SEARCH

SEARCH BY CITATION

Keywords:

  • Canadian carbon program;
  • climate change;
  • ecosystem net CO2 exchange;
  • Fluxnet-Canada

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Continuous half-hourly net CO2 exchange measurements were made using nine automatic chambers in a treed fen in northern Alberta, Canada from June–October in 2005 and from May–October in 2006. The 2006 growing season was warmer and drier than in 2005. The average chamber respiration rates normalized to 10 °C were much higher in 2006 than in 2005, while calculations of the temperature sensitivity (Q10) values were similar in the two years. Daytime average respiration values were lower than the corresponding, temperature-corrected respiration rates calculated from night-time chamber measurements. From June to September, the season-integrated estimates of chamber photosynthesis and respiration were 384 and 590 g C m−2, respectively in 2006, an increase of 100 and 203 g C m−2 over the corresponding values in 2005. The season-integrated photosynthesis and respiration rates obtained using the eddy covariance technique, which included trees and a tall shrub not present in the chambers, were 720 and 513 g C m−2, respectively, in 2006, an increase of 50 and 125 g C m−2 over the corresponding values in 2005. While both photosynthesis and respiration rates were higher in the warmer and drier conditions of 2006, the increase in respiration was more than twice the increase in photosynthesis.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Current consensus suggests that climate warming will result in increased release of CO2 from ecosystems, but there is large uncertainty about all the mechanisms involved, the magnitude of the expected positive feedback and the consequences for future ecosystem function (Luo 2007; Heimann & Reichstein 2008). The net exchange of CO2 in ecosystems depends on the balance between carbon uptake during photosynthesis and carbon loss during respiration. Most current global-scale carbon cycle models assume that photosynthesis in boreal regions is stimulated by increasing atmospheric CO2 levels and by warmer temperatures that affect maximum leaf area production and the length of the growing season, but these effects are expected to demonstrate saturating responses (Heimann & Reichstein 2008). By contrast, calculation of ecosystem respiration in global-scale models often assumes an exponential increase with higher temperatures. Thus, climate warming is predicted to result in an increased release of CO2 to the atmosphere when temperature-induced stimulation of respiration exceeds the stimulation of ecosystem photosynthesis caused by CO2 fertilization, higher leaf area production and longer growing seasons. Some uncertainties are introduced because of the strong interacting relationships between photosynthesis and respiration and the involvement of other environmental factors, such as soil water and nitrogen availability (Davidson & Janssens 2006; Luo 2007; Heimann & Reichstein 2008). In addition to control by temperature, ecosystem respiration is also strongly affected by plant growth and development. Photosynthetically-produced carbohydrates that are exuded from plant roots are a major source of carbon that affects the respiration rates of soil microbes (Högberg et al. 2001; Davidson & Janssens 2006; Johnsen et al. 2007). The rate and timing of the release of these respiratory substrates and the growth and respiration of roots strongly influence ecosystem respiration rate. The important linkage between photosynthesis and respiration and the interaction of these processes with multiple environmental factors cause large uncertainties in the predictions of ecosystem response to climate warming (Davidson & Janssens 2006; Heimann & Reichstein 2008).

Field studies of temporal (diurnal, seasonal and annual) variation in ecosystem CO2 exchange can provide important empirical insights into the effects of environmental variation on ecosystem physiological processes and potential ecosystem response to climate change (Baldocchi 2003, 2008). Eddy covariance measurements allow study of whole ecosystem net CO2 exchange during the day and the rates of total ecosystem respiration at night when photosynthesis is not active. However, night-time eddy covariance measurements can underestimate ecosystem respiration because of lack of turbulent exchange necessary for the application of the technique and because under some conditions advection can transport CO2 horizontally away from the measuring system (Goulden et al. 1996; Massman & Lee 2002). A variety of methods have been developed to correct for these problems associated with the eddy covariance technique, but the problems do introduce uncertainty in calculating annual sums and determining the sink-source status of an ecosystem (Goulden et al. 1996; Aubinet et al. 2000; Barr et al. 2004; Reichstein et al. 2005; Papale et al. 2006). Automatic chamber systems can be used to provide additional insights about controls on night-time respiration and daytime respiration and net exchange, at least on some ecosystem sub-components that are of a size that is feasible to enclose in chambers (Drewitt et al. 2002; Jassal et al. 2008). The chamber measurements also offer better opportunity to study the effect of light-inhibition of leaf respiration, which is often ignored in eddy covariance studies, but can result in values that are only 40–80% of leaf respiration rates measured in the dark (Wohlfahrt et al. 2005a).

In this study we combined the application of eddy covariance and automatic chamber measurements to test the response of a peatland ecosystem to warmer and drier conditions associated with interannual variation in weather. In Canada, peatlands are defined as wetland ecosystems with a minimum organic soil depth of 40 cm (National Wetlands Working Group 1988). Peatlands contain approximately one-third of the world's soil carbon pool and represent the largest pool of carbon in the Canadian terrestrial biosphere (Gorham 1991). The large carbon stocks in peatlands develop because of moderate rates of net primary production, while soil respiration is inhibited by low temperatures and anaerobic conditions associated with high water tables (Gorham 1991). It has been hypothesized that warming and lowered water tables could increase peat decomposition by exposing a greater volume of peat to aerobic conditions, thereby accelerating the release of CO2 to the atmosphere and altering the balance between peatland ecosystem photosynthesis and respiration (Moore, Roulet & Waddington 1998). Our main objective was to determine the changes in peatland ecosystem photosynthesis and respiration in the contrasting environmental conditions of two study years. We predicted that the warmer and drier conditions in 2006 would lead to higher ecosystem photosynthesis compared to 2005, but that the enhanced photosynthesis would not be high enough to offset the increase in ecosystem respiration, so the that the final result would be lower net ecosystem carbon sequestration during the warmer and drier conditions of 2006 relative to 2005.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Study site

The study site – the western peatland flux station – was established in 2003 as part of Fluxnet-Canada Research Network (FCRN, Margolis, Flanagan & Amiro 2006) and since 2008 is part of the follow-on Canadian Carbon Program. The flux tower (54.95384°N, 112.46698°W) was located approximately 80 km northeast of Athabasca, Alberta (54.82°N, 113.52°W). The mean annual temperature for the region was 2.1 °C, and annual precipitation was 504 mm, with approximately 382 mm from rainfall and 122 mm from snowfall (as recorded at Athabasca, Alberta; Environment Canada 2009).

A detailed description of vegetation and site characteristics was previously provided by Syed et al. (2006), and the following is a brief summary of that information. The site is a ‘moderately-rich treed fen’ according to the classification of Vitt (1994) and Vitt et al. (1998). The vegetation was dominated by stunted trees of Picea mariana and Larix laricina, with high abundance of a tall shrub, Betula pumila, and a wide range of moss species, including Sphagnum spp. (S. angustifolium, S. fuscum and others), brown moss species (Drepanocladus aduncus, Aulocomium palustre, and others) and a feather moss species (Pleurozium schreberi). The total ecosystem carbon stock (51 kg C m−2, total live and dead material in belowground peat plus live aboveground plant material) was dominated by carbon accumulated in below ground peat (maximum peat depth was approximately 2 m), with only about 1% of the total ecosystem carbon stock in aboveground live plant tissue. The two tree species accounted for about two-thirds of this aboveground live biomass. The total leaf area index for the site was 2.6 ± 0.16 m2 m−2 (average ± SE). The terrain is quite flat, with relatively homogeneous vegetation throughout a fetch of 1.5 to 2 km in all directions, except directly north where the fetch is approximately 1 km before the peatland grades into an upland aspen forest.

Eddy covariance and meteorological measurements

Syed et al. (2006) described in detail the meteorological and eddy covariance measurements at the site, here we provide a short summary of the relevant information for this study. Incoming photon flux density of photosynthetically active radiation (PAR) (Qt) was measured using a quantum sensor (LI-190SA, Li-Cor Inc., Lincoln, NE, USA) mounted at a height of 9 m on an instrument tower. Air temperature (Ta) was measured using an air temperature and relative humidity probe (Vaisala HMP45C, Campbell Scientific, Edmonton, Canada) mounted in a ventilated radiation shield at a height of 5 m. Water table depth relative to average hummock height (within a 2 m radius) was measured in a well using a custom float and counterweight system attached to a potentiometer. Volumetric soil moisture content was measured with soil water content reflectometers (CS616-L, Campbell Scientific), installed at depths of 7.5, 10 and 12.5 cm below the peat surface. The reflectometers had been previously calibrated in the lab in containers of commercial peat moss held at a range of water contents. With the exception of the rain gauge, all meteorological sensors were scanned at 5 s intervals and recorded as half-hourly means by a data logger (CR23X, Campbell Scientific) located in an insulated, heated and air-conditioned instrument hut. Cumulative precipitation was monitored with a weighing gauge (T-200B, Geonor Inc., Oslo, Norway) located approximately 800 m north of the flux tower. Any missing meteorological data due to power failure or system maintenance (<1% of possible half-hour periods) was filled by linear interpolation.

The eddy covariance (EC) technique (Baldocchi, Hicks & Meyers 1988; Aubinet et al. 2000; Baldocchi 2003) was used to measure net ecosystem fluxes of CO2, water vapour and sensible heat. The EC sensors, mounted on an instrument tower at a height of 9 m, consisted of a three-dimensional sonic anemometer-thermometer (SAT) (Solent R3, Gill Instruments, Lymington, UK) and a closed-path infrared gas (CO2/H2O) analyser (IRGA) (model LI-7000, Li-Cor Inc.). Half-hourly net CO2 flux was calculated as inline image, where inline image is the mean molar density of dry air, inline image is the covariance between instantaneous vertical wind speed (w) and CO2 mixing ratio (sc) (mol CO2 mol−1 dry air). The overbar and prime denote time average (half-hour) and fluctuations from the average, respectively. The rate of change in CO2 storage in the air column beneath the EC sensors was approximated as inline image, where hm is the measurement height (i.e., 9 m), inline image is the difference between inline image (half-hourly mean of sc) of the current and previous half-hours, and Δt = 1800 s. Half-hourly net ecosystem exchange (NEEec) of CO2 was then calculated as NEEec = Fc + Fs. Ecosystem photosynthesis was calculated as the difference between daytime ecosystem respiration (RecD) and daytime NEE. In this study, daytime and night-time periods were defined as Qt ≥ 20 µmol m−2 s−1 and Qt < 20 µmol m−2 s−1, respectively. We applied the FCRN standard procedure for partitioning eddy covariance measurements into photosynthesis and total ecosystem respiration (Barr et al. 2004). The sign convention in this paper is that negative NEE values denote CO2 being taken up by the ecosystem (C gain), and positive NEE values denote CO2 being released to the atmosphere (C loss).

Measurements of chamber net CO2 exchange

A non-steady state, automatic chamber system, described in detail by Gaumont-Guay et al. (2009) and Jassal et al. (2008), was used to make net CO2 exchange measurements. A brief description of the chamber system along with our site specific changes is included below. The chambers had domed-shaped, transparent acrylic tops (52.5 cm internal diameter, 20.5 cm in height) that were attached via a hinge assembly to a short (4 cm tall) polyvinyl chloride collar. The chamber top and short collar were subsequently attached to a secondary collar that was inserted into the soil/peat. Because of the uneven peat substrate, and relatively tall herbs and dwarf shrubs at our site, we used secondary collars that were 30 cm in height (52.5 cm internal diameter) and the collars were set approximately 5 cm deep into the peat. The secondary collars were installed in the peat in September 2004 by carefully slicing into the peat around the secondary collar until it could be secured about 5 cm below the surface and levelled. Measurements with the chambers did not start until June 2005, so ample time was allowed for any slight disturbance to the peat and vegetation to recover. Care was taken not to cause any significant disturbance to large roots, although some fine roots were severed during collar installation. With the chamber top installed on the secondary collar, a 1 m long metal support rod was driven into the peat and the portion of rod remaining above ground was attached to the chamber hinge assembly to prevent the chamber lid from tipping backward when in the open position. The metal rod and secondary collar were left in place over the winter, after the chamber tops were removed and stored off site. The moss, herb and dwarf shrub vegetation inside the secondary collars was left intact and undisturbed through the measurement periods and during the winter period between measurements when the chamber tops were removed from the site. While these chambers were originally designed to measure soil respiration in forest ecosystems, in our application the chambers measured net CO2 exchange (NEEch), resulting from photosynthesis, autotrophic respiration and heterotrophic respiration occurring in the chambers. The chamber side walls (mostly the secondary collar) and transparent top obscured some of the direct sunlight that would normally be incident on the ground vegetation, but this interference was relatively minor and significant photosynthetic activity was observed in the chambers during the daytime. The chamber tops were monitored and cleaned regularly during measurement periods. The chambers and collars did not include either of the two tree species or the tall shrub (B. pumila) at the site, so the net CO2 exchange measurements recorded by the chambers could not be expected to represent the exact same system that was measured by the eddy covariance flux system.

During a measurement, the chamber top was closed for a 2.5 min period. When a chamber was selected for a measurement and the chamber top closed, air was circulated between the chamber and a closed-path infrared gas analyser (LI-6262) at a rate of 9.5 L min−1. The tubing (6.3 mm OD Synflex 1300) used to connect the chambers and gas analyser was 15 m in length. The gas analyser was maintained at 37.5 ± 1 °C in a temperature-controlled housing. The analogue signal of the gas analyser was sampled at 1 Hz, averaged every 5 s, and stored on a data logger (CR23X, Campbell Scientific). The change in CO2 concentration in the chamber was recorded for 2.5 min (150 s) but only the data recorded during a 60-second period starting 20 s after the chamber was closed were used in calculating CO2 exchange rates. The operation of the chamber system (chamber selection, timing of opening and closing of lids, activation of solenoid valves, etc.) was controlled by a data logger (CR23X, Campbell Scientific). This data logger and associated pumps, and solenoid valves were housed in a temperature-controlled box. A total of nine chambers were measured in the main growing season months of 2005 (June-October) and 2006 (May-October).

Half-hourly net CO2 exchange rates were calculated as NEEch = ρa(Ve/A)(dsc/dt), where ρa is the molar density of dry air in the chamber headspace (mol m−3), Ve is the effective volume of the chamber (m3), A is the surface area covered by the chamber (m2), dsc/dt is the rate of change of CO2 mixing ratio in the chamber headspace over a 1-min interval following lid closure, and t is time. The effective volume of the chambers was measured once daily using a gas injection technique (with 10% CO2 in air), as described in detail by Drewitt et al. (2002). Calculation of the total system volume from dimensions of the components (0.081022 m3) was similar to average effective volume calculations (0.086775 m3). The slightly higher effective volume was likely due to adsorption of CO2 on the chamber materials. The gas analyser calibration was checked once daily by alternately injecting CO2-free nitrogen gas (zero offset) and air of known CO2 concentration (near ambient, span check) from high pressure gas cylinders into the air stream flowing to the gas analyser. All data recorded by the data logger were processed off-line using custom MatLab programs to calculate effective volume, adjust the calibration of CO2 concentration measurements and to calculate net CO2 exchange rates.

Calculation of chamber respiration capacity and temperature sensitivity

We assumed that the temperature dependence of respiration rate could be described by the following relationship:

  • image(1)

where Rch is chamber respiration rate, R10 is the respiration rate normalized to 10 °C (or respiration capacity), and Q10 is the temperature sensitivity coefficient describing the change in Rch for a 10 °C change in Ta.

We applied two approaches to calculation of R10 and Q10 values. First, for night-time measurements of chamber respiration (RchN), we calculated R10 and Q10 values by fitting a linear regression to the logarithmic transformation of Eqn 1; ln NEEchN = ln R10 + (Ta/10 − 1)ln Q10, where night-time NEEch (NEEchN) represents night-time chamber respiration (RchN). We also averaged the half-hourly values of NEEchN and Ta for each night, and the average of NEEchN was taken as mean night-time chamber respiration (inline image) and the average of Ta(inline image) was taken as the Ta associated with inline image.

Second, R10 and Q10 values were calculated for daytime chamber respiration measurements by assuming a linear relationship between daytime NEEch and Qt, when measured Qt values were low (between 20 and 400 µmol m−2 s−1):

  • image((2a))

where NEEchD are half-hourly daytime measurements of NEEch, α is the apparent quantum efficiency, and Qt values are the associated measurements of incident photon flux density (PAR). We set the lower bound of Q10 to 1.1 and its upper bound to 2.1, but no bounds were set for α and R10 during these linear regression calculations. We calculated mean daytime chamber respiration rate using a modified version of Eqn (2a):

  • image((2b))

where inline image is the mean daytime chamber respiration at inline image. The value of inline image was obtained as the average of Ta values associated with relevant measurements of NEEchD, not the average of all daytime Ta values.

We adjusted inline image from inline image to inline image using Eqn 3, so that inline image and inline image could be compared at the same temperature:

  • image(3)

where Adjusted inline image is inline image at inline image. The Q10 value used in Eqn 3 was fixed at 1.44, which was the average of night-time Q10 values (see below).

We employed a 5-day moving-window-average to smooth the daily estimates of R10 and Q10 values. The 5-day moving-window was centred at the value for the current day calculated using Eqns 1 and 2. The R10 and Q10 values from the 5-day moving-window-averages were then assigned to the current day as its final R10 and Q10 values.

Two-way analysis of variance (anova) with replication was used to test for significant differences (between effects of ‘month’ and ‘year’) in the daily estimates of R10 values calculated from the 5-day moving-window-average for each chamber. Because only 29 d of data were available for October 2006, we used the first 29 d of data for all months (June–October) for both years in the two-way anova in order to keep sample sizes equal for all months. Because only seven chambers were functional in June 2005, missing data for these two chambers in 2005 were replaced with the averages of data from the other seven chambers for the anova calculations only. The monthly averages of R10 values were then calculated from nine chambers × 29 d, for a total sample size of n = 261 for each month.

Correlation of chamber NEE with eddy covariance NEE

Half-hourly values of NEEec and NEEch were integrated over daytime and night-time periods, respectively, to obtain their corresponding daily totals. Daily totals of daytime NEEec (NEEecD) and night-time NEEec (NEEecN) were correlated with corresponding daily totals of NEEch using a multiple linear regression:

  • image((4a))
  • image((4b))

where ai and bi are empirical coefficients for the ith chamber, and NEEchD(i) and NEEchN(i) are daily totals of daytime and night-time NEEch, respectively for the ith chamber.

All the regression calculations, conducted to determine values for coefficients in Eqns 1–4, were performed using MatLab software.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Comparison of environmental conditions between 2005 and 2006

Large variation in monthly average temperature was observed in both 2005 and 2006, with the highest mean monthly temperature recorded in July and lowest in October (Fig. 1a). Monthly average temperatures were higher in 2006 than 2005 for all growing season months except October. In 2006, all months, except October, had average temperatures within one standard deviation of the long-term mean at nearby Athabasca, Alberta (1971–2000), while the temperature in October 2006 was 3 °C below the long-term average (Fig. 1a). It was cooler than normal in the growing season of 2005, as four months (June, July, August and September) had average temperatures greater than one standard deviation lower than the long-term average, while May and October 2005 were similar to normal temperatures.

image

Figure 1. Comparison of monthly average air temperature (a), monthly total precipitation (b) measured at the western peatland flux station during 2005 and 2006. Also shown is the long-term normal (30-year average ± SD) for air temperature and precipitation measured at the nearby, climate station in Athabasca, Alberta during 1971–2000 (Environment Canada 2009).

Download figure to PowerPoint

Throughout the growing season, monthly precipitation values in both 2005 and 2006 were within one standard deviation of the respective 30-year means (Fig. 1b). However, in June and July of 2006, precipitation was lower than that recorded in 2005. The warmer and drier conditions in June and July 2006 resulted in a decline of approximately 20 cm in the average water table depth relative to values recorded in 2005 (Fig. 2b). This decline in water table depth was associated with reductions in the shallow soil (peat) water content in 2006 compared to 2005 (Fig. 2a).

image

Figure 2. Comparison of daily average soil (peat) water content (a) and daily average water table depth (b), at the western peatland flux station during 2005 and 2006. Water content measurements represent the average of three probes located at depths of 7.5, 10, and 12.5 cm below the peat surface.

Download figure to PowerPoint

Chamber net carbon dioxide exchange measurements

As an example to illustrate the response of individual chamber NEEch to variation in short-term environmental conditions, we plot the continuous half-hourly measurements of four chambers during a period of eleven days in May 2006 (Fig. 3). In general night-time respiration measurements varied in association with changes in air temperature (although with a slight lag time), increasing during May 20–23 and then decreasing during May 24–26 (Fig. 3b–f). In particular, during May 24 a change in weather system resulted in a strong reduction in night-time air temperature from before sunrise to after sunset (shaded area in Fig. 3). All chambers showed associated reductions in night-time respiration from high rates before sunrise to lower rates after sunset on May 24. Daytime NEEch measurements were also very sensitive to changes in Qt. All chambers illustrated in Fig. 3 showed net uptake of CO2 during daylight periods with net loss of CO2 during night-time. In addition, a short-term reduction in Qt during the daytime, as indicated by the vertical arrows on May 25, led to net CO2 loss and subsequent rise in Qt led to net CO2 uptake in all the chambers (Fig. 3).

image

Figure 3. Temporal variation in incident photon flux density (Qt) (a), air temperature (Ta) (b), measurements of chamber (chambers 2, 3, 5, 9) net CO2 exchange rate (NEEch) (c–f) during selected days in May 2006.

Download figure to PowerPoint

There was strong seasonal variation in chamber photosynthetic and respiratory activity as illustrated by the mean diurnal pattern of NEEch during May–October (nine-chamber-average for all months, except June 2005 when only seven chambers were functioning, no chamber measurements were available in May 2005) (Fig. 4). Both photosynthetic and respiratory activity increased from May to July, remained similar in August and then declined during September and October. Night-time respiration rates were always higher in 2006 than those measured in 2005. Chamber photosynthetic activity, as approximately indicated by the difference between night-time respiratory CO2 loss and daytime net CO2 uptake, was higher in 2006 than 2005 for all months except October, when the photosynthetic activity was similar for 2 years (Fig. 4). However, the peak rates of net CO2 uptake during the day were very similar in 2005 and 2006 for all months except June, when peak net CO2 uptake was higher in 2006.

image

Figure 4. Comparison of seasonal variation in the mean diurnal pattern of chamber net CO2 exchange rate (NEEch) during 2005 and 2006. The NEEch values for each month were binned by time of day and averaged. The averages are for nine chambers in all months except June 2005, when only 7 chambers were functioning. No chamber measurements were available in May 2005. Open squares and filled circles represent data for 2005 and 2006, respectively.

Download figure to PowerPoint

There was large spatial variation in measured chamber CO2 exchange rates (Fig. 5). We used the temperature response of night-time and mid-day NEEch to illustrate this spatial variability among the nine chambers. Mid-day hours were defined as those from 11:00 am to 3:00 pm inclusive. Night-time and midday NEEch values for each chamber in each year were pooled for all the measurement periods, grouped into 2 °C Ta bins, and then averaged (Fig. 5). Average night-time NEEch at Ta = 20 °C for chamber 5 in 2006 was approximately 8 µmol m−2 s−1, while chamber 9 had a rate of approximately 4 µmol m−2 s−1, only 50% of that observed for chamber 5 (Fig. 5b). All chambers except chamber 2 showed net CO2 uptake during the day, while chamber 2 normally had small rates of net release of CO2 during daylight periods. All chambers showed higher rates of night-time respiration (at equivalent temperatures) during 2006 than in 2005 (Fig. 5a,b). Similarly some chambers had higher rates of net CO2 uptake (at equivalent temperatures) during the daytime in 2006 than in 2005, although the pattern was reversed for other chambers (Fig. 5c,d).

image

Figure 5. Variability among chambers (spatial variability) in night-time and daytime measurements of net CO2 exchange rate (NEEch). Night-time and mid-day NEEch values for each chamber in each year were pooled for all the measurement periods, grouped into 2 °C Ta bins, and then averaged. Mid-day hours were defined as those from 11:00 am to 3:00 pm inclusive.

Download figure to PowerPoint

Calculations of respiration capacity and temperature sensitivity

We first present data from one example chamber to illustrate our two approaches to calculation of R10 and Q10 values. For night-time measurements of chamber respiration (RchN), we calculated R10 and Q10 values by fitting a linear regression to the logarithmic transformation of Eqn 1. As shown for a single chamber (chamber 3) in 2006, we obtained good fits of Eqn 1 to our measured data (Table 1) and this was consistent for all chambers. Large (four-fold) seasonal variation occurred for calculated daily R10 values (Fig. 6b). The seasonal variation in R10 was correlated with associated changes in air temperature and average night-time respiration rate (inline image). In addition, for chamber 3 we observed higher peak values of R10 and inline image in 2006 than in 2005 (Fig. 6). The majority of the calculated Q10 values were within the range of 1.0 and 2.0 (Fig. 6c). The average (±SD) Q10 values for chamber 3 were 1.44 ± 0.26 and 1.44 ± 0.39 for 2005 and 2006, respectively. The temperature variation (ΔTa = maximum Ta − minimum Ta) used in the regressions for calculating R10 and Q10 values were mostly between 5–10 °C (Fig. 6e).

Table 1.  Comparison of calculated parameters (R10, Q10) and proportion of variance explained (r2) by fitting equation 1 to night-time respiration rate and air temperature measurements for chamber 3 on selected days in 2006
 24-May20-Jun17-Jul20-Aug26-Sep17-Oct
  1. Also shown are the average night-time chamber respiration rate (inline image) and average night-time air temperature (inline image). ΔTaN was calculated as the difference between maximum and minimum night-time air temperatures, and n is the number of data points used in the regression calculations.

R10 (µmol m−2 s−1)1.92.222.913.632.811.74
Q102.041.911.981.751.841.55
r20.810.660.840.840.470.15
inline image (µmol m−2 s−1)1.622.162.872.992.110.96
inline image (°C)7.478.929.256.254.88−3.56
ΔTaN (°C)8.279.839.579.9210.046.03
n151416192329
image

Figure 6. Seasonal variation in (a) average night-time air temperature (inline image), (b) night-time chamber respiration capacity (R10N) calculated using equation 1, (c) temperature sensitivity (Q10N) calculated using equation 1, (d) average night-time chamber respiration (inline image), and (e) the difference between maximum night-time Ta and minimum night-time TaTaN) for chamber 3 in 2005 and 2006. Lines in each plot were obtained using a 5-day-moving-average, as described in the main text.

Download figure to PowerPoint

Our second approach for calculating R10 and Q10 values made use of daytime chamber CO2 exchange measurements and assumed a linear relationship between daytime NEEch and Qt, when measured Qt values were low (between 20 and 400 µmol m−2 s−1). The y-intercept of this linear relationship was taken to represent the respiration rate in daylight (Eqn 2). As shown for chamber 3 in 2005, we obtained good fits of Eqn 2a to our measured data (Table 2) and this was consistent for all chambers. Seasonal changes in daytime R10 values were quite similar to those of night-time R10 values (compare Fig. 7b with Fig. 6b). The seasonal trend in calculated R10 values was also similar to that for calculated daytime inline image values (Fig. 7). In addition, for chamber 3 we observed higher peak values of R10 and inline image in 2006 than in 2005 (Fig. 7). The calculated Q10 values obtained using Eqn 2a were bound by the predetermined lower (Q10 = 1.1) and upper (Q10 = 2.1) limits. Average (±SD) calculated daytime Q10 values were 1.34 ± 0.27 and 1.38 ± 0.27 for 2005 and 2006, respectively. As expected, ΔTa values for daytime regression calculations were slightly larger than that for night-time calculations (compare Fig. 7e with Fig. 6e).

Table 2.  Comparison of calculated parameters (α, R10, Q10) and proportion of variance explained (r2) by fitting equation 2a to daytime measurements of chamber net CO2 exchange rate (NEEchD) and incident light intensity (Qt) for chamber 3 on selected days in 2005
 18-Jun9-Jul11-Aug16-Sep10-Oct
  1. Also shown are calculations of the average daytime respiration rate (inline image) by fitting Eqn 2b to daytime measurements of chamber net CO2 exchange rate and incident light intensity for chamber 3 on selected days in 2005. ΔTaD was calculated as the difference between maximum and minimum relevant daytime air temperatures, and n is the number of data points used in the regression calculations.

Equation 2a     
 α−0.007−0.01−0.013−0.008−0.005
 R10 (µmol m−2 s−1)2.062.943.011.461.04
 Q101.11.321.271.11.1
 r20.870.950.940.870.87
Equation 2b     
 α−0.007−0.011−0.013−0.008−0.005
 inline image (µmol m−2 s−1)2.063.533.121.421.06
 r20.880.910.930.870.89
 inline image (°C)9.8815.6127.5510.38
 ΔTaD (°C)4.425.983.056.288.53
 n2310101415
image

Figure 7. Seasonal variation in (a) average daytime air temperature (inline image), (b) daytime chamber respiration capacity (R10D) calculated using equation 2a, (c) temperature sensitivity (Q10D) calculated using equation 2a, (d) average daytime chamber respiration (inline image) calculated using equation 2b, and (e) the difference between maximum daytime Ta and minimum daytime TaTaD) for chamber 3 in 2005 and 2006. Lines in each plot were obtained using a 5-day-moving-average, as described in the main text.

Download figure to PowerPoint

In comparisons using data collected with all nine chambers, there was a very strong linear relationship between R10 values calculated using our two approaches in both years (2005: inline imageinline image, r2 = 0.935, n = 1302; 2006: inline imageinline image, r2 = 0.913, n = 1638). In addition, average night-time respiration rates that were adjusted to correspond to average daytime temperatures (adjusted inline image) had a strong linear relationship with average daytime respiration rates (inline image) (2005: inline image Adjusted inline image, r2 = 0.948, n = 1302; 2006: inline image Adjusted inline image, r2 = 0.942, n = 1638). However, inline image values were consistently lower than associated adjusted night-time respiration rates (regression slopes less than 1).

The mean monthly R10 values, averaged over all chambers, showed strong seasonal variation with peak values in July and August and much lower values in May and October (Fig. 8). In addition, the average R10 values were consistently higher in 2006 than 2005 for all months. Two-way anova indicated significant effects of ‘year’, ‘month’ and their interaction [year, F = 555.1, df = 1, P < 0.001; month, F = 407.0, df = 4, P < 0.001; interaction, F = 21.3, df = 4, P < 0.001; n = 261 (9 chambers × 29 d/month)].

image

Figure 8. Comparison of seasonal variation in monthly average values for (a) calculated respiration capacity (R10) and (b) calculated temperature sensitivity (Q10) during 2005 and 2006. Values represent the average of daytime and night-time calculations for all nine chambers for every day in a given month.

Download figure to PowerPoint

Correlation of chamber and eddy covariance NEE measurements

There was a strong correlation between eddy covariance measurements of NEE and calculations of NEE using chamber measurements weighted in a multiple linear regression equation (Eqn 5ab) (Table 3). The empirical coefficients calculated during the fitting of Eqn 4 were different for the 2 years, possibly because of the variation in environmental conditions and associated changes in respiration and photosynthetic rates. The empirical coefficients were also different for the daytime and night-time periods within a given year, because daytime NEE was controlled mainly by light and temperature, and night-time NEE was controlled mainly by temperature (Table 3).

Table 3.  Comparison of the calculated coefficients and proportion of variance explained (r2) by fitting equations 4ab to chamber and eddy covariance measurements of net CO2 exchange rate
Chamber20052006
DayNightDayNight
  1. The values are the coefficients, ai (for day) and bi (for night), from equations 4a (day) and 4b (night) for each chamber 1–9, respectively.

10.3230.0610.405−0.037
2−0.566−0.043−0.588−0.034
30.8541.4031.5690.313
41.1060.098−1.3040.503
50.0160.085−0.8430.019
6−3.8830.3620.0970.123
72.151−0.0280.453−0.006
80.193−1.2130.1870.003
91.8850.2221.5300.043
r20.940.810.790.82

Comparison of integrated carbon budget calculations for 2005 and 2006

The chamber and eddy covariance measurements of NEE and calculations of photosynthesis and respiration were integrated in order to provide daily, monthly and seasonally integrated values of carbon dioxide exchange. For chamber measurements, the integrated-respiration values were higher in 2006 than in 2005 for all months (Fig. 9a), consistent with the higher temperatures (Fig. 1a) and higher respiration capacity in 2006 (Fig. 8a). Similarly, integrated-photosynthesis rates were higher in 2006 than in 2005 for all months except October, when photosynthesis was very slightly higher in 2005 (Fig. 9c). However, the monthly integrated chamber respiration rates were always greater than associated measurements of photosynthesis, so there was a net loss of CO2 from the portion of the ecosystem contained in the chambers (Fig. 9e). The magnitude of the seasonally-integrated net CO2 loss was larger in 2006 than 2005, because in 2006 chamber respiration increased more than did chamber photosynthesis (Table 4).

image

Figure 9. Comparison of seasonal variation in monthly-integrated values of (a) respiration, (c) photosynthesis, and (e) net CO2 exchange (NEE) measured with chambers during 2005 and 2006. Also shown are the monthly-integrated values of (b) respiration, (d) photosynthesis, and (f) NEE measured with the eddy covariance technique during 2005 and 2006. For the chambers, the monthly values represent averages of the data from the nine chambers.

Download figure to PowerPoint

Table 4.  Comparison of the seasonally-integrated (June–September only) calculations of carbon exchange (g C m−2 period−1) for respiration, photosynthesis and net CO2 exchange rate (NEE) obtained using the chamber and eddy covariance techniques during 2005 and 2006
 20052006
Chamber  
 Respiration387590
 Photosynthesis−284−384
 NEE103206
Eddy covariance  
 Respiration388513
 Photosynthesis−670−720
 NEE−282−207

For eddy covariance measurements, the integrated-respiration values were higher in 2006 than in 2005 for all months except May and October, when rates were approximately equal in both years (Fig. 9b). Integrated-photosynthesis rates were approximately the same or slightly higher in 2006 than in 2005 for all months except July (Fig. 9d). The entire ecosystem measured by eddy covariance was a net sink for carbon dioxide integrated over the entire growing season in both years, but the magnitude of the sink was larger in 2005 (Fig. 9f), primarily because in 2006 ecosystem respiration increased more than did ecosystem photosynthesis (Fig. 9bd, Table 4).

The differences in environmental conditions during the growing seasons of 2005 and 2006 did not have any influence on carbon exchange rates during the following fall and winter periods (Table 5). All three carbon budget components (respiration, photosynthesis and net ecosystem CO2 exchange) had very similar integrated values based on eddy covariance measurements during the fall and fall-winter periods of 2005–06 and 2006–07.

Table 5.  Comparison of the integrated calculations of carbon exchange (g C m−2 period−1) for respiration, photosynthesis and net CO2 exchange rate (NEE) obtained using the eddy covariance technique during fall (October-December) and fall &winter (October–March)
 20052006
October–December  
 Respiration4647
 Photosynthesis−17−14
 NEE2933
2005–062006–07
October–March  
 Respiration6268
 Photosynthesis−19−16
 NEE4352

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Calculation of a standardized rate of respiration (R10) provides insight into the capacity of the ecosystem for respiration, a variable that will depend on factors influencing both autotrophic and heterotrophic respiration (Reichstein et al. 2002; Flanagan & Johnson 2005). The magnitude of autotrophic respiration is influenced by factors that control plant growth and development, photosynthesis and carbon allocation patterns (Davidson & Janssens 2006). Heterotrophic respiration is dependent on the supply of substrate, which is primarily provided from plant litter and plant root exudates, in addition to environmental conditions (temperature and moisture) that control microbial growth and development (Davidson & Janssens 2006). We calculated R10 values separately for day and night time periods. Using the chamber data, the two independent techniques produced R10 values that were very strongly correlated (2005, r = 0.967; 2006, r = 0.956). In addition, our data showed that respiration capacity varied temporally during the growing season and between the two study years (Fig. 8a). The seasonal variation in R10 values was a result of changes in temperature and correlated variation in autotrophic and heterotrophic activities within the growing season. The higher R10 values in 2006 were due to a combination of factors including the warmer temperatures stimulating higher ecosystem photosynthesis (Table 4), possibly because of higher leaf area production in the deciduous vascular plants and/or higher leaf-level photosynthetic capacity. The lower water table (Fig. 2b) would also provide for a larger aerobic zone in the surface peat which would contribute to better growth conditions for vascular plants and higher rates of ecosystem photosynthesis and respiratory capacity. The R10 values we calculated were similar to ecosystem respiration rates (0.5–5 µmol m−2 s−1) standardized to 12  °C for bogs and fens in Finland (Silvola et al. 1996), and similar to, or slightly higher than, R10 values calculated for soil respiration in several other ecosystem types (Reichstein et al. 2002; Kowalski et al. 2003; Flanagan & Johnson 2005; Wohlfahrt et al. 2005b; Gaumont-Guay et al. 2009). The cool and water logged conditions of most of the peat substrate, and generally low aboveground biomass at our peatland should reduce respiratory activity relative to photosynthetic activity compared to such ratios for productive forest and grassland ecosystems (Frolking et al. 1998; Bubier et al. 1999). The generally high annual rates of net CO2 sequestration that have been measured in this peatland are illustrative of low ratios of ecosystem respiration relative to photosynthesis (Syed et al. 2006).

The Q10 values we calculated from both night-time and daytime chamber measurements were quite stable within a growing season and between study years (Fig. 8b). Our averaged night-time and daytime Q10 values were 1.44 and 1.34, respectively. These Q10 values were lower than 2, probably as a result of our use of air temperature in the calculations. Air temperature was used because of the expected large contribution from autotrophic (aboveground vegetation and root) respiration, and the uniformity of air temperature relative to the potentially large spatial variation in soil temperature profiles across an uneven peatland surface.

Daytime chamber respiration values were lower than the corresponding night-time chamber respiration values. This suggests that chamber respiration was inhibited in the light relative to that in darkness (Kok 1948; Sharp, Matthews & Boyer 1984; Brooks & Farquhar 1985; Shapiro et al. 2004). Wohlfahrt et al. (2005a) surveyed several leaf-level studies and showed that reduction of foliar respiration in the light occurred quickly with increasing light intensity and then stabilized when Qt was greater than 100 µmol m−2 s−1. In general, Wohlfahrt et al. (2005a) determined that foliar respiration measured in the light was only 40–80% of that measured in the dark. In our study, the integrated daytime respiration during the period of June to September was 252 and 390 g C m−2 for 2005 and 2006, respectively (Table 4). If we had estimated daytime chamber respiration by extrapolating the night-time R10-Q10 relationship obtained using Eqn 1 to daytime periods, the integrated daytime chamber respiration for the same period would have been 276 and 411 g C m−2 for 2005 and 2006, respectively. Therefore, the reduction of daytime chamber respiration by light inhibition was 9% (1–252/276) and 5% (1–390/411) in 2005 and 2006, respectively. These reductions are lower than the 40–80% noted above because our chamber measurements include respiration from several components in addition to plant leaf respiration, and light inhibition of respiration occurs only in leaves.

Our growing season-integrated estimates of respiration derived from chamber measurements were approximately the same or higher than that determined by eddy covariance measurements (Table 4). Syed et al. (2006) calculated that the uncertainty on annual NEE at our peatland site was approximately 30 g C m−2 year−1, and similar uncertainty values should apply for ecosystem photosynthetic and respiratory rates during the growing season. Including this consideration of uncertainty, the chamber technique did produce higher seasonally-integrated respiration rates than the eddy covariance technique in 2006, despite the fact that the chambers did not include all ecosystem components, as the two tree species and a large shrub species were not present in the chambers. Several studies in forest ecosystems have also observed that scaled-up chamber respiration measurements were higher than eddy covariance estimates of respiration (Goulden et al. 1996; Lavigne et al. 1997; Law et al. 2000; Drewitt et al. 2002; Griffis et al. 2004). One major problem in scaling chamber measurements up to the entire ecosystem is the large spatial variability often observed among different chambers and the lack of information necessary to weight the representation of each different chamber (Lavigne et al. 1997). We also observed a great deal of variation among CO2 exchange measurements in different chambers (Fig. 5). However, strong correlations were observed between weighted-average chamber and eddy covariance measurements of NEE during the daytime and night-time in both study years (Table 3). The coefficients used to weight the different chamber measurements were quite variable between day and night periods and between study years (Table 3). In addition, the coefficients were simply calculated via multiple linear regression, and so were not based on knowledge of the spatial distribution of the ecological attributes of the different chambers within the footprint sampled by the eddy covariance system. The data in Table 3 do indicate, however, very good agreement between the temporal patterns of net CO2 exchange measurements made with the two techniques. We also observed quite good quantitative agreement between seasonally-integrated estimates of photosynthesis made using the chamber and eddy covariance techniques, when allowance was made for the fact that leaf area in the chambers was only approximately 50% of that of the entire ecosystem. Syed et al. (2006) previously measured a combined leaf area index of 1.3 for the herbs, dwarf shrubs and moss that were components in our chambers, while the two tree species (P. mariana and L. laricina) and a tall shrub (B. pumila) added another 1.3 for a total ecosystem leaf area index of 2.6. The seasonally integrated (June–September) photosynthesis rates calculated for chambers were 284 and 383 g C m−2 for 2005 and 2006, respectively, which were 42 and 53% of the photosynthesis rates calculated from eddy covariance measurements during 2005 and 2006, respectively (Table 4). It is possible that chamber photosynthesis was slightly underestimated because the chamber walls and domes slightly shaded the vegetation inside the chambers, and that chamber respiration was slightly overestimated because the air inside the chambers was moderately warmed when the domes were closed for measurements, and this warming would slightly increase respiration but have little effect on photosynthesis in the chambers.

The warmer and drier conditions of 2006 resulted in higher June-September integrated photosynthesis rates, measured by both chamber and eddy covariance techniques, than occurred in 2005 (Table 4). However, net CO2 sequestration was reduced in 2006 relative to 2005, because respiration rates were stimulated more in 2006 than was photosynthesis. The higher integrated respiration rates were the combined result of the warmer temperatures and increased respiration capacity (R10) in 2006 (no significant change occurred in Q10). These results are consistent with predictions of several modelling studies that calculate increased release of CO2 from ecosystems to the atmosphere in response to climate change (Heimann & Reichstein 2008). Additional global-scale studies have indicated that climate change has stimulated earlier starts to the growing season in boreal and temperate zones, but that the length of the growing season has not increased because of shifts in the ratio of photosynthesis and respiration at the end of the growing season (Piao et al. 2008). Drought-induced reductions in photosynthesis combined with stimulations of respiration by higher temperatures may have caused the earlier reduction in net carbon uptake at the end of the growing season (Piao et al. 2008). A recent remote sensing study has indicated that higher net carbon sequestration, associated with an earlier spring start to the growing season, can be offset by late-summer drought effects on photosynthesis in boreal regions (Angert et al. 2005). However, comparative eddy covariance studies conducted in boreal forests of northern Manitoba in Canada showed that the dominant effect of warmer temperatures was an earlier spring period and an extension of the growing season length, with no evidence of significant drought-induced reduction in photosynthesis, at least in relative old, mature coniferous forests (dominated by P. mariana) (McMillan, Winston & Goulden 2008). Drought reduction in photosynthesis was a secondary effect noted mainly in younger forests with an abundance of broad-leaf deciduous trees in addition to evergreen conifers (McMillan et al. 2008). In our study, there was no evidence of any significant shift in the seasonal pattern of either photosynthesis or respiration, and rates of both respiration and photosynthesis in May and October of 2005 and 2006 were very similar based on eddy covariance measurements (Fig. 9b,d). The largest difference in eddy covariance-measured NEE between 2005 and 2006 occurred during July, in the middle of the growing season (Fig. 9f). Despite the June–September integrated photosynthesis rates being higher in 2006 than 2005, photosynthesis in July 2006 was actually lower than in July 2005 (Fig. 9f). This pattern was not apparent for the chamber estimates of photosynthesis, which showed consistently higher rates in 2006 than in 2005 for all months during June-September (Fig. 9c). This suggests that photosynthesis in the two tree species and B. pumila during July 2006 were affected by some conditions that did not have the same magnitude of effect on the lower stature herbs and dwarf shrubs at the site. The warmer and drier conditions in 2006 were likely associated with decreased diffuse radiation and higher VPD, which can result in reduced photosynthesis in tree canopies (Roderick et al. 2001; Gu et al. 2002; Cai et al. 2009). Despite the relatively open tree and tall shrub canopy at our study site, the tightly-clustered needle leaves and dense branching structure of P. mariana and L. laricina trees could have been negatively influenced by relatively less diffuse radiation under clear days in July. In addition, P. mariana trees do show strong response of stomatal closure and increased stomatal limitation of photosynthesis under high VPD conditions (Dang et al. 1997). The major effect of the warmer and drier conditions in our study, however, was that ecosystem respiration was stimulated more than photosynthesis. This was consistent with long-held physiological principles, the temperature optimum for photosynthesis occurs at a much lower temperature than that for respiration, and the optimum region of the photosynthesis–temperature response curve has a broad plateau (Larcher 1995). While this study examined only two growing seasons and the peatland was still a strong net CO2 sink on an annual basis, if the trend to larger stimulation of respiration than photosynthesis continues with further warming, the ecosystem could eventually switch to become a CO2 source. Our study was consistent with the results of a warming experiment conducted in a subarctic peatland. Dorrepaal et al. (2009) showed that warming of about 1 °C resulted in 60% higher ecosystem respiration in spring and 52% higher rates in summer, effects that were maintained over an eight-year study. However, further studies are required to examine the possible interactions and feedbacks that may occur to respiration as photosynthesis is reduced by the combined effects of drought and temperatures that become warmer than the temperature optimum for photosynthesis.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

This research was part of the Fluxnet-Canada Research Network and the Canadian Carbon Program, and was supported by grants to LBF from Natural Sciences and Engineering Council of Canada (NSERC), the Canadian Foundation for Climate and Atmospheric Sciences (CFCAS), BIOCAP Canada, and the University of Lethbridge. We thank Peter Carlson for help with the field measurements and the initial setup of eddy covariance and chamber systems.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES
  • Angert A., Biraud S., Bonfils C., Henning C.C., Buermann W., Pinzon J., Tucker C.J. & Fung I. (2005) Drier summers cancel out the CO2 uptake enhancement induced by warmer springs. Proceedings of the National Academy of Sciences of the United States of America 102, 1082310827.
  • Aubinet M., Grelle A., Ibrom A., et al. (2000) Estimates of the annual net carbon and water exchange of forests: the EUROFLUX methodology. Advances in Ecological Research 30, 113176.
  • Baldocchi D.D. (2003) Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of ecosystems: past, present and future. Global Change Biology 9, 479492.
  • Baldocchi D.D. (2008) Breathing' of the terrestrial biosphere: lessons learned from a global network of carbon dioxide flux measurements. Australian Journal of Botany 56, 126.
  • Baldocchi D.D., Hicks B.B. & Meyers T.D. (1988) Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods. Ecology 69, 13311340.
  • Barr A.G., Black T.A., Hogg E.H., Kljun N., Morgenstern K. & Nesic Z. (2004) Inter-annual variability in the leaf area index of a boreal aspen-hazelnut forest in relation to net ecosystem production. Agricultural and Forest Meteorology 126, 237255.
  • Brooks A. & Farquhar G.D. (1985) Effect of temperature on the CO2/O2 specificity of ribulose-1,5-bisphosphate carboxylase/oxygenase and the rate of respiration in the light. Planta 165, 397406.
  • Bubier J.L., Frolking S., Crill P.M. & Linder E. (1999) Net ecosystem productivity and its uncertainty in a diverse boreal peatland. Journal of Geophysical Research 104, 2768327692.
  • Cai T., Black T.A., Jassal R.S., Morgenstern K. & Nesic Z. (2009) Incorporating diffuse photosynthetically active radiation in a single-leaf model of canopy photosynthesis for a 56-year-old Douglas-fir forest. International Journal of Biometeorology 53, 135148.
  • Dang Q-L., Margolis H.A., Coyea M.R., Sy M. & Collatz G.J. (1997) Regulation of branch-level gas exchange of boreal trees: roles of shoot water potential and vapor pressure difference. Tree Physiology 17, 521535.
  • Davidson E.A. & Janssens I.A. (2006) Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440, 165173.
  • Dorrepaal E., Toet S., Van Logtestijn R.S.P., Swart E., Van De Weg M.J., Callaghan T.V. & Aerts R. (2009) Carbon respiration from subsurface peat accelerated by climate warming. Nature 460, 616619.
  • Drewitt G.B., Black T.A., Nesic Z., Humphreys E.R., Jork E.M., Swanson R., Ethier G.J., Griffis T. & Morgenstern K. (2002) Measuring forest floor CO2 fluxes in a Douglas-fir forest. Agricultural and Forest Meteorology 110, 299317.
  • Environment Canada (2009) Canadian climate normals or averages: 1971–2000. [WWW document]. URL http://climate.weatheroffice.ec.gc.ca/climate-normals/index_e.html[accessed on 8 December 2009].
  • Flanagan L.B. & Johnson B.G. (2005) Interacting effects of temperature, soil moisture, and plant biomass production on ecosystem respiration in a northern temperate grassland. Agricultural and Forest Meteorology 130, 237253.
  • Frolking S.E., Bubier J.L., Moore T.R., et al. (1998) Relationship between ecosystem productivity and photosynthetically active radiation for northern peatlands. Global Biogeochemical Cycles 12, 115126.
  • Gaumont-Guay D., Black T.A., McCaughey H., Barr A.G., Krishnan P., Jassal R.S. & Nesic Z. (2009) Soil CO2 efflux in contrasting boreal deciduous and coniferous stands and its contribution to the ecosystem carbon balance. Global Change Biology 15, 13021319. Doi: 10.1111/j.1365-2486.2008.01830.x.
  • Gorham E. (1991) Northern peatlands: role in the carbon cycle and probable responses to climatic warming. Ecological Applications 1, 182195.
  • Goulden M.L., Munger J.M., Fan S.-M., Daube B.C. & Wofsy S.C. (1996) Measurements of carbon sequestration by long-term eddy covariance: methods and a critical evaluation of accuracy. Global Change Biology 2, 169182.
  • Griffis T.J., Black T.A., Gaumont-Guay D., Drewitt G.B., Nesic Z., Barr A.G., Morgenstern K. & Kljun N. (2004) Seasonal variation and partitioning of ecosystem respiration in a southern boreal aspen forest. Agricultural and Forest Meteorology 125, 207223.
  • Gu L., Baldocchi D.D., Verma S.B., Black T.A., Vesala T., Falge E.M. & Dowty P.R. (2002) Advantages of diffuse radiation for terrestrial ecosystem productivity. Journal of Geophysical Research (D6), 107, 4050.
  • Heimann M. & Reichstein M. (2008) Terrestrial ecosystem carbon dynamics and climate feedbacks. Nature 451, 289292.
  • Högberg P., Nordgren A., Buchmann N., Taylor A.F.S., Ekblad A., Högberg M., Nyberg G., Ottosson-Lofvenius M. & Read D.J. (2001) Large-scale forest girdling shows that current photosynthesis drives soil respiration. Nature 411, 789792.
  • Jassal R.S., Black T.A., Novak M.D., Gaumont-Guay D. & Nesic Z. (2008) Effect of soil water stress on soil respiration and its temperature sensitivity in an 18-year-old temperate Douglas-fir stand. Global Change Biology 14, 13051318.
  • Johnsen K., Maier C., Sanchez F., Anderson P., Butnor J., Waring R. & Linder S. (2007) Physiological girdling of pine trees via phloem chilling: proof of concept. Plant, Cell & Environment 30, 128134.
  • Kok B. (1948) A critical consideration of the quantum yield of Chlorella photosynthesis. Enzymology 13, 156.
  • Kowalski S., Sartore M., Burlett R., Berbigier P. & Loustau D. (2003) The annual carbon budget of a French pine forest (Pinus pinaster) following harvest. Global Change Biology 9, 10511065.
  • Larcher W. (1995) Physiological Plant Ecology: Ecophysiology and Stress Physiology of Functional Groups, 3rd edition. Springer-Verlag, Berlin, Germany.
  • Lavigne M.B., Ryan M.G., Anderson D.E., et al. (1997) Comparing nocturnal eddy covariance measurements to estimates of ecosystem respiration made by scaling chamber measurements. Journal of Geophysical Research 102, 2897728986.
  • Law B.E., Williams M., Anthoni P.M., Baldocchi D.D. & Unsworth M.H. (2000) Measuring and modeling seasonal variation of carbon dioxide and water vapour exchange of Pinus ponderosa forest subject to soil water deficit. Global Change Biology 6, 613630.
  • Luo Y. (2007) Terrestrial carbon-cycle feedback to climate warming. Annual Review of Ecology, Evolution and Systematics 38, 683712.
  • McMillan A.M., Winston G.C. & Goulden M.L. (2008) Age-dependent response of boreal forest to temperature and rainfall variability. Global Change Biology 14, 19041916. Doi: 10.1111/j.1365-2486.2008.01616.x.
  • Margolis H.A., Flanagan L.B. & Amiro B.D. (2006) The Fluxnet-Canada research network: influence of climate and disturbance on carbon cycling in forests and peatlands. Agricultural and Forest Meteorology 140, 15.
  • Massman W.J. & Lee X. (2002) Eddy covariance flux corrections and uncertainties in long-term studies of carbon and energy exchanges. Agricultural and Forest Meteorology 113, 121144.
  • Moore T.R., Roulet N.T. & Waddington J.M. (1998) Uncertainty in predicting the effect of climatic change on the carbon cycling of Canadian peatlands. Climatic Change 40, 229245.
  • National Wetlands Working Group (1988) Wetlands of Canada. Ecological Land Classification Series No. 24. Sustainable Development Branch, Environment Canada, Ottawa, Ontario, Canada. Polyscience Publications, Inc., Montreal, Quebec.
  • Papale D., Reichstein M., Aubinet M., et al. (2006) Towards a standardized processing of net ecosystem exchange measured with the eddy covariance technique: algorithms and uncertainty estimation. Biogeosciences 3, 571538.
  • Piao S., Ciais P., Friedlingstein P., et al. (2008) Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature 451, 4953.
  • Roderick M.L., Farquhar G.D., Berry S. & Noble I.R. (2001) On the direct effect of clouds and atmospheric particles on the productivity and structure of vegetation. Oecologia 129, 2130.
  • Reichstein M., Tenhunen J.D., Roupsard O., Ourcival J.-M., Rambal S., Dore S. & Valentini R. (2002) Ecosystem respiration in two Mediterranean evergreen Holm Oak forests: drought effects and decomposition dynamics. Functional Ecology 16, 2739.
  • Reichstein M., Falge E., Baldocchi D.D., et al. (2005) On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm. Global Change Biology 11, 14241439.
  • Shapiro J.B., Griffin K.L., Lewis J.D. & Tissue D.T. (2004) Response of Xanthium strumarium leaf respiration in the light to elevated CO2 concentration, nitrogen availability and temperature. New Phytologist 162, 377386.
  • Sharp R.E., Matthews M.A. & Boyer J.S. (1984) Kok effect and the quantum yield of photosynthesis: light partially inhibits dark respiration. Plant Physiology 75, 95101.
  • Silvola J., Alm J., Ahlholm U., Nykanen H. & Martikainen P.J. (1996) CO2 fluxes from peat in boreal mires under varying temperature and moisture conditions. Journal of Ecology 84, 219228.
  • Syed K.H., Flanagan L.B., Carlson P.J., Glenn A.J. & Van Gaalen K.E. (2006) Environmental control of net ecosystem CO2 exchange in a treed, moderately rich fen in northern. Alberta. Agricultural and Forest Meteorology 140, 97114.
  • Vitt D.H. (1994) An overview of the factors that influence the development of Canadian peatlands. Memoirs of the Entomological Society of Canada 169, 720.
  • Vitt D.H., Halsey L.A., Thormann M.N. & Martin T. (1998) Peatland Inventory of Alberta. Phase 1: Overview of Peatland Resources in the Natural Regions and Subregions of the Province. Sustainable Forest Management Network, University of Alberta, Edmonton, AB, Canada.
  • Wohlfahrt G., Bahn M., Haslwanter A., Newesely C. & Cernusca A. (2005a) Estimation of daytime ecosystem respiration to determine gross primary production of a mountain meadow. Agricultural and Forest Meteorology 130, 1325.
  • Wohlfahrt G., Anfang C., Bahn M., Haslwanter A., Newesly C., Schmitt M., Drosler M., Pfadenhauer J. & Cernusca A. (2005b) Quantifying nighttime ecosystem respiration of a meadow using eddy covariance, chambers and modeling. Agricultural and Forest Meteorology 128, 141162.