Surface O₂ influx related to soil O₂ profiles in a drained tropical peatland

Abstract

Tropical peatlands are potentially the highest-ranked carbon sources among various types of soil in the world. The O₂ consumption rate is one of the deterministic factors for soil carbon release through aerobic decomposition of soil organic matter in reclaimed tropical peatlands. The present study examined in-situ O₂ influx at the soil surface in relation to below-ground O₂ consumption in a palm oil plantation field on a tropical peatland in Thailand. The surface O₂ influx rate was measured using a closed-chamber method. Below-ground O₂ concentrations were also measured. The O₂ influx rates obtained from three sampling points were 3.06, 3.66 and 7.63 mmol m⁻² h⁻¹, and did not show marked responses to changes in soil temperature. When the surface chambers were kept closed beyond the influx measurement period, the O₂ concentrations in the chambers dropped to different steady-state concentrations even in the two chambers that showed similar surface O₂ influx rates to each other, suggesting a difference in the effective depth range of O₂ consumption. The O₂ concentrations at depths of 5, 10 and 20 cm reached 0.181, 0.131 and 0.070 m³ m⁻³, respectively, at one monitoring point, whereas the concentrations at the other point were 0.194, 0.149 and 0.144 m³ m⁻³, respectively. The drop in O₂ concentrations after the installation of the O₂ sensors into the monitoring depths were rapid and linear over time at the former monitoring point, in contrast to the slower convergent lowering behaviors observed at the other point. The fast linear lowering at a monitoring depth implied the O₂ consumption rate surpassing the diffusive O₂ transport at the depth, suggesting that because of low soil gas diffusivity the depth range of O₂ consumption could be confined to just a shallow portion of the unsaturated zone in a peat soil layer and could make the other deeper portion anaerobic.

Key words:

• carbon dioxide
• decomposition
• oxygen consumption
• soil respiration

Introduction

Soil carbon release associated with the reclamation of tropical peatlands is attracting concern in the context of the regulation of atmospheric carbon. This type of land-use change has resulted in more than 10 Gt of CO₂ release since 1985, and even now soil carbon release continues with a likelihood rate of 632 Mt CO₂ year⁻¹ from 27.1 million ha of South-East Asian peatlands (Hooijer et al. 2006). These values rank tropical peatlands highest in terms of soil carbon release among the various types of soils around the world.

Emission of CO₂ from peatlands is mainly the product of aerobic decomposition of peat soils and, therefore, soil carbon release can be greater at deeper groundwater table levels (Furukawa et al. 2005). The release of soil carbon is enhanced by drainage because lowering of the groundwater table simply thickens the unsaturated zone. However, some studies have suggested that there is a certain threshold groundwater table level that gives the maximum CO₂ efflux, meaning that even if the groundwater table drops below the threshold depth no further increases in CO₂ efflux occur (e.g. Hirano et al. 2009; Page et al. 2009). In addition, in these studies no consistent response of CO₂ emission to groundwater table level could be found in some cases.

The O₂ flux into a peatland may be one of the determining factors for a possible threshold groundwater table depth because no matter how deeply the unsaturated zone could penetrate with increases in drainage depth, the rate of O₂ supply from the soil surface may not necessarily rise to fully meet the O₂ consumption rate in the unsaturated zone. In this situation, all of the O₂ supplied from the soil surface is consumed completely on the way to the groundwater table and no more O₂ can penetrate below a certain depth. Thus, an unsaturated anaerobic layer may occur. Validation of this assumption will contribute to the development of a model to predict the change in the rate of carbon release from drained tropical peatlands and, therefore, examining the O₂ influx and below-ground O₂ consumption in a peatland will increase our understanding of the background mechanisms supporting the model development. The aim of the present study was to examine in-situ O₂ influx at the soil surface and to relate the influx to below-ground O₂ consumption in a drained tropical peatland in Thailand.

Materials and methods

The study was conducted in a palm oil field in Nakhon-Si-Thammarat, Thailand (08°00′ 55.4″N, 100°04′ 09.6″E; Fig. 1) from 24 to 28 August 2009. The palm oil plantation has been run at the study site for 2 years. Open ditches drain groundwater in the planted area and are dug in parallel approximately every 14 m.

The surface O₂ influx rate was measured using closed chambers that were made from stainless tubes with a 50-mm inner diameter and 60-mm length of insertion. The thickness of the chamber head space was 5 mm. The basements of the chambers were inserted into the soil at data sampling points [1], [2] and [3] in Fig. 1 and remained buried throughout the entire field experiment. Galvanic cell sensors (O2-204G, Gastec Corporation, Ayase, Kanagawa, Japan) were used to sense lowering drop in O₂ concentration in the chamber headspace. The O₂ concentration was determined as the partial pressure of O₂ using a linear relationship between the output voltage of the galvanic cell sensor and the O₂ level. Through calibration of the sensors using N₂ gas, it was confirmed that the sensor responded within 1 min to an immediate change in O₂ concentration from the atmospheric level to the O₂-free level. Therefore, the O₂ concentration in the chamber headspace was measured at 1-min intervals. In a pilot study to identify the adequate period of time for evaluating O₂ influx rates, the chambers remained closed for up to 5.5 h. The identified period of time was 10 min, excluding the first 2 min, after the chamber was closed; within this time the lowering rate became constant.

Figure 1 Site description at Nakhon Si Thammarat, Thailand. The three circles (numbers [1], [2] and [3]) denote the data sampling points for the surface O2 influx rates. The O2 concentration profiles were examined at points [4] and [5] (□). The soil temperature profile was monitored at point [6] (□).

The profile of the O₂ concentration in the peat layer was also examined. Polyvinyl chloride (PVC) pipes containing the galvanic cell sensors described in Iiyama and Hasegawa (2009) were installed at points [4] and [5] in Fig. 1. After installation, O₂ in the 5 cm³ space in which the sensor in the pipe was in direct contact began diffusing into the surrounding soil. Then, the drop in the O₂ concentration from the atmospheric level was monitored every 20 min. The lowering rate of the recorded O₂ concentration should depend on both gas diffusivity and the O₂ consumption rate of the soil adjacent to the PVC pipes. The monitoring depths were at 5, 10 and 20 cm for each sampling point.

Soil temperatures were monitored at depths of 0, 2, 4, 8, 16, 32 and 45 cm by using cupper-constantan thermocouples at point [6] every 20 min. All galvanic cell sensors and thermocouples were regulated by a data logger (CR10X; Campbell, Campbell Scientific, Logan, UT, USA) wired with a multiplexer (AM25T; Campbell).

A soil profile of the study site comprised wooden peat layers sandwiching a clay layer. The thickness of the top wooden peat layer varied spatially, ranging from 10 to 25 cm. The clay layer underlay the top peat layer and continued to a depth of 65 cm, overlying the second peat layer, which was recorded at a depth of at least 130 cm. The groundwater table level was detected in a borehole with a scale once per day from 25 to 28 August 2009; the level was maintained at a depth of 30–31 cm.

Results and discussion

Surface O₂influx rates

In Fig. 2, the triplicate-sampled influx rates were plotted against measurement time. The surface O₂ influx rate at point [2] was consistently higher than the rate at the other two points. The averaged influx rates at points [1], [2] and [3] were 3.66, 7.63 and 3.06 mmol m⁻² h⁻¹, respectively. Weak or no consistent responses to changes in soil temperature were found, implying that spatial variability rather than changes in temperature affected the measured influx rates. In-situ O₂ influx rates in peatlands have not been evaluated as much as other types of soil gas. Haraguchi et al. (2003) examined O₂ influx rates in sphagnum-dominated mires in central Japan and recorded rates ranging from 0.275 to 1.688 mmol m⁻² h⁻¹, with groundwater table depths ranging from 0.5 to 6.3 cm. The O₂ influx rates in the present study were slightly higher than those reported in this study. The deeper groundwater table level and the higher temperature might be responsible for the relatively high O₂ influx rates recorded in the present study. The spatially variable depth to the clay layer, which underlay the top peat layer, might also have affected the results.

Figure 2 Time series of the surface O2 influx at sampling points [1], [2] and [3] with the soil temperatures at depths of 0, 2, 4, 8 and 16 cm on 27 August 2009.

Surface influx rates in relation to the depth range of O₂ consumption

Figure 3 shows the drop in O₂ concentrations in the closed chambers at points [1], [2] and [3]. The order in the degree of slope among the three curves in the first 1 h corresponded well with the order obtained and used to evaluate the O₂ influx rates shown in Fig. 2. However, each curve had a unique lowering behavior, namely “a low influx rate with a high steady-state concentration” at point [1], “a high influx rate with a low steady-state concentration” at point [2] and “a low influx rate with a low steady-state concentration” in thate point [3]. In order to discuss these lowering behaviors, the depth profiles of O₂ concentration was shown in Figure 4.

Figure 4 shows the time-series data obtained from the O₂ sensors installed at the three depths at points [4] and [5]. A monitored O₂ concentration that remains steady with time should be equivalent to that in soil air adjacent to the monitoring pipe. By the elapsed date of 3.75 from the installation, the O₂ concentrations at point [4] dropped to 0.181, 0.131 and 0.070 m³ m⁻³ at depths of 5, 10 and 20 cm, respectively. At point [5], the concentration values reached 0.194, 0.149 and 0.144 m³ m⁻³ at depths of 5, 10 and 20 cm, respectively, higher than at point [4]. The monitored O₂ concentrations declined linearly at point [4] with rates of 0.056 and 0.103 m³ m⁻³ day⁻¹ at depths of 10 and 20 cm, respectively. At point [5] the O₂ concentration at a depth of 10 cm was equivalent to that of soil air by the elapsed date of 3, whereas the O₂ concentration at a depth of 20 cm continuously declined at a rate of 0.039 m³ m⁻³ day⁻¹.

Figure 3 Drop in O2 concentrations in the chambers that were kept closed for longer than the time used to evaluate the O2 influx rates. The lines represent 1-min interval datasets and the symbols on the lines were plotted every 3 h and denote the sampling locations.

Figure 4 O2 concentration profiles at points [4] and [5] at monitoring depths of 5, 10 and 20 cm. The horizontal axis denotes the elapsed time after the installation of the monitoring pipes.

The O₂ concentration in the closed chamber at point [1] remained steady at 0.19 m³ m⁻³ (Fig. 3), almost the same concentration level as that observed at a depth of 5 cm at points [4] and [5] (Fig. 4). Therefore, “the low influx rate with a high steady-state concentration” at point [1] implied that O₂ consumption was weaker than at the other two points and occurred mainly in the top 5 cm from the soil surface.

In contrast, at points [2] and [3] the O₂ concentration in the chambers dropped to <0.16 m³ m⁻³ (Fig. 3). These concentrations were 0.03 m³ m⁻³ lower than the concentration observed at a depth of 5 cm, but marginally higher than that recorded at a depth of 10 cm (Fig. 4). Therefore, “the low steady-state concentrations” at points [2] and [3] implied that the depth range of O₂ consumption was spread more deeply than at point [1]. This further suggested that the O₂ consumption per depth at point [3] was smaller than that at point [1] because the surface O₂ influx rates at points [1] and [3], in other words, the entire O₂ consumption per unit time at two points, were at a similar level to each other.

Relative importance of O₂ consumption to diffusive transport

The drops in O₂ concentration shown in Fig. 4 reflect the relative importance of O₂ consumption to the diffusive transport of O₂ at depth. According to the mass conservation equation of a soil gas, a change in the O₂ concentration (C) can be described as follows (Hillel 1998; Jury and Horton 2004):

(1)

where t is time [T], a, D and S are air-filled porosity [L³ L⁻³], O₂-air binary diffusion coefficient [L² T⁻¹] and O₂ consumption rate [M L³ T⁻¹], respectively. Equation 1 indicates that when the profile of a remains stable during measurement, the O₂ concentration at any depth should drop linearly with time as the O₂ consumption becomes more dominant on the right-side of Eq. 1.

In Fig. 4, a linear drop in the O₂ concentration was observed at depths of 10 and 20 cm at point [4] as well as at a depth of 20 cm at point [5], implying that diffusive transport of the soil O₂ around the monitoring depths was severely restricted and that the O₂ concentration dropped mainly as a result of biological O₂ consumption in the soil adjacent to the O₂ sensors. Among these three monitoring depths, the curve observed at a depth of 20 cm at point [5] declined less steeply than the curves observed at depths of 10 and 20 cm at point [4], suggesting weaker O₂ consumption. In contrast, at a depth of 10 cm at point [5], gas diffusion was likely to be more responsible for the drop in the monitored O₂ concentration than biological O₂ consumption because the drop in the observed O₂ concentration was slowing, implying dissipation of the O₂ concentration gradient between the soil gas and the air with which the O₂ sensor was directly in contact.

The faster and more linear drop observed at depths of 10 and 20 cm at point [4] suggested high O₂ consumption throughout the soil layers and could explain the high surface O₂ influx recorded at point [2], which was close to point [4] (Fig. 1). However, the strongly linear lowering behavior could give an assumption that a possible maximum O₂ consumption rate at a depth might surpass far beyond a diffusive O₂ supply rate from a soil surface to the depth. On this assumption, the depth range of O₂ consumption could be confined to a shallow portion of the unsaturated zone because of low soil gas diffusivity in the peat soil layer and an unsaturated anaerobic layer could take place.

To confirm our implications from the present study, further studies examining soil gas diffusivity combined with long-term monitoring of the soil gaseous phase are required in tropical peatlands with various soil layer profiles.

Acknowledgments

This study was supported by the research funding program Core Research of Evolutional Science and Technology provided by the Japan Science and Technology Agency.