectively. Clearly, for both periods, lower carbon uptake by the canopy (more positive NEE) was mainly due to lower photosynthetic capacity of the canopy, indicated by lower NDVI values. Lower Ta or lower PAR (circle II period) was also likely to play a role in this pattern.

image

Figure 9. Response of net ecosystem carbon exchange (NEE) to soil water content (SWC). Data were compiled from the growing period from 1 May to 31 September 2003. The curve is the quadratic best fit: NEE = −(4.682 ± 0.951) + (0.641 ± 0.190) × SWC − (0.025 ± 0.009) × SWC2 (n = 153, adjusted r2 = 0.10, F = 9.45, P < 0.005). Solid triangles with standard error bars are SWC bin averaged data. Bin size is 0.5%. See text for encircled points.

Download figure to PowerPoint

3.4.6. Combined Controls of Environmental Variables Over NEE

[30] The above analyses indicate that NEE responds differently to various environmental factors. The scatter in these response models (Figures 5 to 9) illustrates the complication of or the interaction among various environmental factors. Although each separate factor was important in governing NEE by itself, our concern here is which one was most advantageous to NEE. To answer this question, we computed the Pearson product-moment correlation coefficient between NEE and various environmental factors. For the 30-min mean NEE data (n = 153 × 48), the Pearson product-moment correlation coefficient was −0.704 between NEE and PAR, −0.340 between NEE and Ta, −0.252 between NEE and VPD, and 0.086 between NEE and SWC. This suggests that in the 30-min timescale the effect of PAR on NEE is dominant and NEE was least correlated with SWC. For the daily NEE (n = 153), the Pearson product-moment correlation coefficient was −0.789 between NEE and NDVI, −0.549 between NEE and Ta, −0.250 between NEE and PAR, −0.158 between NEE and VPD, and 0.252 between NEE and SWC. A stepwise multiple regression of NEE as a function of NDVI, Ta, PAR, and VPD for the active growth period displayed that 66% of the day-to-day variability in NEE was accounted for by the linear relationship: NEE = 1.897 ± 0.380 − (5.813 ± 0.715) × NDVI − (0.022 ± 0.007) × PAR − (0.030 ± 0.035) × Ta + (0.776 ± 0.473) × VPD + (0.079 ± 0.021) × SWC, (F = 59.4 and P < 0.001 for regression, P < 0.0001 for NDVI, P = 0.002 for PAR, P = 0.39 for Ta, P = 0.10 for VPD, and P = 0.0004 for SWC: adjusted r2 = 0.658, n = 153). Clearly, the effect of NDVI on NEE is dominating by far. Compared to the simple regression between NEE and NDVI (section 3.4.1) the additional gain in explained variance was marginal.

3.5. Comparison With NEE of Other Forest Ecosystems

[31] The peak value of mean hourly CO2 uptake rates observed during midsummer at our larch forest stand was about −10 μmol CO2 m−2 s−1. This is comparable with boreal evergreen coniferous forests [Baldocchi et al., 1997; Goulden et al., 1998; Vesala et al., 1998; Markkanen et al., 2001] (Table 2). It was greater than that reported for the Dahurian larch (Larix gmelinii (Rupr.) Rupr.) forest in eastern Siberia [Hollinger et al., 1998] (Table 2). The higher NEE rates observed at the Mongolian larch forest as compared to eastern Siberian larch forest are probably due to the larger LAI, which enhances radiation interception by the canopy for photosynthetic activity. The peak values of mean CO2 release rates (positive NEE) from our larch forest were low as compared with those reported for other forests with similar peak CO2 uptake rates (Table 2). However, this rate is close to the average observed for the eastern Siberian larch forest [Hollinger et al., 1998]. Similar small CO2 release rates were also reported as the lower end of the range found in a mountain Beech stand [Valentini et al., 1996] and a boreal jack pine forest [Baldocchi et al., 1997] (Table 2).

Table 2. Comparison of Net Ecosystem CO2 Exchange (NEE) for the Larch Forest in This Study With Other Selected Boreal Forest Ecosystems
SiteBoreal Forest TypeAge, yearsLAI, m2 m−2Mean Minimal NEE, μmol CO2 m−2 s−1Mean Maximal NEE, μmol CO2 m−2 s−1Annual NEE, g C m−2 yr−1Annual Reco, g C m−2 yr−1Annual GEP, g C m−2 yr−1Reference
Saskatchewan, Canadajack pine 2.8–3.3−82.3–4.5   Baldocchi et al. [1997]
Manitoba, Canadablack spruce53–1554.8−7.46.2−36∼20782∼861−864∼−815Goulden et al. [1998]
Yakutsuk, RussiaDahurian larch1252.4−5∼−88 (mean 2.4)   Hollinger et al. [1998]
Norunda, SwedenScots pine and Norway spruce50–1003–6−22.414.4−251∼−163596∼922−1173∼−759Lindroth et al. [1998]
Yenisei, RussiaScots pine2001.5−12.5 −163∼−145414∼464−628∼−559Lloyd et al. [2002]
Hyytiala, FinlandScots pine353−12.85.9−228720−948Markkanen et al. [2001], Vesala et al. [1998]
Northeastern MongoliaSiberian larch70–1503.9–4.4−10.12.5−85440−525this study

[32] The annual NEE rate of −85 g C m−2 yr−1 for our larch forest is lower than that observed in boreal evergreen Scots pine forest in central Siberia [Lloyd et al., 2002] (Table 2). The annual NEE rate at our site is however much larger than values reported from a boreal black spruce forest [Goulden et al., 1998]. Annual NEE rates vary widely among different biomes or even within the same biome, depending on differences in climate conditions, disturbance history, soil nutrition status, and architecture and eco-physiological functions of the canopy [Schulze et al., 1999; Law et al., 2002]. This variation is also partly associated with differences in ecosystem respiration [Law et al., 2002], suggesting that correct estimate of ecosystem respiration is critical in quantifying NEE. At our site, there existed a linear relationship between daily NEE and Ta, and lower air temperature is likely to be a limiting factor for carbon uptake (Figure 7). Annual NEE rate was sensitive to u* threshold, with which we corrected nighttime ecosystem respiration when air conditions were very stable with weak turbulence (Table 3). Roughly, annual NEE rate in absolute terms showed an decreasing tendency with an increase in u* before a threshold (0.5 m s−1 in this study) (Table 3). In addition, the long-term steady state CO2 budget expected in this larch forest is believed to be currently out of balance owing to substantial disturbances that have occurred over the past several decades. The stand is now recovering from fire and lumbering disturbance and represents an early successional stage of ecosystem development. Patchy appearance of young white birch community following fire and lumbering might also be an important contributor to the NEE dynamics.

Table 3. Sensibility Analysis of Annual Total of Net Ecosystem CO2 Exchange (NEE) and its Componentsa
u* Threshold, m s−1RecoNEEGEPReco/∣GEP∣
  • a

    Components are gross ecosystem production (GEP) and total ecosystem respiration (Reco) with respect to u*-correction. Reco, NEE, and GEP are given in g C m−2 yr−1. The ratio of Reco to absolute GEP is also presented.

No u* correction232−144−3760.62
>0.1244−145−3890.63
>0.2305−133−4380.70
>0.3380−116−4960.77
>0.4434−100−5340.81
>0.5440−85−5250.84
>0.6460−82−5420.85
>0.7433−81−5140.84

3.6. Partitioning of NEE Into its Components

[33] Gross ecosystem production (GEP) can be estimated from the difference between NEE and total ecosystem respiration (Reco) when NEE and Reco are quantified from the eddy covariance (EC) flux measurements. In this exercise, exact estimation of Reco is crucial. Since no photosynthesis is performed at night, the nighttime NEE represent ecosystem respiration only, which, however, might be low estimates at times with weak wind and stable stratification of the atmosphere, because an unknown fraction of CO2 flux (storage flux within the canopy) cannot be sampled by EC systems [Aubinet et al., 2000]. We excluded such critical conditions by rejecting nocturnal data obtained in weak turbulence (u* < 0.5 m s−1) and plotted the remaining data against soil temperature at 10 cm depth or air temperature at 30 m height. We used various exponential models [Lloyd and Taylor, 1994] to fit these data, and found that using equation (4) yielded the best fits to bimonthly, seasonal, and annual data sets. Thus we adopted equation (4) to fill nighttime NEE data gaps caused by low turbulence, in order to estimate daily Reco. On an hourly basis, the minimal Reco was estimated at 0.83 μmol CO2 m−2 s−1 at an air temperature of 0°C, 3.14 μmol CO2 m−2 s−1 at an air temperature of 20°C, and the Q10, the factor by which respiration increases for each 10°C increase in temperature, was estimated to be 1.61. On a daily basis, Reco was estimated at 0.99 g C m−2 day−1 at an air temperature of 0°C, and 3.31 g C m−2 d−1 at an air temperature of 20°C, and the Q10 was estimated to be 1.81. Integration of Reco over the daily and annual courses made it possible to also compute gross ecosystem production (GEP) from measured NEE as GEP = NEE − Reco. On an annual basis, GEP and Reco were −525 and 440 g C m−2 yr−1, respectively. The ratio of Reco to GEP was about 0.84. This ratio lies in the lower half of the range of 0.55–1.2 found in the literature [Law et al., 2002], and is slightly higher than the values observed in an undisturbed Larix gmelinii (Rupr.) Rupr. forest (0.63) in eastern Siberia [Kelliher et al., 1999] and in a 215-year-old stand of Pinus sylvestris L. (0.77) in central Siberia [Hollinger et al., 1998]. However, when the effect of different u* thresholds is considered (Table 3), it becomes clear that such comparisons strongly depend on assumptions made about data quality under nighttime low-turbulence conditions. In addition, because our site was located on a gentle mountain slope, divergence or convergence of the flux caused by the topographical features might also have a confounding, yet unquantifiable, advection effect [Moncrieff et al., 1996; Massman and Lee, 2002].

4. Conclusions

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

[34] The forests in the mountain areas of the Mongolian Plateau are important for the central Asian climate since they are located in a transitional area from the Siberian taiga to the Asian steppe and are expected to be vulnerable to climate change and anthropogenic disturbances. This study reports on the first CO2 flux measurements over a montane larch forest in northeastern Mongolia. One year-round observations of net ecosystem CO2 exchange (NEE) indicate that this forest was a net carbon sink for atmospheric CO2. The daily and seasonal patterns of NEE at this larch forest site are similar to those measured in other boreal forest types. The dynamics of NEE was dominantly related to growth status of the canopy as represented by the normalized difference vegetation index and controlled primarily by such environmental factors as air temperature, incident photosynthetically active radiation, and air water vapor pressure deficit.

Acknowledgments

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

[35] Financial support for this research was provided by Japan Science and Technology Corporation to M. Sugita. We acknowledge the assistance of Climatec, Inc., of Japan for installation of the eddy covariance measurement system and the Seemon Company of Mongolia for field assistance during field observation campaigns. We also thank D. Azzaya, T. Ganbold, and J. Bayasaa of Institute of Meteorology and Hydrology, Mongolia, for providing meteorological data of the study area and their valuable support in the field.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Amiro, B. D. (2001), Paired-tower measurements of carbon and energy fluxes following disturbance in the boreal forest, Global Change Biol., 7, 253268.
  • Amthor, J. S. (1995), Terrestrial higher-plant response to increasing atmospheric [CO2] in relation to the global carbon cycle, Global Change Biol., 1, 243274.
  • Apps, M. J., W. A. Kurz, R. J. Luxmoore, L. O. Nilsson, R. A. Sedjo, R. Schmidt, L. G. Simpson, and T. S. Vinson (1993), Boreal forests and tundra, Water Air Soil Pollut., 70, 3953.
  • Aubinet, M., et al. (2000), Estimates of the annual net carbon and water exchange of forests: The EUROFLUX methodology, Adv. Ecol. Res., 30, 113175.
  • Baldocchi, D. D. (2003), Assessing ecosystem carbon balance: Problems and prospects of the eddy covariance technique, Global Change Biol., 9, 478492.
  • Baldocchi, D. D., and P. C. Harley (1995), Scaling carbon dioxide and water vapor exchange from leaf to canopy in a deciduous forest: II. Model testing and application, Plant Cell Environ., 18, 11571173.
  • Baldocchi, D. D., C. A. Vogel, and B. Hall (1997), Seasonal variation of carbon dioxide exchange rates above and below a boreal jack pine forest, Agric. For. Meteorol., 83, 147170.
  • Baldocchi, D., E. Falge, and K. Wilson (2001), A spectral analysis of biosphere-atmosphere trace gas flux densities and meteorological variables across hour to multi-year time scales, Agric. For. Meteorol., 107, 127.
  • Bastian, O. (2000), Ecological situation and land use changes in Mongolia, in Consequences of Land Use Changes, Adv. Ecol. Sci., vol. 5, edited by Ü. Mander, and R. H. G. Jongman, pp. 199225, Witpress, Boston, Mass.
  • Batjargal, Z., and A. Enkhbat (1998), Biological diversity of Mongolia. First national report, report, Min. for Nature and Environ. of Mongolia, Ulaanbaatar.
  • Beerling, D. J. (1999), Long-term responses of boreal vegetation to global change: An experimental and modeling investigation, Global Change Biol., 5, 5574.
  • Black, T. A., W. J. Chen, A. G. Barr, M. A. Arain, Z. Chen, Z. Nesic, E. H. Hogg, H. H. Neumann, and P. C. Yang (2000), Increased carbon sequestration by a boreal deciduous forest in years with a warm spring, Geophys. Res. Lett., 27, 12711274.
  • Bonan, G. B. (1993), Physiological controls of the carbon balance of boreal ecosystems, Can. J. For. Res., 23, 14531471.
  • Bonan, G. B., and H. H. Shugart (1989), Environmental factors and ecological processes in boreal forests, Annu. Rev. Ecol. Syst., 20, 128.
  • Bonan, G. B., S. F. Chapin III, and S. L. Thompson (1995), Boreal forest and tundra ecosystems as components of the climate system, Clim. Change, 29, 145167.
  • Cao, M. K., and F. I. Woodward (1998), Dynamic responses of terrestrial ecosystem carbon cycling to global climate change, Nature, 393, 249252.
  • 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, Suppl., 211223.
  • Chen, J., W. J. Chen, J. Liu, J. Cihlar, and S. Gray (2000), Annual carbon balance of Canada's forest during 1895–1996, Global Biogeochem. Cycles, 14, 839849.
  • Chuluunbaatar, T. (2002), Forest fires in northern Mongolian mountains, Int. For. Fire News, 27, 9297.
  • Ciais, P., P. Peylin, and P. Bousquet (2000), The regional distribution of the biospheric carbon fluxes as inferred from atmospheric CO2 measurements and models, Ecol. Appl., 10, 15741589.
  • Dagvadorj, D., and R. Mijiddorj (1996), Climate change issues in Mongolia, in Hydrometeorological Issues in Mongolia: Papers in Hydrometeorology, edited by D. Dagvadorj, and L. Natsagdorj, pp. 7988, Mongolian Inst. of Meteorol. and Hydrol., Ulaanbaatar.
  • D'Arrigo, R., G. Jacoby, N. Pederson, D. Frank, B. Buckley, B. Nachin, R. Mijiddorj, and C. Dugarjav (2000), Mongolian tree rings, temperature sensitivity and reconstructions of Northern Hemisphere temperature, Holocene, 10, 669672.
  • Dixon, R. K., S. Brown, R. A. Houghton, A. M. Solomon, M. C. Trexler, and J. Winiewski (1994), Carbon pools and flux of global forest ecosystems, Science, 263, 185190.
  • Dixon, R. K., O. N. Krankina, and K. I. Kobak (1996), Global climate change adaptation: Examples from Russian boreal forests, in Adapting to Climate Change: Assessments and Issues, edited by J. B. Smith et al., pp. 359373, Springer, New York.
  • Erdenesaikhan, N., and M. Erdenetuya (1999), Forest and steppe fire monitoring in Mongolia using satellite remote sensing, Int. For. Fire News, 21, 7174.
  • Eugster, W., and W. Senn (1995), A cospectral correction model for measurements of turbulent NO2 flux, Boundary Layer Meteorol., 74, 321340.
  • Eugster, W., J. P. McFadden, and F. S. Chapin III (1997), A comparative approach to regional variation in surface fluxes using mobile eddy correlation towers, Boundary Layer Meteorol., 85, 293307.
  • Falge, E., et al. (2001), Gap filling strategies for defensible annual sums of net ecosystem exchange, Agric. For. Meteorol., 107, 4369.
  • Field, C. B., M. J. Behrenfeld, J. T. Randerson, and P. Falkowski (1998), Primary production of the biosphere: Integrating terrestrial and oceanic components, Science, 281, 237240.
  • Foley, J. A., J. E. Kutzbach, M. T. Coe, and S. Levis (1994), Feedbacks between climate and boreal forests during the Holocene epoch, Nature, 371, 5254.
  • Foley, J. A., M. H. Costa, C. Delire, N. Ramankutty, and P. Snyder (2003), Green surprise? How terrestrial ecosystems could affect Earth's climate, Frontiers Ecol. Environ., 1, 3844.
  • Frazer, G. W., R. A. Fournier, J. A. Trofymow, and R. J. Hall (2001), A comparison of digital and film fisheye photography for analysis of forest canopy structure and gap light transmission, Agric. For. Meteorol., 109, 249263.
  • Geider, R., et al. (2001), Primary productivity of planet Earth: Biological determinants and physical constraints in terrestrial and aquatic habitats, Global Change Biol., 7, 849882.
  • Gifford, R. M. (1994), The global carbon cycle: A viewpoint on the missing sink, Aust. J. Plant Physiol., 21, 115.
  • Goldammer, J. G. (2002), Fire situation in Mongolia, Int. For. Fire News, 26, 7583.
  • Goulden, M. L., B. C. Daube, S.-M. Fan, D. J. Sutton, A. Bazzaz, J. W. Munger, and S. C. Wofsy (1997), Physiological responses of a black spruce forest to weather, J. Geophys. Res., 102, 28,98728,996.
  • Goulden, M. L., et al. (1998), Sensitivity of boreal forest carbon balance to soil thaw, Science, 279, 214217.
  • Gower, S. T., O. Krankina, R. J. Olson, M. Apps, S. Linder, and C. Wang (2001), Net primary production and carbon allocation patterns of boreal forest ecosystems, Ecol. Appl., 11, 13951411.
  • Gunin, P. D., E. A. Vostokova, N. I. Dorofeyuk, P. E. Tarasov, and C. C. Black (1999), Vegetation Dynamics of Mongolia, Springer, New York.
  • Hall, F. G. (1999), Introduction to special section: BOREAS in 1999: Experiment and science overview, J. Geophys. Res., 104, 27,62727,639.
  • Hall, F. G. (2001), Introduction to special section: BOREAS III, J. Geophys. Res., 106, 33,51133,516.
  • Halldin, S., S. E. Gryning, L. Gottschalk, A. Jochum, L. C. Lundin, and A. A. van den Griend (1999), Energy, water and carbon exchange in a boreal forest landscape—NOPEX experiences, Agric. For. Meteorol., 98–99, 529.
  • Hicke, J. A., G. P. Asner, J. T. Randerson, C. J. Tucker, S. O. Los, R. Birdsey, J. C. Jenkins, C. B. Field, and E. A. Holland (2002), Satellite-derived increases in net primary productivity across North America, 1982–1998, Geophys. Res. Lett., 29(10), 1427, doi:10.1029/2001GL013578.
  • Hicke, J. A., G. P. Asner, E. S. Kasischke, N. H. F. French, J. T. Randerson, B. J. Stocks, C. J. Tucker, S. O. Los, and C. B. Field (2003), Postfire response of North American net primary productivity analyzed with satellite observations, Global Change Biol., 9, 11451157.
  • Hilbig, W. (1995), The Vegetation of Mongolia, SPB Acad., Amsterdam.
  • Högberg, P., A. Nordgren, N. Buchmann, A. F. Taylor, A. Ekblad, M. N. Högberg, G. Nyberg, M. Ottosson-Lofvenius, and D. J. Read (2001), Large-scale forest girdling shows that current photosynthesis drives soil respiration, Nature, 411, 789792.
  • Hollinger, D. Y., F. M. Kelliher, J. N. Byers, J. E. Hunt, T. M. McSeveny, and P. L. Weir (1994), Carbon dioxide exchange between an undisturbed old-growth temperature forest and the atmosphere, Ecology, 75, 134150.
  • Hollinger, D. Y., et al. (1998), Forest-atmosphere carbon dioxide exchange in eastern Siberian, Agric. For. Meteorol., 90, 291306.
  • Houghton, R. A., E. A. Davidson, and G. M. Woodwell (1998), Missing sinks, feedbacks, and understanding the role of terrestrial ecosystems in the global carbon balance, Global Change Biol., 12, 2534.
  • Intergovernmental Panel on Climate Change (1997), Special Report of IPCC Working Group II, The Regional Impacts of Climate Change: An Assessment of Vulnerability, edited by R. T. Watson, M. C. Zinyowera, and R. H. Moss, Cambridge Univ. Press, New York. (Available at http://www.grida.no/climate/ipcc/regional/267.htm).
  • Jacoby, G. C., R. D. D'Arrigo, and T. Davaajamts (1996), Mongolian tree rings and 20th century warming, Science, 273, 771773.
  • Jacoby, G. C., R. D. D'Arrigo, N. Pederson, B. M. Buckley, C. Dugarjav, and R. Mijiddorj (1999), Temperature and precipitation in Mongolia based on dendroclimatic investigations, Int. Assoc. Wood Anatomists J., 20, 339350.
  • Kasischke, E. S., and B. J. Stocks (2000), Fire, Climate Change, and Carbon Cycling in the Boreal Forest, Ecol. Stud., vol. 138, Springer, New York.
  • Kelliher, F. M., et al. (1999), Carbon dioxide efflux from the floor of a central Siberian pine forest, Agric. For. Meteorol., 94, 217232.
  • Kurz, W. A., and M. J. Apps (1999), A 70-yr retrospective analysis of carbon fluxes in the Canadian forest sector, Ecol. Appl., 9, 526547.
  • Larsen, J. A. (1980), The Boreal Ecosystem, Elsevier, New York.
  • Law, B. E., et al. (2002), Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation, Agric. For. Meteorol., 113, 97120.
  • Lee, X., and T. A. Black (1994), Relating eddy correlation sensible heat flux to horizontal sensor separation in the unstable atmospheric surface layer, J. Geophys. Res., 99, 18,54518,553.
  • Leuning, R., and J. Moncrieff (1990), Eddy covariance CO2 flux measurements using open- and closed-path analyzers: Corrections for analyzer water vapor sensitivity and damping fluctuations in air sampling tubes, Boundary Layer Meteorol., 53, 6376.
  • Leuning, R., F. M. Kelliher, D. G. G. de Pury, and E.-D. Schulze (1995), Leaf nitrogen, photosynthesis, conductance and transpiration: Scaling from leaves to canopies, Plant Cell Environ., 18, 11831200.
  • LI-COR Inc. (2000), LI-7500 CO2/H2O Analyzer Instruction Manual, LI-COR Inc., Lincoln, Nebr.
  • Lindroth, A., A. Grelle, and A.-S. Morén (1998), Long-term measurements of boreal forest carbon balance reveal large temperature sensitivity, Global Change Biol., 4, 443450.
  • Lloyd, J., and J. A. Taylor (1994), On the temperature dependence of soil respiration, Funct. Ecol., 8, 315323.
  • Lloyd, J., O. Shibistova, D. Zolothukhine, O. Kolle, A. Arneth, J. Styles, N. M. Tchebakova, and E.-D. Schulze (2002), Seasonal and annual variations in the photosynthetic productivity and carbon balance of a central Siberian Pine forest, Tellus, Ser. B, 54, 590609.
  • Mahrt, L. (1998), Flux sampling errors for aircraft and towers, J. Oceanic Atmos. Technol., 15, 416429.
  • Malhi, Y., D. D. Baldocchi, and P. G. Jarvis (1999), The carbon balance of tropical, temperate and boreal forests, Plant Cell Environ., 22, 715740.
  • Markkanen, T., U. Rannik, P. Keronen, T. Suni, and T. Vesala (2001), Eddy covariance fluxes over a boreal Scots pine forest, Boreal Environ. Res., 6, 6578.
  • Massman, W. J., and X. Lee (2002), Eddy covariance flux corrections and uncertainties in long-term studies of carbon and energy exchanges, Agric. For. Meteorol., 113, 121144.
  • McMillen, R. T. (1988), An eddy correlation technique with extended applicability to non-simple terrain, Boundary Layer Meteorol., 43, 231245.
  • Melillo, J. M., A. D. McGuire, D. W. Kicklighter, B. Moore III, C. J. Vorosmarty, and A. L. Schloss (1993), Global climate change and terrestrial net primary production, Nature, 363, 234240.
  • Moncrieff, J. B., Y. Malhi, and R. Leuning (1996), The propagation of errors in long-term measurements of land-atmosphere fluxes of carbon and water, Global Change Biol., 2, 231240.
  • Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemanl (1997), Increased plant growth in the northern high latitudes from 1981 to 1991, Nature, 386, 698702.
  • Pederson, N., G. C. Jacoby, R. D. D'Arrigo, E. R. Cook, and B. M. Buckley (2001), Hydrometeorological reconstructions for northeastern Mongolia derived from tree rings: 1651–1995, J. Clim., 14, 872881.
  • Plochl, M., and W. Cramer (1995), Coupling global models of vegetation structure and ecosystem processes, Tellus, Ser. B, 47, 240250.
  • Prentice, I. C., et al. (2001), The carbon cycle and atmospheric carbon dioxide, in Climate Change 2001: The Scientific Basis: Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, edited by J. T. Houghton et al., pp. 183237, Cambridge Univ. Press, New York.
  • Price, D. T., and T. A. Black (1990), Effects of short-term variation in weather on diurnal canopy CO2 flux and evapotranspiration of a juvenile Douglas-fir stand, Agric. For. Meteorol., 50, 139158.
  • Price, D. T., and T. A. Black (1991), Effects of summertime changes in weather and root-zone soil water storage on canopy CO2 flux and evapotranspiration of two juvenile Douglas-fir stands, Agric. For. Meteorol., 53, 303323.
  • Randerson, J. T., C. B. Field, I. Fung, and P. Tans (1999), Increases in early season ecosystem uptake explain changes in the seasonal cycle of atmospheric CO2 at high northern latitudes, Geophys. Res. Lett., 26, 27652768.
  • Reed, B. C., J. F. Brown, D. VanderZee, T. R. Loveland, J. W. Merchant, and D. Ohlen (1994), Measuring phenological variability from satellite imagery, J. Veg. Sci., 5, 703714.
  • Ruimy, A., P. G. Jarvis, D. D. Baldocchi, and B. Saugier (1995), CO2 fluxes over plant canopies and solar radiation: A review, Adv. Ecol. Res., 26, 151.
  • Schulze, E.-D., et al. (1999), Productivity of forests in the Eurosiberian boreal region and their potential to act as a carbon sink–A synthesis, Global Change Biol., 5, 703722.
  • Schulze, E.-D., et al. (2002), The Eurosiberian Transect: An introduction to the experimental region, Tellus, Ser. B, 54, 421428.
  • Sellers, P. J., et al. (1997), BOREAS in 1997: Experiment overview, scientific results, and future directions, J. Geophys. Res., 102, 28,73128,769.
  • Thornton, P. E., et al. (2002), Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests, Agric. For. Meteorol., 113, 185222.
  • Valendik, E. N., G. A. Ivanova, Z. O. Chuluunbaatar, and J. G. Goldammer (1998), Fire in forest ecosystems of Mongolia, Int. For. Fire News, 19, 5863.
  • Valentini, R., P. de Angelis, G. Matteucci, R. Monaco, S. Dore, and G. E. Scarascia Mugnozza (1996), Seasonal net carbon dioxide exchange of a beech forest with the atmosphere, Global Change Biol., 2, 199207.
  • Valentini, R., et al. (2000), Respiration as the main determinant of carbon balance in European forests, Nature, 404, 861865.
  • Vesala, T., et al. (1998), Long-term field measurements of atmosphere-surface interactions in boreal forest combining forest ecology, micrometeorology, aerosol physics and atmospheric chemistry, Trends Heat Mass Momentum Transfer, 4, 1735.
  • Webb, E. K., G. I. Pearman, and R. Leuning (1980), Correction of flux measurements for density effects due to heat and water vapour transfer, Q. J. R. Meteorol. Soc., 106, 85100.
  • Weiss, A., and J. M. Norman (1985), Partitioning solar radiation into direct and diffuse, visible and near-infrared components, Agric. For. Meteorol., 34, 205213.
  • White, A., M. G. R. Cannell, and A. D. Friend (2000), The high-latitude terrestrial carbon sink: A model analysis, Global Change Biol., 6, 227245.
  • Wilson, K. B., et al. (2002), Energy balance closure at FLUXNET sites, Agric. For. Meteorol., 113, 223243.
  • Wofsy, S. C., M. L. Goulden, J. W. Munger, S.-M. Fan, P. S. Bakwin, B. C. Daube, S. L. Bassow, and F. A. Bazzaz (1993), Net exchange of CO2 in a mid-latitude temperate forest, Science, 260, 13141317.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
jgrd11858-sup-0001-t01.txtplain text document1KTab-delimited Table 1.
jgrd11858-sup-0002-t02.txtplain text document1KTab-delimited Table 2.
jgrd11858-sup-0003-t03.txtplain text document1KTab-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.