Soil temperature synchronisation improves representation of diel variability of ecosystem respiration in Sphagnum peatlands

https://doi.org/10.1016/j.agrformet.2016.03.021Get rights and content

Highlights

  • Quantification of the time-delay between ER fluxes and temperatures in peatlands.

  • Time-delay estimations are higher than those found in mineral soils.

  • Synchronised data can improves ER daily representation in sphagnum peatlands.

Abstract

The temperature dependence of Ecosystem Respiration (ER) is often assessed based on the temperature of one specific layer. Air temperature or temperatures in the first ten centimetres of the soil profile are the most frequently used temperatures in models. However, previous studies showed that the relationship between ER and temperature is depth dependent, making depth selection for temperature measurements an important issue, especially at short time-scales. The present study explores one possible way to assess this relationship by synchronising the ER and temperature signals and to test if the relationship between ER and temperature differs between daytime and nighttime. To do so, ER measurements were undertaken in 2013 in four Sphagnum-peatlands across France using the closed chamber method. The ER fluxes were measured hourly during 72 h in each of four replicates in each site. Synchronisations between ER and T signal were determined for each depth (from surface to 30 cm depth) by selecting the time-delay leading to the best correlation between ER and soil temperatures and ER was then modelled. Our results showed that: (i) the delay between ER and soil temperature is greater in peat than in mineral soils; (ii) at a daily time-scale synchronisation can improve the model representation using soil temperatures.

Introduction

At a global scale, Ecosystem Respiration (ER) and photosynthesis are the largest carbon (C) fluxes between the atmosphere and the biosphere, accounting for 98 and 123 PgC yr−1, respectively (Bond-Lamberty and Thomson, 2010, Beer et al., 2010). By contrast the fossil fuel and cement production flux is one order of magnitude lower, at 7.8 PgC yr−1 (Ciais et al., 2014). Consequently, even small variations in the ecosystem fluxes may result in substantial changes in net C storage dynamics. This can have a significant effect on the global C budget, in particular on the atmospheric C concentration. The C stock in natural ecosystems is divided into two pools: vegetation, which contains 450–650 PgC, and the soil which contains 1500–2400 PgC (Ciais et al., 2014, Carvalhais et al., 2014). Across the world, the soil organic C (SOC) pool is spatially heterogeneous in terms of source and physical conditions, leading to variable storage rates between ecosystem types. Peatlands are efficient C storage ecosystems. They cover only 3% of the global terrestrial area, but contain from 270 to 455 PgC as SOC, i.e. from 10 to 30% of the world’s soil C (Gorham, 1991, Turunen et al., 2002, Limpens et al., 2008). Thus, peatlands are considered as a hot spots for SOC storage, and their evolution under current environmental changes deserves attention.

As in many other terrestrial ecosystems, many factors affect ER variability in peatlands: temperature, soil water content, vegetation, and substrate supply (Luo and Zhou, 2006). All these factors are thought to be affected by global change, with unknown consequences on the C balance (Limpens et al., 2008). More specifically the temperature affect ER directly (biochemical reaction rates are related to temperature) and indirectly (vegetation, and particularly root growth, transport rates) (Luo and Zhou, 2006) and is thus largely utilized to model ER. Different temperature may be used: either air (e.g., Bortoluzzi et al., 2006), or soil temperature. The most commonly used soil temperatures are those at −5 cm (Ballantyne et al., 2014, Görres et al., 2014) and −10 cm (Kim and Verma, 1992, Zhu et al., 2015). In some studies, different depths are used and the selected one depends on the goodness-of-fit (Günther et al., 2014, Zhu et al., 2015). All these studies use the chamber method to measure gas fluxes and even though most studies use −5 cm soil temperature, no clear consensus exists.

The relationship between ER and temperature is often described using the Q10 indicator, which represents the proportional increase of a reaction rate due to a 10 °C rise in temperature. However, even if the Q10 seems coherent at a global scale (Mahecha et al., 2010), reported values show a significant variability at the ecosystem level (Graf et al., 2008). Because the calculated Q10 are not linked to a single reaction but to multiple processes, numerous issues arise (Davidson et al., 2006). Among them are the time-scale considered (Curiel Yuste et al., 2004), the depth (Graf et al., 2008) and the time-delays between ER and soil temperatures (Phillips et al., 2011).

More specifically Pavelka et al. (2007) and Graf et al. (2008) showed that the relationship between ER and temperature is depth dependent since heat transfer in the soil profile is not instantaneous and leads to a time-delay between the temperature and the ER signals. One way to deal with the time-delays might be to synchronise ER fluxes and temperature measurements according to Pavelka et al. (2007). Another issue is the difference between the daytime and nighttime ER relationship with temperature. Juszczak et al. (2012), for example, showed that there are significant differences between ER modelled with daytime and nighttime data. Assessing these differences may be important when working at a daily timescale and when treating data from eddy-covariance measurements.

Based on these previous studies, we expected that time-delays in Sphagnum-dominated peatlands would be significant, even in the first 10 cm depth and that they would lead to a better description of observed data once taken into account, especially through data synchronisation. To our knowledge no studies have explored the time-delay between ER and soil temperature in peatlands yet. To test these predictions, ER fluxes, during the growing season in 4 Sphagnum-dominated peatlands were measured in 2013. Continuous measurements over 72 h were carried out in each site using static dark chambers. Air and soil temperature were also monitored. Specifically, the relationship between ER and temperature, measured at different depths in peat was studied.

The aim of this study was (i) to highlight any time-delay at the daily timescale between ER and soil temperature at different depths in peatlands (ii) to assess the effect of synchronisation between ER and temperature in the model representation of the diel ER variations.

Section snippets

Study sites

The study was performed on four French Sphagnum-dominated peatlands: Bernadouze (BDZ, Ariège; 3.75 ha, N 42°4809, E 1°2524, 1400 m), Frasne (FRN, Doubs; 98 ha, N 46°4935, E 6°1020, 836 m), Landemarais (LDM, Ille-et-vilaine; 23 ha, N 48°2630, E 1°1054, 154 m), and La Guette (LGT, Cher; 26 ha, N 47°1944, E 2°1704, 145 m). Mean annual air temperatures and annual rainfalls were 6, 7.5, 11, 11 °C, and 1700, 1400, 870, 880 mm for BDZ, FRN, LDM and LGT respectively. During the measurements the water Table level

Air temperature and ER variability

During the period of experiments, mean surface air temperatures were about 14–15 °C for all sites, except for LGT which was 20.8 ± 7.4 °C, (Fig. 2H). The lowest mean temperature and standard deviation were found at BDZ: 14.4 ± 3.3 °C (Fig. 2E). In LDM and FRN, the mean surface air temperatures were respectively 14.9 ± 8.7 °C and 15.0 ± 10.3 °C (Fig. 2F, G). Surface air temperature was the highest in FRN.

At −5 cm depth, BDZ and LGT had lower mean peat temperatures than their air surface counterparts: 14.1 ± 1.5 

ER differences between sites

The ER fluxes calculated in the 4 sites were in the same order of magnitude as those of peatlands found in the literature: Bortoluzzi et al. (2006), found ER values ranging from 2 to 5 μmol m−2 s−1 during the same period as this study, i.e. July to October 2004, as well as Juszczak et al. (2013) with value between 2.6 to 5.4 μmol m−2 s−1 (June to August 2008–2009). In the present study, the models performed poorly in 2 sites, BDZ and LDM. For BDZ, amplitudes of both ER and temperatures were low (Fig.

Conclusions

We showed that the time-delays between ER and soil temperatures in peat soils at different depths are significant on a daily timescale. The signals are shifted approximately 30 min every centimetre in all studied sites, leading to longer time-delay than those found in mineral soils. At this scale the use of synchronised soil temperature, to take into account these time-delays, can improve the model representation of ER particularly in the first 10 cm. Thus the synchronised temperature at the −5

Acknowledgements

The work was funded as part of the Peatland National Observatory Service (Service national d’observation Tourbières, certified by the CNRS/INSU) as the four studied sites are part of this Service. The authors are also indebted to the site managers for permitting access to the studied peatlands. We also acknowledge support from Labex VOLTAIRE (ANR-10-LABX-100-01). Finally we would like to thank Elizabeth Rowley-Jolivet for corrections to the manuscript.

References (30)

  • N. Carvalhais et al.

    Global covariation of carbon turnover times with climate in terrestrial ecosystems

    Nature

    (2014)
  • P. Ciais et al.

    Carbon and other biogeochemical cycles

    Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change

    (2014)
  • J. Curiel Yuste et al.

    Annual Q10 of soil respiration reflects plant phenological patterns as well as temperature sensitivity

    Glob. Change Biol.

    (2004)
  • E.A. Davidson et al.

    On the variability of respiration in terrestrial ecosystems: moving beyond Q10

    Glob. Change Biol.

    (2006)
  • O. Farouki

    Thermal Properties of Soils. Series on Rock and Soil Mechanics

    (1981)
  • Cited by (14)

    • The responses of soil respiration to changed precipitation and increased temperature in desert grassland in northern China

      2021, Journal of Arid Environments
      Citation Excerpt :

      In this case, the increase in temperature will further cause the soil to emit more CO2 into the atmosphere. Therefore, the increase of the SR rate caused by climate change may significantly reduce the carbon storage of high altitude and high latitude ecosystems (D'Angelo et al., 2016). There is no significant interaction between the temperature rising and precipitation increasing and soil parameters, the reason maybe is that short-term experiment has little effect on them, thus it need to do long-term experiment to verify.

    • Modelling and upscaling ecosystem respiration using thermal cameras and UAVs: Application to a peatland during and after a hot drought

      2021, Agricultural and Forest Meteorology
      Citation Excerpt :

      Nevertheless, future modelling approaches could use mixed effects models to account for these issues (Kravchenko 2015). In terms of model accuracy, our results for ERmod_surf (mean R2 = 0.54, mean NMRSE = 47%) are in line with previous work using chamber data in peatland sites and simple empirical ER models based on temperature: R2 between 0.2-0.9 and NRMSE between 15-50% (Acosta et al., 2017; D'Angelo et al., 2016; Juszczak et al., 2013; Mäkiranta et al., 2009). We modelled ER using only temperature data, which is a common approach in the greenhouse gas flux and remote sensing communities (Ai et al., 2018; Reichstein et al., 2005; Reichstein and Beer, 2008).

    • Response of C and N cycles to N fertilization in Sphagnum and Molinia-dominated peat mesocosms

      2019, Journal of Environmental Sciences (China)
      Citation Excerpt :

      GHG measurements were performed with the closed chamber method between once to twice per week during the growing season (April–October 2015 and April–June 2016) and every 2 weeks during winter (November 2015–March 2016). CO2 fluxes were measured during 5 minute using a GMP343 Vaisala probe inserted in a transparent PVC chamber (D'Angelo et al., 2016). A clear chamber was used to measure the net ecosystem exchange (NEE), the balance between GPP (absorption of CO2 by photosynthesis) and ER (release of CO2 into the atmosphere).

    • Production rates of bacterial tetraether lipids and fatty acids in peatland under varying oxygen concentrations

      2017, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      Initially oxic incubations were prepared on a surface sterilized with ethanol, then sealed to prevent escape of labeled 13CO2, which would have compromised estimates of autotrophic production. Given the initial amount of 1.25 mmol O2 in the headspace and typical peat respiration rates of 1.75 μmol m−2 s−1 in Frasne (D’Angelo et al., 2016), the incubations were expected to remain aerobic for at least 22 days, then transition to suboxic and eventually anoxic conditions (Table S1). All samples were incubated at 12 °C, consistent with the mean air temperature at the sampling site in spring and summer (Huguet et al., 2013a), i.e. when the peatland is not snow-covered and when the microbial activity is expected to be maximal (e.g. Blume et al., 2002).

    View all citing articles on Scopus
    View full text