Peatlands are globally important long-term sinks of atmospheric carbon dioxide (CO₂). However, there is concern that climate change-mediated drying will reduce gross primary productivity (GPP) and increase ecosystem respiration (ER) making peatlands vulnerable to a weaker carbon sink function and potential net carbon loss. While large and deep peatlands are usually resilient to moderate summer drying, CO₂ exchange in shallow Boreal Shield peatlands is likely more sensitive to drying given the reduced groundwater connectivity and water storage potential. To better understand the carbon cycling responses of Boreal Shield peatlands to meteorological conditions, we examined ecohydrological controls on CO₂ fluxes using the eddy covariance technique at a shallow peatland during the summer season for 5 years, from 2016–2020. We found lower GPP in dry summer years. Mean summer water table depth (WTD) was found to be significantly correlated with summer total net ecosystem CO₂ exchange (R² = 0.78; p value = 0.046) and GPP (R² = 0.83; p value = 0.03), where wet summers with a WT close to the peat surface sequestered more than twice the amount of CO₂ than dry summers. Our findings suggest that shallow Boreal Shield peatland GPP may be sensitive to climate-mediated drying as they may switch to a net CO₂ source in the summer season when WTDs exceed a critical ecohydrological threshold for a prolonged period of time.

Plain Language Summary

Peatlands take in carbon from the atmosphere and store it in the ground as peat, a process that helps to regulate climate change. Peatlands in the Boreal Shield are positioned in bedrock basins. Their water table (WT) is controlled by precipitation, which can trigger water flow over the bedrock between wetland ecosystems. Due to this unique setting, we expected Boreal Shield peatlands to be more sensitive to differences in growing season meteorological conditions from year to year. We used 5 years of carbon dioxide (CO₂) exchange measurements between land and atmosphere during the summer season in a Boreal Shield peatland in Ontario, Canada. We found the peatland vegetation took up more CO₂ from the atmosphere during summers with more rain which keeps the WT in the peatland closer to the peat surface. The peatland took up less CO₂ when the summer was dry. Our findings provide insight into how Boreal Shield peatlands are responding to summer droughts under current climate conditions.

Key Points

Summers with higher rainfall (and lower P-PET) maintain a water table near the peat surface and have greater net carbon dioxide (CO₂) uptake
Interannual differences in summer net ecosystem CO₂ exchange are attributable to changes in summer gross primary production
Shallow Boreal Shield peatlands may switch to a net carbon source in the summer season

1 Introduction

Northern peatlands are globally important ecosystems for long-term carbon storage (Blodau et al., 2004; Clymo, 1987), and have exerted a global climate cooling effect over millennia (Frolking et al., 2006). While only covering ∼3% of the global land area, peatlands store approximately one third of the global organic soil carbon pool (Gorham, 1991; Xu et al., 2018; Yu et al., 2010). This long-term accumulation of peat is generally facilitated by a water table (WT) level near the peatland surface, which limits peat decomposition rates (Benscoter et al., 2005; Clymo, 1984) and maintains high photosynthetic capacity for the key peat-forming Sphagnum mosses (Bubier et al., 2003; Hájek & Beckett, 2008; P. A. Moore et al., 2021; Robroek et al., 2009). There is a concern, however, that a climate change-mediated increase in the frequency, magnitude, and duration of summer drying and drought (Helbig et al., 2020) will transition peatlands from net carbon sinks to net carbon sources (van der Velde et al., 2021) through increased peat decomposition (Ise et al., 2008) and vegetation moisture stress (P. A. Moore et al., 2021).

While several studies have observed a decline in net carbon dioxide (CO₂) uptake in summers with a low WT (Aurela et al., 2007; Bubier et al., 2003; Fortuniak et al., 2021; Moore & Knowles, 1989; Strachan et al., 2016) due to lower gross primary productivity (GPP) (Lund et al., 2012; Strachan et al., 2016) and increased ecosystem respiration (ER) (Aurela et al., 2007; Lund et al., 2012), peatlands are generally considered resilient to moderately dry periods (Waddington et al., 2015). Peatland ecohydrological resilience to drought is maintained through a suite of negative autogenic ecohydrological feedbacks that act to generally maintain a shallow WT and wet conditions in the near-surface peat (Nijp et al., 2017; Waddington et al., 2015) thereby limiting the response of ER and GPP to summer water deficits (e.g., Blodau et al., 2004). However, recent research suggests that peatland ecological resilience to summer water deficits is likely lower in shallow peatlands (e.g., P. A. Moore et al., 2021; Wilkinson et al., 2020) as many of the key autogenic negative feedbacks are greater in deeper peatlands (e.g., Waddington et al., 2015). For example, deeper peatlands may be more resilient to rapid decay during dry periods because of slow pore water turnover in deep layers (Morris & Waddington, 2011) as facilitated through the WT depth—peat decomposition feedback (Waddington et al., 2015). Moreover, small and shallow peatlands, such as those which form in Boreal Shield bedrock depressions, have limited water storage (Devito et al., 1989) and rely on fill-and-spill landscape connectivity with lateral inflow from adjacent upland areas (Spence & Woo, 2003) which is low or absent during dry periods (Devito et al., 1989; Spence & Woo, 2003). As such, these shallow peatlands are especially prone to deep summer WT drawdown where depths can exceed the critical depth (30–50 cm) where productivity of Sphagnum mosses (the dominant species in many bogs and poor fen peatlands) have been shown to dramatically decline (e.g., McCarter & Price, 2014; Thompson & Waddington, 2008) and could undergo a complete loss of the WT altogether for a large portion of the peatland (P. A. Moore et al., 2021). However, most studies of peatland CO₂ exchange have focused on deep and large peatlands (e.g., Humphreys et al., 2006; Lund et al., 2010), while the interannual variability in peatland summer CO₂ exchange and its resilience to summer water deficit in shallow peatlands has not been widely studied. Given that shallow peatlands are abundant in Boreal Shield landscapes and remain understudied in comparison to their deeper counterparts (P. A. Moore et al., 2021), the response of shallow peatland CO₂ fluxes to summer drying represents an important research gap. Here, we quantify the ecosystem scale CO₂ flux response of a small, shallow Boreal Shield peatland to hydrological conditions during the growing season. We hypothesized the peatland would switch to a weak net ecosystem CO₂ sink or a net ecosystem CO₂ source in summers where the WT depths (WTDs) dropped below a critical threshold (e.g., 30–50 cm) and/or when the WT was completely lost for a prolonged period of time in a large portion of the peatland. To test this hypothesis, we assess the growing season net ecosystem exchange (NEE), GPP, and ER response to interannual WT variability throughout five growing seasons between 2016 and 2020 in a shallow Boreal Shield peatland.

2 Methods

2.1 Study Area

The study site is located in a peatland ∼20 km north of Parry Sound, Ontario, Canada in the Georgian Bay Biosphere, Mnidoo Gamii, a UNESCO Biosphere situated within the Robinson-Huron Treaty of 1850 and the Williams Treaty of 1923, and located on Anishinabek territory. The EGB region is located in a rock barrens and peatland landscape of the Boreal Shield ecozone (Markle et al., 2020; P. A. Moore et al., 2019). The region has a cool-temperate and humid climate with a 30-year (1981–2010 climate normal) average air temperature for May–October of 15.7 ± 4.5°C (Dunchurch ∼45 km NE from site; Government of Canada, 2021). The 30-year (1981–2010 climate normal) average cumulative rainfall for May 1 to October 31 is 563 mm (Government of Canada, 2021).

The studied peatland area is a 4,800 m² oligotrophic peatland with a mean depth of 0.49 ± 0.20 m, a maximum depth of 1.33 m (Figure S1 in Supporting Information S1), and has hummock-hollow-lawn microtopography in some areas of the peatland. The vegetation community is comprised of mosses (Polytrichum strictum, Sphagnum palustre, Sphagnum fallax, and Sphagnum cuspidatum) and a variety of shrubs (Rhododendron groenlandicum and Chamaedaphne calyculata), with leaf area index (LAI) ranging from 0.64 ± 0.30 to 0.90 ± 0.50 at the peatland margins and middle, respectively. The site also has a few jack pine (Pinus banksiana) and white pine (Pinus strobus) trees (diameter at breast height = 11.8 ± 6.6 cm) scattered in the middle of the peatland but mostly located on the peatland margins (Markle et al., 2020).

The studied peatland is surrounded predominantly by open granite rock with some moss cushions and lichen mats (hereafter referred to as open rock), coniferous trees on shallow peat and moss-dominated ephemeral wetlands (hereafter referred to as coniferous), and deciduous trees on mineral soil in valleys (hereafter referred to as deciduous) (Figure 1).

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Footprint climatology for May–October overlayed on a land cover map of the study area. The 60% and 70% flux source areas are displayed with a solid and dashed black line, respectively.

2.2 Instrumentation

The eddy covariance (EC) technique was used to measure net CO₂ fluxes (Baldocchi, 2003). High-frequency CO₂ concentrations were measured simultaneously with wind velocities at 10 Hz by an integrated three-dimensional sonic anemometer and open-path infrared gas analyzer (IRGASON, Campbell Scientific, Canada), along with fine wire thermocouple (FW10 Type E, Campbell Scientific, Canada) and recorded on a CR5000 datalogger (Campbell Scientific, Canada). Instruments were mounted 7.8 m above the land surface and oriented into the dominant wind direction (∼285°).

Supporting meteorological variables were measured at the tower and averages/totals were recorded half-hourly. Incoming solar radiation and net radiation were measured using a four-component net radiometer (CNR1, Kipp & Zonen, the Netherlands). Photosynthetic photon flux density (PPFD) was calculated from incoming shortwave radiation using a conversion factor of 2.3, the product of the fraction of incoming radiation that is photosynthetically active radiation (∼0.5) and a unit conversion factor (Knauer et al., 2018). Air temperature and relative humidity were measured with a temperature and relative humidity probe (HMP60, Vaisala, Finland) mounted ∼6 m above the ground surface, while rainfall was measured using a tipping bucket rain gauge several meters from the base of the tower (TE525M, Texas Electronics Inc., USA). WTD was measured every 15 min in a relatively deep surveyed location of the peatland (∼1 m depth) using a Solinst levelogger pressure transducer (Solinst, Georgetown, ON) in 0.05 m diameter slotted PVC wells. Gaps in WT data (∼6% of May–October data over the 5 years) were filled using WT data from near-by peatlands (≤1 km distance) with offsets added to shift the datum.

2.3 EC Data

NEE was measured over 5 years (2016–2020) during the growing season, from May 1 to October 31. The NEE of CO₂ was derived from the sum of turbulent net CO₂ flux and the storage term. Before calculation of half-hourly fluxes, we post-processed the high-frequency EC data. We applied spike detection and removal to the high-frequency data using an algorithm analogous to Vickers and Mahrt (1997) (see P. A. Moore et al., 2013 for additional details). The three-component wind speed data underwent coordinate rotation using the planar fit method (Wilczak et al., 2001). Scalar data was lagged relative to rotated vertical wind speed data to maximize the average covariance. Frequency response corrections (Massman, 2000) and density correction for temperature and humidity compensation (Webb et al., 1980) were applied to half-hourly fluxes.

A detailed flux footprint analysis was completed using the Kljun et al. (2015) flux footprint prediction model in R (R Core Team, 2020). The footprint was overlaid on a land classification map (Figure 1) to determine the proportions of land classes within each half-hour flux footprint. On average, 48% of the measured (uncensored) flux sourced from the study peatland with the second largest contributing area (28%) coming from the surrounding rock barrens (Table S1 in Supporting Information S1). To minimize the influence of deciduous classified areas on measured NEE in this study, we censored all half-hours where the ratio of peatland to deciduous flux footprint contribution was less than 4:1 (Figure S2 in Supporting Information S1). While the peatland is not the sole contributor to the measured fluxes, it is by far the dominant flux source with open rock not contributing to CO₂ fluxes. Our analyses focus on within-season and between-season comparisons of measured fluxes, and not the carbon balance of the study peatland, per se. Consequently, we evaluated the flux source area for systematic variations at diurnal, within-season, and between-season times scales. From the uncensored data, source area varied diurnally with a maximum average peatland source area of 62% (daytime) and minimum of 41% (nighttime). The diurnal change in source area was largely offset by rock barren source area with a minimum of 21% during the day and maximum of 31% at night. However, both the footprint censoring and friction velocity threshold (see below) were biased toward removing nighttime data. Variation in the average monthly peatland source area was small within-season, having a standard deviation ranging from 1% to 3% across years. The average peatland source area was consistent across years ranging from 48% to 53% for uncensored data, and 63% to 67% for the censored data set (Table S1 in Supporting Information S1).

Flux data were further processed using the REddyProc package in R for u* filtering, gap-filling, and flux partitioning (Wutzler et al., 2018). Vapor pressure deficit (VPD) was calculated from relative humidity and air temperature. NEE were filtered by a friction velocity threshold of 0.14 m s⁻¹, derived according to Papale et al. (2006) representing the limit of turbulent conditions whereby below the threshold flux data may have large uncertainties due to poorly developed turbulence (Papale et al., 2006; Wutzler et al., 2018). The proportion of each year's data set identified as gaps was 55% (2016), 56% (2017), 57% (2018), 57% (2019), and 62% (2020). While not mutually exclusive, gaps (% of full May–October study period) were the result of footprint censoring (42%–49%), friction velocity filtering (23%–28%), EC quality assurance (18%–30%), and power/instrument failure (<1%–14%). NEE was gap-filled using the marginal distribution sampling method with incoming solar radiation, VPD, and T_air as input variables (Reichstein et al., 2005; Wutzler et al., 2018).

Flux partitioning of NEE into GPP and ER was completed using daytime-based methods where a modified light-response curve that incorporates the temperature-respiration relationship is used to model NEE (Falge et al., 2001; Lasslop et al., 2010; Lloyd & Taylor, 1994). NEE is the difference between GPP and ER (Equation 1):

$urn:x-wiley:21698953:media:jgrg22395:jgrg22395-math-0001$ (1)

where the sign convention for GPP used herein is negative, meaning that a negative NEE represents net CO₂ uptake.

Half-hour and daily gap-filled data were used to quantify uncertainty associated with random error in the cumulative flux estimates (Liu et al., 2009; Moncrieff et al., 1996; Richardson et al., 2012).

2.4 Analyses and Supplementary Data

Throughout, we colloquially refer to two general time periods: (a) growing season (May 1 to October 31); and (b) summer (June 1 to August 31). Separately, the meteorological growing season was identified as the first day of 7 consecutive days when mean daily T_air was above 5°C (Lund et al., 2010) to the first of 7 consecutive days where mean daily T_air was below 5°C.

To define the end of the carbon uptake period (CUP), a 10-day moving average on daily NEE was calculated (Fu et al., 2017). The start of the CUP was not considered as net CO₂ uptake was already observed at the start of the data set in May. The moving average was then used to define the transition day from sink to source, whereby the end day of the CUP was selected if the moving average was a positive value (NEE changing from negative to positive, indicating net release to the atmosphere) for more than 3 consecutive days.

Random forest models were used to examine the potential nonlinear relations and complex interactions between CO₂ fluxes (GPP, ER, and NEE) and several environmental predictors (air temperature, PPFD, and WT). Random forest models were fitted to daily average values from the May–October period using the MATLAB function treeBagger() (MATLAB, 2021 - The Mathworks, Inc.), where data was pooled across years. Only 10 regression trees were grown per random forest, with a minimum leaf size of five observations, since classification error did not decrease appreciably with a higher number of regression trees.

Light response curves were fitted for each year from the relationship between measured NEE (μmol m⁻² s⁻¹) and incoming PPFD (μmol m⁻² s⁻¹) using the nonlinear least squares analysis in R. NEE values at a PPFD of 0 were excluded. The light response relationship was evaluated based on the rectangular hyperbola equation from Frolking et al. (1998) (Equation 2):

$urn:x-wiley:21698953:media:jgrg22395:jgrg22395-math-0002$ (2)

where 𝜶 represents the quantum yield (initial slope), R₀ is the y-intercept (or nighttime respiration value; μmol m⁻² s⁻¹), GPP_max is the maximum gross photosynthesis (μmol m⁻² s⁻¹). GPP_max and 𝜶 were determined through nonlinear least-square regression (nls) in R.

PET (mm d⁻¹), which was used to estimate the cumulative P-PET, was calculated according to Hargreaves and Samani (1985) as:

$urn:x-wiley:21698953:media:jgrg22395:jgrg22395-math-0003$ (3)

where HS (0.2542 based on unpublished site-specific data) is an empirical coefficient relating daily temperature variation to atmospheric transmissivity, ⍺ is the albedo (0.148 – site-specific data), 𝝀 is the latent heat of vaporization (J kg⁻¹), Q_ex is the top-of-atmosphere incoming radiation (J m⁻² d⁻¹) modeled according to Stull (1988), T_mean, T_max, and T_min are the daily mean, maximum, and minimum temperatures (°C), respectively.

Climate normal data was obtained for the Environment and Climate Change Canada Dunchurch weather station (45°37′N, 79°53′W, 268.2 m elevation; climate ID #6112133).

The results presented herein are based on rejecting half-hourly CO₂ fluxes without a high peatland contribution based on flux footprint analysis (see above) before gap filling and flux partitioning. While the focus of the manuscript is on the measured peatland dynamics contained within the EC measurements, the analyses presented herein were also run on the uncensored data set. For comparison, we provide equivalent versions of select CO₂ flux figures in the Supporting Information S1 based on the analysis done using uncensored data.

3 Results

3.1 Environmental Variables

Mean monthly air temperatures (May–October) did not generally fall outside the normal range (Figure S3 in Supporting Information S1), but where 2016 had a preponderance of both warmer (4 of 6) and drier (4 of 6) months, while 2017 has a preponderance of cooler (5 of 6) and wetter (5 of 6) months relative to the 30-year (1981–2010 climate normal) monthly values. Cumulative rainfall for May 1 to October 31 differed by up to 260 mm between years (i.e., 2017: 689 mm; and 2016: 429 mm) (Table 1), with a mean of 561 mm over the study years compared to the 30-year average of 563 mm). This is generally reflected in the minimum ∑(P−PET) (mm) (2016 = −290 mm; 2017 = −38 mm) but where the greatest estimated summer moisture deficit occurred in 2018 (Table 1). The 3 years experiencing a dry summer (JJA) (2016, 2018, 2019—Table 1) had relatively deep WT drawdown, where WT variability generally followed the pattern of cumulative P-PET (Figure 2).

Table 1. May to October Mean Air Temperature, Water Table, and Daytime (>10 Umol m⁻² s⁻¹) PPFD, Minimum Cumulative P-PET, Total Rainfall and NEE

	2016	2017	2018	2019	2020
Air temperature (°C)	16.0	15.0	16.1	14.4	14.7
Water table (m)	−0.29	−0.15	−0.26	−0.26	−0.14
Daytime PPFD (μmol m⁻² s⁻¹)	885	792	792	795	822
Minimum ∑(P - PET) (mm)	−290	−38	−326	−171	−157
Rainfall (mm)	429	689	513	586	590
Rainfall (JJA) (mm)	228	334	246	196	329
NEE (g C m⁻²) (± s.d.)	−6 (55)	−140 (76)	−31 (55)	−38 (52)	−50 (58)
End day of carbon uptake period (DOY)	July 10 (188)	Sept 14 (257)	July 13 (194)	August 3 (215)	August 22 (235)
Growing season length (days T_air > 5°C)	192	178	179	198	165

Note. Years with pronounced summer WT drawdown (2016, 2018, 2019) are bolded.

In general, the growing season WT position followed a similar pattern between early spring and early June in all years, where the daily WT was less than 0.2 m below the surface. WT position diverges between years after the beginning of June, with the dry years (2016, 2018, and 2019) experiencing a more rapid WT drawdown than other years with WT position dropping below a depth of 0.4 m for several weeks. The WT dropped below a depth of 0.6 m in 2016 and 2018. WT position remained in the upper 0.2 m in the wet years (2017 and 2020) (Figure 2b).

3.2 Growing Season Length and CUP

Growing season length ranged from 165 days in 2020 to 198 days in 2019, and the mean growing season length was 182 days (±12 days (s.d.)) (Table 1). Much of the interannual variability in growing season length was due to a wider range in the growing season start day (DOY 106–134) compared to the end day (DOY 298–305) (Figure S4 in Supporting Information S1).

The transition from daily CO₂ uptake to daily CO₂ emission is described as the end of the CUP (Fu et al., 2017), which occurred as early as mid-July in 2016 and 2018, and as late as mid-September in 2017 (Table 1). CUP end day was moderately correlated to mean growing season WTD (R²_adj = 0.74, p = 0.04) (Figure S5 in Supporting Information S1). Relatively early transitions to average daily CO₂ emission occurred in both 2016 and 2018 (Table 1), corresponding with the years with the lowest growing season rainfall and lowest WTD (Figure S3 in Supporting Information S1). In 2017, the year with the highest growing season rainfall and the second shallowest WTD, the CUP end occurred at the latest (DOY: 257). Both 2016 and 2018, water deficits began to accumulate from the start of May, reaching deficits of around −300 mm (Figure 2). 2017, the year with the latest CUP end date, was the wettest and effectively did not experience any cumulative water deficit (P-PET) from May to October.

3.3 CO₂ Exchange

All growing seasons (May–October) were a net CO₂ carbon sink to near neutral (−0.76 to −0.03 g C m⁻² d⁻¹), with the wettest (2017) and driest (2016) years corresponding with the strongest and weakest sink strengths, respectively (Table 1 and Table S2 in Supporting Information S1). Summer (JJA) NEE exhibited trends similar to the growing season, with interannual differences in NEE (both growing season and JJA) being driven more so by GPP compared to ER (Table S2 in Supporting Information S1). Across study years, both GPP and ER followed similar seasonal patterns, tending to peak in July (Figure 3). Similarity in seasonal patterns amongst GPP and ER is reflected in the daily data, where ER is significantly correlated with GPP (R²_adj = 0.43; p << 0.01), where only a small portion of the variance in the daily GPP-ER relation is explained by WT when aggregated across years (i.e., R²_adj increases to 0.46) (Figure S6 in Supporting Information S1).

Daytime maximum NEE was generally greater in the wet summer of 2017 and 2020 (Figure 4). The early transition to a net daily CO₂ source (Figure 3) is reflected in a lower magnitude of daytime NEE in July of 2016 and 2018 (Figure 4), while daytime net uptake remained greatest (most negative) and relatively constant for June to August in 2017 and 2020. Mean growing season PPFD varied between study years, with lowest incoming PPFD in 2017 and greatest in 2016 (Table 1 and Figure 4), which is a reflection of differences in rainfall frequency (and likely overall cloudiness) in wet versus dry years. Daytime diurnal patterns are broadly similar across years in June, but are characteristically different in July (i.e., following WT drawdown—see Figure 2) between wet (2017 and 2020) and dry (2016, 2018, and 2019) years with a late morning peak in NEE uptake during dry years and depressed afternoon uptake compared to wet years (Figure 4).

3.4 Summer CO₂ Exchange and Environmental Controls

Cumulative summer (JJA) GPP ranged from −324 g C m⁻² in 2016 to −445 g C m⁻² in 2020, with a mean cumulative GPP of −388 ± 54 g C m⁻² for 5 years (Figure 5 and Figure S7 in Supporting Information S1). One-way comparisons with several meteorological variables (e.g., WT, rainfall, air temperature, and PPFD) were not significant predictors (p values of 0.19, 0.55, 0.73, and 0.92, respectively) of interannual variability in growing season (May–October) GPP. However, summer GPP was highly correlated with average WTD over the same period (R² = 0.83; p = 0.03; Figure 6 and Figure S8 in Supporting Information S1), suggesting a more complex relationship between GPP and WTD. For example, outside the summer period, GPP tended to be low (Figure 3), where the divergence of the GPP-ER relation associated with WT was greater at high GPP (Figure S6 in Supporting Information S1). Interannual differences in light response curve characteristics (derived from half-hourly data) were also reflected in a greater GPPmax for the wet summer years of 2017 (−14.7 ± 0.31 μmol m⁻² s⁻¹) and 2020 (−17.7 ± 0.40 μmol m⁻² s⁻¹) (Table S3 in Supporting Information S1). Additionally, when segregated according to WT level and aggregated across years, modeled GPP_max was smaller in magnitude for deep (−10.2 ± 0.26 μmol m⁻² s⁻¹; WT < −0.4 m) compared to shallow WT conditions (−14.3 ± 0.16 μmol m⁻² s⁻¹; WT ≥ 0.4 m). Results of the random forest regression show that both daily air temperature and PPFD were positively related to daily GPP (Figure 7) where the two interacted such that the sensitivity of marginal GPP to PPFD roughly doubled when comparing periods with low and high growing season air temperatures. The partial dependence of daily GPP to WT was more complex, with a peak at approximately a WT of −0.35 m and a generally decreasing GPP on either side of the peak (Figure 7). While the average predicted response of GPP to WT has a humpback shape, during warm periods GPP was less sensitive to WT variation in the wet range (i.e., −0.35 to 0 m) compared to cool periods.

Cumulative summer (JJA) ER ranged from 256 g C m⁻² in 2019 to 326 g C m⁻² in 2020, with a mean cumulative flux of 291 ± 26 g C m⁻² for 5 years (Figure 5). For both the summer period (R² = 0.17, p = 0.49, F = 0.60) and growing season (May–October) (R² = 0.29, p = 0.35, F: 1.22), no significant linear relation was observed between ER and WTD. Similarly, one-way comparisons with other meteorological variables (e.g., rainfall, air temperature, and PPFD) were not significant predictors (p values of 0.55, 0.99, and 0.87, respectively) of interannual variability in growing season ER. However, similar to the daily relationship between GPP and ER (Figure S6 in Supporting Information S1), growing season ER was significantly related to GPP (R² = 0.77, p = 0.05, F = 10.2). For the variables examined, the only coherent relationship for daily ER from the random forest model was air temperature (Figure 7) having a positive relationship. Moreover, the partial dependence of ER to predictor variables tended to be about half the magnitude compared to GPP.

Cumulative summer (JJA) (±SE) NEE was variable between years ranging from 4 (±6) g C m⁻² in 2016 to −114 (±8) g C m⁻² in 2017, with a mean cumulative NEE flux of −47 ± 21 g C m⁻² (Figure 5), where cumulative summer NEE was significantly correlated with average WTD (R² = 0.78, p = 0.046) (Figure 6). For the 2 drier years (2016 and 2018; Table 1 and Figure 2), average summer (JJA) NEE was a source or a smaller CO₂ sink compared to the May–October period. In contrast, the average summer sink strength was higher relative to the May–October period in wetter summers. For the growing season (May–October), one-way comparisons with several meteorological variables (e.g., WT, rainfall, air temperature, and PPFD) suggest that only total rainfall was a significant predictor of interannual variability in growing season NEE (R² = 0.82, p = 0.03, F = 13.6).

For the half-hourly data (Figure S6 in Supporting Information S1), the NEE light response model (Figure S9 in Supporting Information S1) suggests that the vegetation light utilization efficiency (𝜶) was greatest in the wet summer year of 2017 (0.049), ranging between 0.028 and 0.037 among the remaining years (Table S3 in Supporting Information S1). Combined with a fairly narrow range in R₀ amongst years (3.7–4.6 μmol m⁻² s⁻¹), interannual differences in the NEE light response relationship appear to be more strongly driven by GPP dynamics.

While the relationship between NEE and environmental predictors can be complicated by differences in response from GPP and ER, the random forest model results from daily NEE over the growing season show a coherent negative relationship (increasing sink) with increasing PPFD (Figure 7). This is complementary to the half-hourly response of NEE to PPFD, where at a daily timescale higher average PPFD is associated with longer daylight hours and clear sky conditions. The modeled response of daily NEE to air temperature and WT was less coherent. The random forest model suggests that, on average, high air temperatures were associated with a daily CO₂ source (Figure 7). However, due to the seasonality in both air temperature and WT (Figure 2), dry conditions typically occur in mid/late summer when average air temperature is high. The partial dependence of modeled daily NEE to WT shows a switch to a CO₂ source for WT deeper than −0.4 m, and similarly for very wet conditions (Figure 7).

4 Discussion

Our results support the hypothesis that the shallow Boreal Shield peatland could switch to a weak net ecosystem CO₂ sink when WT position dropped below a critical threshold for a prolonged period of a dry summer (Figure 6). Our multi-year data set is the first for a peatland in the Boreal Shield region and highlights the importance of interannual summer water variability to ecosystem-scale summer CO₂ fluxes (Figures 4-6). While the range in cumulative CO₂ fluxes was relatively large over the five seasons, our daily and summer CO₂ fluxes were of similar magnitude to recent studies of deeper peatlands in the boreal regions (e.g., Helbig et al., 2019; Humphreys et al., 2006; Lund et al., 2010; Peichl et al., 2014). Nevertheless, we also found a strong WT control on GPP, similar to Strachan et al. (2016) and Sonnentag et al. (2010), where wet years have greater CO₂ uptake than drier years. Here, we discuss the importance of summer water deficit controls on peatland CO₂ fluxes as it is manifested by the interannual variability in the CUP and water storage dynamics and the climate change implications for shallow peatland carbon balances.

4.1 Growing Season and CO₂ Uptake

Despite the long growing season in 2016 and 2019 (Table 1), cumulative NEE was low because of a high water deficit leading to a deep WT. All else being equal, we would expect a greater WT drawdown for these small rock barren peatlands because of their area-volume relationship and lower specific yield at comparatively shallow depths compared to larger/deeper peatlands. For example, given the bedrock elevation and distribution of peat depth, approximately 40% of the study peatland would lack a WT (i.e., at the periphery) when the WT reached 0.4 m depth in the central portion of the peatland (Figure S1 in Supporting Information S1). In the wet years of 2017 and 2020, the steeper slope of the light response relationship (Table S3 and Figure S6 in Supporting Information S1) indicates NEE is greater in those years at comparable light levels (Figure S6 in Supporting Information S1). This may be an indication that more accessible water from the higher WT has allowed for vegetation to photosynthesize more efficiently, and for longer into the summer and fall, as evidenced by a late end to the CUP (Figure S5 in Supporting Information S1).

The interannual variability in meteorological variables did not considerably affect cumulative ER between years. Consequently, it appears that the differences in summer NEE are more attributable to changes in summer GPP between years at our study site. Nevertheless, with measurement campaigns starting in May, we were unable to capture the beginning of the CUP (e.g., Figure S4 in Supporting Information S1), where interannual differences in cumulative growing season CO₂ fluxes may be sensitive to early and late season dynamics. For example, Helfter et al. (2015) found winter air temperature to be the dominant control on summer net CO₂ uptake through its effect on local phenology with warmer air temperature leading to an earlier net CO₂ uptake period start. In regions with substantial winter snowfall, the interannual difference in late winter meteorological conditions can affect the timing of snowpack loss, vegetation phenology (Kreyling, 2010), and biomass production particularly for mosses (Küttim et al., 2019). While our study found that summer and growing seasons CO₂ uptake was strongly controlled by WT, the relationship between growing season WT and ER was not statistically significant (Figure 6c). While other studies (e.g., Helfter et al., 2015) have shown increasing ER with deeper WT on daily timescales, we did not observe a coherent relationship at a daily timescale based on random forest modeling of ER (Figure 7). As with other studies, daily ER was positively related with air temperature (Figure 7) and GPP (Figure S6 in Supporting Information S1) (e.g., Lund et al., 2010).

The transition day from net CO₂ uptake to net CO₂ emission, or end of the CUP, at this site occurred earlier (August 10 ± 29 days) than the well-studied Mer Bleue peatland, which is located at a similar latitude, which transitioned from CO₂ sink to CO₂ source around October 3 (Lafleur et al., 2003; Roulet et al., 2007). At our site, dry summers transitioned from net CO₂ uptake to net CO₂ emission earlier than wet summers, and earlier than dry years at Mer Bleue peatland (Lafleur et al., 2003). Although the variability of temperature and rainfall anomalies in October was greater than in September, the colder and wetter environmental conditions in September may have triggered an early fall vegetation senescence, slowing the photosynthetic activity of vegetation in the ecosystem and enhancing ER leading to a shift from negative NEE to positive (Järveoja et al., 2018; Piao et al., 2008).

4.2 Water Storage as Control on CO₂ Fluxes

Fill and spill hydrological dynamics are important in the Boreal Shield as many peatlands on the rock barrens landscape are positioned in shallow bedrock basins, relying on precipitation and overland flow from the upland rock watershed to fill peatland storage (Devito et al., 1989; Spence & Woo, 2003). This hydrological connectivity from the upland watershed to the peatland landscape units usually decreases or ceases completely during drier periods of the growing season, leaving the peatland vulnerable to drought conditions (Devito et al., 1989). WT dynamics followed similar trends in all study years, where the WT experienced small fluctuations in response to rain events while remaining within 0.2–0.3 m of the peat surface until the beginning of June and maintaining a daily net CO₂ uptake. In the wet summers (2017 and 2020), the WT remained within 0.2 m of the peat surface throughout the summer. The interannual variability in precipitation (coefficient of variation, CV = ∼20%) was larger than for PET (CV = ∼5%) suggesting large rain events throughout the summer maintained a near-surface WT. In contrast, summer dry periods were experienced in 2016, 2018, and 2019. Despite a long growing season, cumulative NEE was low because a high water deficit led to deep WTs with the WT falling below 0.40 m in the peat profile (a critical threshold) for a period of time in all dry summers. While we observed a strong relationship between mean summer WTD and total summer NEE and GPP (Figure 6), the response of peatland GPP to changes in WTD presented in the literature has been varied. For example, Ratcliffe et al. (2019) observed no change in GPP to WT fluctuations, while several studies have demonstrated a drop in GPP with lower WT (e.g., Humphreys et al., 2014; Strachan et al., 2016; Sulman et al., 2010), and much lower GPP with extreme WT drawdown (e.g., Peichl et al., 2014; Ratcliffe et al., 2019; Sonnentag et al., 2010). Our results are similar to these latter studies as we observed greater diurnal daytime CO₂ uptake, higher GPP_max, and cumulative summer GPP in wet summers, while dry summers had lower CO₂ uptake in comparison to other studies (Sonnentag et al., 2010; Strachan et al., 2016). GPP was greater in years with wet summers, as evidenced by larger maximum net CO₂ uptake at high light levels in the wet years of 2017 and 2020 (Figure 7 and Figure S6 in Supporting Information S1), which also had the largest cumulative summer GPP and ER (Figure 6). Nijp et al. (2014) highlighted the importance of frequent precipitation to maintain and promote high Sphagnum moss productivity. Similarly, Aurela et al. (2007) attributed differences in the interannual variability in light response parameters to water availability in the peatlands as a limiting factor for photosynthesis.

In Sphagnum moss-dominated peatlands (bogs and poor fens) like our study site, a WT in the top 0.3–0.4 m of a peatland provides suitable growing conditions for both mosses and vascular vegetation (Peichl et al., 2014; Rydin & McDonald, 1985; Thompson & Waddington, 2008), and our results support this with larger cumulative GPP during the summers where the WT did not fall below 0.3 m depth. Mosses, which are non-vascular, rely on water supplied to the growing surface via capillary rise from deeper in the peat profile and water storage in the capitula (apical bud) during dry periods (McCarter & Price, 2014; Thompson & Waddington, 2008). When the WT is in the near-surface, Sphagnum moss can maintain high capitula water content while meeting evaporative demand, and thus maintain healthy photosynthesis (Bengtsson et al., 2020; Strack & Price, 2009). When the WTD increases, soil water tension increases potentially leading to a disconnection between the moss capitula and the saturated zone (Gauthier et al., 2018; Goetz & Price, 2015). For example, Sphagnum species with higher stem and branch packing densities (e.g., Sphagnum section Acutifolia) have a greater ability to retain and conduct water (Bengtsson et al., 2020; Elumeeva et al., 2011) giving them the ability to maintain a hydrological connection with the WT at greater WT depths than species with weaker moisture retention and hydraulic conductivity (e.g., Sphagnum section Cuspidata), and thus they possess a greater ability to avoid desiccation in a given environment (McCarter & Price, 2014; Rydin & McDonald, 1985).

Despite a relatively low graminoid and shrub LAI at our study site (0.64–0.90 m² m⁻²), the response of CO₂ exchange we observed was likely mediated by the contributions of both Sphagnum mosses and vascular plants. While many vascular species in peatlands are adapted to wet conditions, saturated conditions can suppress productivity. Nevertheless, the length, magnitude, and timing of WT drawdown will affect how peatland NEE and their contributing components respond (Lund et al., 2012). Early in the growing season when potential ET (PET) is low, saturated conditions are common. Consequently, vascular vegetation might benefit from a lower WT during the early growing season. Later in the growing season when LAI and PET are higher and WT typically lower, exceptionally dry conditions can result in early plant senescence and reduced overall growing season productivity (Sonnentag et al., 2010). WTD thresholds for graminoid and shrub productivity likely differ from Sphagnum mosses. However, during periods when the WTD drops below the rooting depth of graminoid and shrub plants, productivity may be limited due to reduced soil water availability (Lafleur et al., 2005), while the response of GPP for conifer stands growing on organic soils to drought is inconclusive (e.g., Bouriaud et al., 2014; Kljun et al., 2006). Nevertheless, lower peatland WT conditions increase root oxygenation which can increase black spruce productivity (Lieffers & Macdonald, 1990), while the productivity of conifers growing on peatland margins has also shown to be sensitive to interannual variation in precipitation (Bouriaud et al., 2014). However, jack pine which is the predominant species growing on the study site peatland margin and surrounding thin organic soils (which represents ∼10% of our tower footprint) have been shown to be less sensitive to climate variation compared to spruce (Bouriaud et al., 2014). The lower sensitivity of jack pine to WT drawdown compared to other tree species and low relative cover in the flux footprint compared to open peatland suggests that it is likely that the large interannual variability in GPP we have observed is due to more than a tree drought stress response. While a seasonal dependence of CO₂ exchange sensitivity to WT has been demonstrated using multi-annual data from 20 northern peatlands (Helbig et al., 2022), the nature of the response to WT may be threshold-like, where sensitivity can differ between GPP and ER (e.g., Peichl et al., 2014). The response of ER and GPP to WTD between growing seasons at our study site follows the same pattern as the relationship for bogs (Sulman et al., 2010), where WTD in fens was shown to have the opposite effect compared to bogs due to substantially higher shrub and sedge biomass in the fens. Nevertheless, at poor fens with moderate LAI (1.6–2.4 m² m⁻²; P. A. Moore et al., 2013, 2022), growing season moss productivity was more variable along a WT gradient (CV ∼28%) compared to shrubs (CV ∼11%) and graminoids (CV ∼5%) (Chimner et al., 2017) suggesting that Sphagnum CO₂ exchange response can be important even at moderate LAI.

Unlike deep peatlands where a WT remains present at depth throughout the summer, two seasons in our study had a WTD greater than 0.6 m, representing ∼80% of the peatland having no WT, thereby limiting water supply to the surface and resulting in a drop in GPP. A loss of the WT increases the risk of moisture stress for both vascular and non-vascular species as it also limits water supply. While we hypothesized that the study peatland would switch to a net source of CO₂ during a dry summer under conditions where a large portion of the peatland was without a WT, none of our summers were a net source of CO₂. However, examining our summer cumulative NEE flux—mean summer WTD relationship (see Figure 6) suggests that our study peatland would switch to a net summer source of CO₂ with a mean summer WTD of ∼0.4 m.

4.3 Implications for Climate Change

While peatlands are generally resilient to climate-mediated disturbance (Nijp et al., 2017; Waddington et al., 2015), there is concern, that an increase in the frequency, magnitude, and duration of summer drying and drought (Helbig et al., 2020) will transition peatlands from net carbon sinks to net carbon sources through increased peat decomposition and/or vegetation moisture stress (Ise et al., 2008). While precipitation is expected to increase in the Boreal Shield region (Almazroui et al., 2021), greater evaporative water loss from peatlands is expected under a warming climate across the boreal region (Helbig et al., 2020) enhancing the risk of decreasing water availability for peatland ecosystems.

Our study demonstrated that a shallow Boreal Shield peatland displayed tremendous interannual variability in summer ecosystem scale CO₂ fluxes in response to interannual variability in summer water deficit as manifested through the peatland WT position. While several studies conducted in deep peatlands have also observed this strong CO₂ sink in wet summers and a weak CO₂ sink in dry summers function (e.g., Aurela et al., 2007; Bubier et al., 2003; Fortuniak et al., 2021; Moore & Knowles, 1989; Strachan et al., 2016), we argue that shallow peatlands are more sensitive to the same summer water deficit than deep peatlands and, as such, their long-term carbon sink function is likely more vulnerable to future climate change-mediated summer drought.

As mentioned earlier, peat depth plays an important role in the strength of peatland autogenic feedbacks that generally maintain a WT close to the surface (Morris & Waddington, 2011). As such, shallow peatlands are more likely to observe a rapid WT drawdown than deeper peatlands with the same summer water deficit (P. A. Moore et al., 2021; Waddington et al., 2015). Given that many Boreal Shield peatlands have thin peat deposits with limited water storage (e.g., P. A. Moore et al., 2021) and often rely on fill and spill hydrological connectivity in bedrock-dominated watersheds, they are especially vulnerable to drought conditions (Devito et al., 1989) with large portions of the peatland potentially losing the WT completely for several weeks of the summer. Two of the five seasons saw ∼80% of the peatland area without a WT in our study for approximately 1 week at the end of July and early August, and we anticipate loss of WT to occur more frequently with enhanced evapotranspiration in a warming climate (e.g., Helbig et al., 2020).

Increased temperatures with climate change will also likely promote an earlier start to the growing season and extend the length of the growing season, and likely increase peatland GPP in wet years. Given that our study did not show an increase in ER with drying, we anticipate the long-term carbon balance of these shallow peatlands will be controlled by moisture controls on plant productivity (e.g., P. A. Moore et al., 2021). Changes to the moss and vascular vegetation community composition may also occur if there are long-term changes in the WT regime (Breeuwer et al., 2009; T. R. Moore et al., 2002; Strack et al., 2006). This would have direct effects on above and below ground biomass and productivity, thereby changing the total C flux and C storage of the ecosystem (T. R. Moore et al., 2002; Peichl et al., 2018). A shift in Sphagnum community composition may occur with climate change, to Sphagnum species dominating that are more tolerant to deeper WTs and drier surface conditions (Dieleman et al., 2015; McCarter & Price, 2014; P. A. Moore et al., 2021) and a greater proportion of vascular vegetation (Breeuwer et al., 2009; Fenner et al., 2007). A persistently low WT may have different effects across microform types, where hummocks may experience a reduction of Sphagnum moss cover and surface moisture as well as enhanced respiration, while the lowering of the WT at hollows increased Sphagnum and vascular vegetation cover at hollows (Strack et al., 2006).

Finally, moisture stress will not only decrease the productivity of the peatland vegetation community (Bengtsson et al., 2020; Breeuwer et al., 2009) but a deep WT also increases the likelihood of wildfire ignition as a result of the disruption of hydroclimatic feedbacks in peat (Turetsky et al., 2015; Waddington et al., 2015). Shallow peatlands in the same region as our study site were found to be more susceptible to deep burning (Wilkinson et al., 2020). Collectively these shallow peatlands are vulnerable to summer carbon loss in a changing climate.

5 Conclusion

The relationship between CO₂ exchange and WTD is especially sensitive in the Boreal Shield landscape as peatlands are often located in shallow bedrock basins and mainly rely on precipitation inputs and overland flow to maintain water storage function. This connection between CO₂ exchange processes and peatland water availability suggests that peatlands in the Canadian Boreal Shield region may be vulnerable to losing their CO₂ uptake capacity with climate change due to changes in precipitation frequency, greater increases in air temperature, and water loss through greater evapotranspiration. Our results suggest that the peatland could become a net summer source of CO₂ with a mean summer WTD over ∼0.4 m or greater. We encourage a coupled water balance and carbon flux modeling study to examine future climate change scenarios.

Acknowledgments

The research published in this paper is part of the Boreal Water Futures project, which is funded by the Global Water Futures program of the Canada First Research Excellence Fund. The funding for this research was also provided by an NSERC Discovery Grant to JMW. The authors thank Ben Didemus, Max Lukenbach, Alex Furukawa, Cam McCann, Alanna Smolarz, and Hope Freeman for assistance with fieldwork. The authors thank Emma Sherwood and Craig Allison for GIS assistance. The comments of two anonymous reviewers greatly improved this manuscript and the authors thank them for their thoughtful suggestions.

Open Research

Data Availability Statement

The data set used in this publication is available for download from the following site: https://doi.org/10.5281/zenodo.6395387 (McDonald et al., 2022).

Supporting Information

References

Almazroui, M., Islam, M. N., Saeed, F., Saeed, S., Ismail, M., Ehsan, M. A., et al. (2021). Projected changes in temperature and precipitation over the United States, Central America, and the Caribbean in CMIP6 GCMs. Earth Systems and Environment, 5, 1–24. https://doi.org/10.1007/s41748-021-00199-5
Aurela, M., Riutta, T., Laurila, T., Tuovinen, J., Vesala, T., Tuittila, E., et al. (2007). CO₂ exchange of a sedge fen in southern Finland-the impact of a drought period. Tellus B: Chemical and Physical Meteorology, 59(5), 826–837. https://doi.org/10.1111/j.1600-0889.2007.00309.x
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(4), 479–492. https://doi.org/10.1046/j.1365-2486.2003.00629.x
Bengtsson, F., Granath, G., Cronberg, N., & Rydin, H. (2020). Mechanisms behind species-specific water economy responses to water level drawdown in peat mosses. Annals of Botany, 126(2), 219–230. https://doi.org/10.1093/aob/mcaa033
Benscoter, B. W., Vitt, D. H., & Wieder, R. K. (2005). Association of postfire peat accumulation and microtopography in boreal bogs. Canadian Journal of Forest Research, 35(9), 2188–2193. https://doi.org/10.1139/x05-115
Blodau, C., Basiliko, N., & Moore, T. R. (2004). Carbon turnover in peatland mesocosms exposed to different water table levels. Biogeochemistry, 67(3), 331–351. https://doi.org/10.1023/B:BIOG.0000015788.30164.e2
Bouriaud, O., Frank, D., & Bhatti, J. S. (2014). Assessing the influence of climate–water table interactions on jack pine and black spruce productivity in western central Canada. Écoscience, 21(3–4), 315–326. https://doi.org/10.2980/21-(3-4)-3707
Breeuwer, A., Robroek, B. J. M., Limpens, J., Heijmans, M. M. P. D., Schouten, M. G. C., & Berendse, F. (2009). Decreased summer water table depth affects peatland vegetation. Basic and Applied Ecology, 10(4), 330–339. https://doi.org/10.1016/j.baae.2008.05.005
Bubier, J. L., Bhatia, G., Moore, T. R., Roulet, N. T., & Lafleur, P. M. (2003). Spatial and Temporal variability in growing-season net ecosystem carbon dioxide exchange at a large peatland in Ontario, Canada. Ecosystems, 6, 353–367. https://doi.org/10.1007/s10021-003-0125-0
Chimner, R. A., Pypker, T. G., Hribljan, J. A., Moore, P. A., & Waddington, J. M. (2017). Multi-decadal changes in water table levels alter peatland carbon cycling. Ecosystems, 20(5), 1042–1057. https://doi.org/10.1007/s10021-016-0092-x
Clymo, R. S. (1984). The limits to peat bog growth. Philosophical Transactions of the Royal Society B: Biological Sciences, 303(1117), 605–654. https://doi.org/10.1098/rstb.1984.0002
Clymo, R. S. (1987). The ecology of peatlands. Science Progress, 71, 593–614.
Devito, K. J., Dillon, P. J., & Lazerte, B. D. (1989). Phosphorus and nitrogen retention in five Precambrian shield wetlands. Biogeochemistry, 8(3), 185–204. https://doi.org/10.1007/BF00002888
Dieleman, C. M., Branfireun, B. A., McLaughlin, J. W., & Lindo, Z. (2015). Climate change drives a shift in peatland ecosystem plant community: Implications for ecosystem function and stability. Global Change Biology, 21(1), 388–395. https://doi.org/10.1111/gcb.12643
Elumeeva, T. G., Soudzilovskaia, N. A., During, H. J., & Cornelissen, J. H. (2011). The importance of colony structure versus shoot morphology for the water balance of 22 subarctic bryophyte species. Journal of Vegetation Science, 22(1), 152–164. https://doi.org/10.1111/j.1654-1103.2010.01237.x
Environment Canada. (2021). Daily data report for Parry Sound CCG [data file]. Retrieved from https://climate.weather.gc.ca/climate_data/daily_data_e.html?StationID=32128
Falge, E., Baldocchi, D., Olsen, R., Anthoni, P., Aubinet, M., Bernhofer, C., et al. (2001). Gap filling strategies for defensible annual sums of net ecosystem exchange. Agricultural and Forest Meteorology, 107(1), 43–69. https://doi.org/10.1016/S0168-1923(00)00225-2
Fenner, N., Ostle, N. J., McNamara, N., Sparks, T., Harmens, H., Reynolds, B., & Freeman, C. (2007). Elevated CO₂ effects on peatland plant community carbon dynamics and DOC production. Ecosystems, 10(4), 635–657. https://doi.org/10.1007/s10021-007-9051-x
Fortuniak, K., Pawlak, W., Siedlecki, M., Chambers, S., & Bednorz, L. (2021). Temperate mire fluctuations from carbon sink to carbon source following changes in water table. Science of the Total Environment, 756, 144071. https://doi.org/10.1016/j.scitotenv.2020.144071
Frolking, S. E., Bubier, J. L., Moore, T. R., Ball, T., Bellisario, L. M., Bhardwaj, A., et al. (1998). Relationship between ecosystem productivity and photosynthetically active radiation for northern peatlands. Global Biogeochemical Cycles, 12(1), 115–126. https://doi.org/10.1029/97GB03367
Frolking, S. E., Roulet, N., & Fuglestvedt, J. (2006). How northern peatlands influence the Earth's radiative budget: Sustained methane emission versus sustained carbon sequestration. Journal of Geophysical Research: Biogeosciences, 111(G1), G01008. https://doi.org/10.1029/2005JG000091
Fu, Z., Stoy, P. C., Luo, Y., Chen, J., Sun, J., Montagnani, L., et al. (2017). Climate controls over the net carbon uptake period and amplitude of net ecosystem production in temperate and boreal ecosystems. Agricultural and Forest Meteorology, 243, 9–18. https://doi.org/10.1016/j.agrformet.2017.05.009
Gauthier, T. L. J., McCarter, C. P. R., & Price, J. S. (2018). The effect of compression on Sphagnum hydrophysical properties: Implications for increasing hydrological connectivity in restored cutover peatlands. Ecohydrology, 11(8), e2020. https://doi.org/10.1002/eco.2020
Goetz, J. D., & Price, J. S. (2015). Role of morphological structure and layering of Sphagnum and Tomenthypnum mosses on moss productivity and evaporation rates. Canadian Journal of Soil Science, 95(2), 109–124. https://doi.org/10.4141/cjss-2014-092
Gorham, E. (1991). Northern peatlands: Role in the carbon cycle and probable response to climatic warming. Ecological Applications, 1(2), 182–195. https://doi.org/10.2307/1941811
Government of Canada. (2021). Canadian climate normals 1981-2010 [Dunchurch]. Retrieved from https://climate.weather.gc.ca/climate_normals/index_e.html
Hájek, T., & Beckett, R. P. (2008). Effect of water content components on desiccation and recovery in Sphagnum mosses. Annals of Botany, 101(1), 165–173. https://doi.org/10.1093/aob/mcm287
Hargreaves, G. H., & Samani, Z. A. (1985). Reference crop evapotranspiration from temperature. Applied Engineering in Agriculture, 1(2), 96–99. https://doi.org/10.13031/2013.26773
Helbig, M., Humphreys, E. R., & Todd, A. (2019). Contrasting temperature sensitivities of CO₂ exchange in peatlands of the Hudson Bay Lowlands, Canada. Journal of Geophysical Research: Biogeosciences, 124(7), 2126–2143. https://doi.org/10.1029/2019JG005090
Helbig, M., Waddington, J. M., Alekseychik, P., Amiro, B. D., Aurela, M., Barr, A. G., et al. (2020). Increasing contribution of peatlands to boreal evapotranspiration in a warming climate. Nature Climate Change, 10(6), 555–560. https://doi.org/10.1038/s41558-020-0763-7
Helbig, M., Živković, T., Alekseychik, P., Aurela, M., El-Madany, T. S., Euskirchen, E. S., et al. (2022). Warming response of peatland CO₂ sink is sensitive to seasonality in warming trends. Nature Climate Change, 12(8), 743–749. https://doi.org/10.1038/s41558-022-01428-z
Helfter, C., Campbell, C., Dinsmore, K. J., Drewer, J., Coyle, M., Anderson, M., et al. (2015). Drivers of long-term variability in CO₂ net ecosystem exchange in a temperate peatland. Biogeosciences, 12(6), 1799–1811. https://doi.org/10.5194/bg-12-1799-2015
Humphreys, E. R., Charron, C., Brown, M., & Jones, R. (2014). Two bogs in the Canadian Hudson Bay Lowlands and a temperate bog reveal similar annual net ecosystem exchange of CO₂. Arctic Antarctic and Alpine Research, 46(1), 103–113. https://doi.org/10.1657/1938-4246.46.1.103
Humphreys, E. R., Lafleur, P. M., Flanagan, L. B., Hedstrom, N., Syed, K. H., Glenn, A. J., & Granger, R. (2006). Summer carbon dioxide and water vapour fluxes across a range of northern peatlands. Journal of Geophysical Research, 111, G04011. https://doi.org/10.1029/2005JG000111
Ise, T., Dunn, A. L., Wofsy, S. C., & Moorcroft, P. R. (2008). High sensitivity of peat decomposition to climate change through water-table feedback. Nature Geoscience, 1(11), 763–766. https://doi.org/10.1038/ngeo331
Järveoja, J., Nilsson, M. B., Gažovič, M., Crill, P. M., & Peichl, M. (2018). Partitioning of net CO₂ exchange using an automated chamber system reveals plant phenology as key control of production and respiration fluxes in a boreal peatland. Global Change Biology, 24(8), 3436–3451. https://doi.org/10.1111/gcb.14292
Kljun, N., Black, T. A., Griffis, T. J., Barr, A. G., Gaumont-Guay, D., Morgenstern, K., et al. (2006). Response of net ecosystem productivity of three boreal forest stands to drought. Ecosystems, 10(6), 1039–1055. https://doi.org/10.1007/s10021-007-9088-x
Kljun, N., Calanca, P., Rotach, M. W., & Schmid, H. P. (2015). A simple two-dimensional parameterisation for Flux Footprint Prediction (FFP). Geoscientific Model Development, 8(11), 3695–3713. https://doi.org/10.5194/gmd-8-3695-2015
Knauer, J., El-Madany, T. S., Zaehle, S., & Migliavacca, M. (2018). Bigleaf—An R package for the calculation of physical and physiological ecosystem properties from eddy covariance data. PLoS One, 13(8), e0201114. https://doi.org/10.1371/journal.pone.0201114
Kreyling, J. (2010). Winter climate change: A critical factor for temperate vegetation performance. Ecology, 91(7), 1939–1948. https://doi.org/10.1890/09-1160.1
Küttim, M., Laine, A. M., Küttim, L., Ilomets, M., & Robroek, B. J. (2019). Winter climate change increases physiological stress in calcareous fen bryophytes. Science of the Total Environment, 695, 133867. https://doi.org/10.1016/j.scitotenv.2019.133867
Lafleur, P. M., Hember, R. A., Admiral, S. W., & Roulet, N. T. (2005). Annual and seasonal variability in evapotranspiration and water table at a shrub-covered bog in southern Ontario, Canada. Hydrological Processes, 19(18), 3533–3550. https://doi.org/10.1002/hyp.5842
Lafleur, P. M., Roulet, N. T., Bubier, J. L., Frolking, S., & Moore, T. R. (2003). Interannual variability in the peatland-atmosphere carbon dioxide exchange at an ombrotrophic bog. Global Biogeochemical Cycles, 17(2), 1–14. https://doi.org/10.1029/2002gb001983
Lasslop, G., Reichstein, M., Papale, D., Richardson, A. D., Arneth, A., Barr, A., et al. (2010). Separation of net ecosystem exchange into assimilation and respiration using a light response curve approach: Critical issues and global evaluation. Global Change Biology, 16(1), 187–208. https://doi.org/10.1111/j.1365-2486.2009.02041.x
Lieffers, V. J., & Macdonald, S. E. (1990). Growth and foliar nutrient status of black spruce and tamarack in relation to depth of water table in some Alberta peatlands. Canadian Journal of Forest Research, 20(6), 805–809. https://doi.org/10.1139/x90-106
Liu, M., He, H., Yu, G., Luo, Y., Sun, X., & Wang, H. (2009). Uncertainty analysis of CO₂ flux components in sub-tropical evergreen coniferous plantation. Science in China - Series D: Earth Sciences, 52(2), 257–268. https://doi.org/10.1007/s11430-009-0010-6
Lloyd, J., & Taylor, J. A. (1994). On the temperature dependence of soil respiration. Functional Ecology, 8(3), 315–323. https://doi.org/10.2307/2389824
Lund, M., Christensen, T. R., Lindroth, A., & Schubert, P. (2012). Effects of drought conditions on the carbon dioxide dynamics in a temperate peatland. Environmental Research Letters, 7(4), 045704. https://doi.org/10.1088/1748-9326/7/4/045704
Lund, M., Lafleur, P. M., Roulet, N. T., Lindroth, A., Christensen, T. R., Aurela, M., et al. (2010). Variability in exchange of CO₂ across 12 northern peatland and tundra sites. Global Change Biology, 16, 2436–2448. https://doi.org/10.1111/j.1365-2486.2009.02104.x
Markle, C. E., North, T. D., Harris, L. I., Moore, P. A., & Waddington, J. M. (2020). Spatial heterogeneity of surface topography in peatlands: Assessing overwintering habitat availability for the eastern massasauga rattlesnake. Wetlands, 40(6), 2337–2349. https://doi.org/10.1007/s13157-020-01378-2
Massman, W. J. (2000). A simple method for estimating frequency response corrections for eddy covariance systems. Agricultural and Forest Meteorology, 104(3), 185–198. https://doi.org/10.1016/S0168-1923(00)00164-7
MATLAB. (2021). Version 9.10.0 (R2021a). The MathWorks.
McCarter, C. P. R., & Price, J. S. (2014). Ecohydrology of Sphagnum moss hummocks: Mechanisms of capitula water supply and simulated effects of evaporation. Ecohydrology, 7(1), 33–44. https://doi.org/10.1002/eco.1313
McDonald, R. M., Moore, P. A., Helbig, M., & Waddington, J. M. (2022). Reduced net CO₂ uptake during dry summers in a Boreal Shield peatland [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.6395387
Moncrieff, J. B., Malhi, Y., & Leuning, R. (1996). The propagation of errors in long-term measurements of land-atmosphere fluxes of carbon and water. Global Change Biology, 2(3), 231–240. https://doi.org/10.1111/j.1365-2486.1996.tb00075.x
Moore, P. A., Didemus, B. D., Furukawa, A. K., & Waddington, J. M. (2021). Peat depth as a control on Sphagnum moisture stress during seasonal drought. Hydrological Processes, 35(4), e14117. https://doi.org/10.1002/hyp.14117
Moore, P. A., Pypker, T. G., Hribljan, J. A., Chimner, R. A., & Waddington, J. M. (2022). Examining the peatland shrubification-evapotranspiration feedback following multi-decadal water table manipulation. Hydrological Processes, 36(11), e14719. https://doi.org/10.1002/hyp.14719
Moore, P. A., Pypker, T. G., & Waddington, J. M. (2013). Effect of long-term water table manipulation on peatland evapotranspiration. Agricultural and Forest Meteorology, 178, 106–119. https://doi.org/10.1016/j.agrformet.2013.04.013
Moore, P. A., Smolarz, A. G., Markle, C. E., & Waddington, J. M. (2019). Hydrological and thermal properties of moss and lichen species on rock barrens: Implications for turtle nesting habitat. Ecohydrology, 12(2), e2057. https://doi.org/10.1002/eco.2057
Moore, T. R., Bubier, J. L., Frolking, S. E., Lafleur, P. M., & Roulet, N. T. (2002). Plant biomass and production and CO₂ exchange in an ombrotrophic bog. Journal of Ecology, 90(1), 25–36. https://doi.org/10.1046/j.0022-0477.2001.00633.x
Moore, T. R., & Knowles, R. (1989). The influence of water table levels on methane and carbon dioxide emissions from peatland soils. Canadian Journal of Soil Science, 69(1), 33–38. https://doi.org/10.4141/cjss89-004
Morris, P. J., & Waddington, J. M. (2011). Groundwater residence time distributions in peatlands: Implications for peat decomposition and accumulation. Water Resources Research, 47(2), W02511. https://doi.org/10.1029/2010WR009492
Nijp, J. J., Limpens, J., Metselaar, K., van der Zee, S. E. A. T. M., Berendse, F., & Robroek, B. J. M. (2014). Can frequent precipitation moderate the impact of drought on peatmoss carbon uptake in northern peatlands? New Phytologist, 203(1), 70–80. https://doi.org/10.1111/nph.12792
Nijp, J. J., Metselaar, K., Limpens, J., Teutschbein, C., Peichl, M., Nilsson, M. B., et al. (2017). Including hydrological self-regulating processes in peatland models: Effects on peat moss drought projections. Science of the Total Environment, 580, 1389–1400. https://doi.org/10.1016/j.scitotenv.2016.12.104
Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W., et al. (2006). Towards a standardized processing of net ecosystem exchange measured with eddy covariance technique: Algorithms and uncertainty estimation. Biogeosciences, 3(4), 571–583. https://doi.org/10.5194/bg-3-571-2006
Peichl, M., Gažovič, M., Vermeij, I., De Goede, E., Sonnentag, O., Limpens, J., & Nilsson, M. B. (2018). Peatland vegetation composition and phenology drive the seasonal trajectory of maximum gross primary production. Scientific Reports, 8, 1–11. https://doi.org/10.1038/s41598-018-26147-4
Peichl, M., Öquist, M., Löfvenius, M. O., Llstedt, U., Sagerfors, J., Grelle, A., et al. (2014). A 12-year record reveals pre-growing season temperature and water table level threshold effects on the net carbon dioxide exchange in a boreal fen. Environmental Research Letters, 9(5), 055006. https://doi.org/10.1088/1748-9326/9/5/055006
Piao, S., Ciais, P., Friedlingstein, P., Peylin, P., Reichstein, M., Luyssaert, S., et al. (2008). Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature, 451(7174), 49–52. https://doi.org/10.1038/nature06444
R Core Team. (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing. Retrieved from https://www.R-project.org/
Ratcliffe, J. L., Campbell, D. I., Clarkson, B. R., Wall, A. M., & Schipper, L. A. (2019). Water table fluctuations control CO₂ exchange in wet and dry bogs through different mechanisms. Science of the Total Environment, 655, 1037–1046. https://doi.org/10.1016/j.scitotenv.2018.11.151
Reichstein, M., Falge, E., Baldocchi, D., Papale, D., Aubinet, M., Berbigier, P., et al. (2005). On the separation of net ecosystem exchange into assimilation and ecosystem respiration: Review and improved algorithm. Global Change Biology, 11(9), 1424–1439. https://doi.org/10.1111/j.1365-2486.2005.001002.x
Richardson, A. D., Aubinet, M., Barr, A. G., Hollinger, D. Y., Ibrom, A., Lasslop, G., & Reichstein, M. (2012). Uncertainty quantification. In Eddy covariance (pp. 173–209). Springer.
Robroek, B. J. M., van Ruijven, J., Schouten, M. G. C., Breeuwer, A., Crushell, P. H., Berendse, F., & Limpens, J. (2009). Sphagnum re-introduction in degraded peatlands: The effects of aggregation, species identity and water table. Basic and Applied Ecology, 10(8), 697–706. https://doi.org/10.1016/j.baae.2009.04.005
Roulet, N. T., Lafleur, P. M., Richard, P. J. H., Moore, T. R., Humphreys, E. R., & Bubier, J. (2007). Contemporary carbon balance and late Holocene carbon accumulation in a northern peatland. Global Change Biology, 13(2), 397–411. https://doi.org/10.1111/j.1365-2486.2006.01292.x
Rydin, H., & McDonald, A. J. S. (1985). Tolerance of Sphagnum to water level. Journal of Bryology, 13(4), 572–578. https://doi.org/10.1179/jbr.1985.13.4.571
Sonnentag, O., van der Kamp, G., Barr, A. G., & Chen, J. M. (2010). On the relationship between water table depth and water vapour and carbon dioxide fluxes in a minerotrophic fen. Global Change Biology, 16(6), 1762–1776. https://doi.org/10.1111/j.1365-2486.2009.02032.x
Spence, C., & Woo, M. (2003). Hydrology of subarctic Canadian shield: Soil-filled valleys. Journal of Hydrology, 279(1–4), 151–166. https://doi.org/10.1016/S0022-1694(03)00175-6
Strachan, I. B., Pelletier, L., & Bonneville, M. (2016). Inter-annual variability in water table depth controls net ecosystem carbon dioxide exchange in a boreal bog. Biogeochemistry, 127(1), 99–111. https://doi.org/10.1007/s10533-015-0170-8
Strack, M., & Price, J. S. (2009). Moisture controls on carbon dioxide dynamics of peat-Sphagnum monoliths. Ecohydrology, 2(1), 34–41. https://doi.org/10.1002/eco.36
Strack, M., Waddington, J. M., Rochefort, L., & Tuitilla, E.-S. (2006). Response of vegetation and net ecosystem carbon dioxide exchange at different peatland microforms following water table drawdown. Journal of Geophysical Research, 111(G2), G02006. https://doi.org/10.1029/2005JG000145
Stull, R. B. (1988). An introduction to boundary layer meteorology. Kluwer Academic Publishers. https://doi.org/10.1007/978-94-009-3027-8
Sulman, B. N., Desai, A. R., Saliendra, N. Z., Lafleur, P. M., Flanagan, L. B., Sonnentag, O., et al. (2010). CO₂ fluxes at northern fens and bogs have opposite responses to inter-annual fluctuations in water table. Geophysical Research Letters, 37(19). https://doi.org/10.1029/2010GL044018
Thompson, D. K., & Waddington, J. M. (2008). Sphagnum under pressure: Towards an ecohydrological approach to examining Sphagnum productivity. Ecohydrology, 1(4), 299–308. https://doi.org/10.1002/eco.31
Turetsky, M. R., Benscoter, B., Page, S., Rein, G., van der Werf, G. R., & Watts, A. (2015). Global vulnerability of peatlands to fire and carbon loss. Nature Geoscience, 8(1), 11–14. https://doi.org/10.1038/ngeo2325
van der Velde, Y., Temme, A. J. A. M., Nijp, J. J., Braakhekke, M. C., van Voorn, G. A. K., Dekker, S. C., et al. (2021). Emerging forest-peatland bistability and resilience of European peatland carbon stores. Proceedings of the National Academy of Sciences of the United States of America, 118(38), e2101742118. https://doi.org/10.1073/pnas.2101742118
Vickers, D., & Mahrt, L. (1997). Quality control and flux sampling problems for tower and aircraft data. Journal of Atmospheric and Oceanic Technology, 14(3), 512–526. https://doi.org/10.1175/1520-0426(1997)014<0512:QCAFSP>2.0.CO;2
Waddington, J. M., Morris, P. J., Kettridge, N., Granath, G., Thompson, D. K., & Moore, P. A. (2015). Hydrological feedbacks in northern peatlands. Ecohydrology, 8(1), 113–127. https://doi.org/10.1002/eco.1493
Webb, E. K., Pearman, G. I., & Leuning, R. (1980). Correction of flux measurements for density effects due to heat and water vapour transfer. Quarterly Journal of the Royal Meteorological Society, 106(447), 85–100. https://doi.org/10.1002/qj.49710644707
Wilczak, J. M., Oncley, S. P., & Stage, S. A. (2001). Sonic anemometer tilt correction algorithms. Boundary-Layer Meteorology, 99(1), 127–150. https://doi.org/10.1023/A:1018966204465
Wilkinson, S. L., Tekatch, A. M., Markle, C. E., Moore, P. A., & Waddington, J. M. (2020). Shallow peat is most vulnerable to high peat burn severity during wildfire. Environmental Research Letters, 15(10), 104032. https://doi.org/10.1088/1748-9326/aba7e8
Wutzler, T., Lucas-Moffat, A., Migliavacca, M., Knauer, J., Sickel, K., Šigut, L., et al. (2018). Basic and extensible post-processing of eddy covariance flux data with REddyProc. Biogeosciences, 15, 5015–5030. https://doi.org/10.5194/bg-2018-56
Xu, J., Morris, P. J., Liu, J., & Holden, J. (2018). PEATMAP: Refining estimates of global peatland distribution based on a meta-analysis. Catena, 160, 134–140. https://doi.org/10.1016/j.catena.2017.09.010
Yu, Z., Loisel, J., Brosseau, D. P., Beilman, D. W., & Hunt, S. J. (2010). Global peatland dynamics since the Last Glacial Maximum. Geophysical Research Letters, 37(13), LI3402. https://doi.org/10.1029/2010GL043584

Volume128, Issue2

February 2023

e2022JG006923

Reduced Net CO₂ Uptake During Dry Summers in a Boreal Shield Peatland

Abstract

Plain Language Summary

Key Points

1 Introduction