Assessing soil CO₂ efflux using continuous measurements of CO₂ profiles in soils with small solid-state sensors

Abstract

This paper describes a new method to monitor continuously soil CO₂ profiles using small solid-state CO₂ sensors buried at different depths of the soil. Based on the measurement of soil CO₂ profile and a gaseous diffusivity model, we estimated soil CO₂ efflux, which was mainly from heterotrophic respiration, and its temporal variation in a dry season in a Mediterranean savanna ecosystem in California. The daily mean values of CO₂ concentrations in soils had small variation, but the diurnal variation was significant and correlated well with soil temperature. The daily mean CO₂ concentration remained steady at 396 μmol mol⁻¹ at 2 cm depth during the dry summer from days 200 to 235 in 2002. Over the same period, CO₂ concentration decreased from 721 to 611 μmol mol⁻¹ at 8 cm depth, and from 1044 to 871 μmol mol⁻¹ at 16 cm. The vertical soil CO₂ concentrations changed almost linearly with depth up to 16 cm, but the gradient varied over time. Based on the soil CO₂ gradient and the diffusion coefficient estimated from the Millington–Quirk model, continuous soil CO₂ efflux was calculated. The daily mean values of CO₂ efflux slightly decreased from 0.43 to 0.33 μmol m⁻² s⁻¹ with a mean of 0.37 μmol m⁻² s⁻¹. The mean diurnal range of CO₂ efflux was greater than the range of daily mean CO₂ efflux within the study period. The diurnal variation of soil CO₂ efflux ranged from 0.32 to 0.45 μmol m⁻² s⁻¹ with the peak value reached between 14:30 and 16:30 h. This pattern corresponded well with the increase in soil temperatures during this time. By plotting CO₂ efflux vs. soil temperature, we found that CO₂ efflux correlated exponentially with soil temperature at the depth of 8 cm, with R² of 0.86 and Q₁₀ of 1.27 in the summer dry season. The Q₁₀ value increased with the depth of soil temperature measurements. The high correlation between CO₂ efflux and temperature explains the diurnal pattern of CO₂ efflux, but moisture may become another factor driving the seasonal pattern when moisture changes over seasons. The estimated CO₂ efflux using this method was very close to chamber measurements, suggesting that this method can be used for long-term continuous measurements of soil CO₂ efflux.

Introduction

Soil surface CO₂ efflux, or soil respiration, is a major component of the biosphere’s carbon cycle because it may constitute about three-quarters of total ecosystem respiration (Law et al., 2001). In recent years, soil CO₂ efflux has been the subject of intense studies because of its potential and controversial role in amplifying global warming (e.g. Trumbore et al., 1996, Liski et al., 1999, Cox et al., 2000, Giardina and Ryan, 2000, Kirschbaum, 2000, Luo et al., 2001). Soil carbon modelers generally view soil CO₂ efflux as a function of soil temperature or a combination of soil temperature and moisture (e.g. Raich and Schlesinger, 1992, Davidson et al., 1998, Epron et al., 1999, Treonis et al., 2002). However, there is no consensus in functional forms and parameterization in these models. The uncertainty is partly due to the instrumentation and methods used to measure soil CO₂ production and efflux (Livingston and Hutchinson, 1995, Davidson et al., 2002).

Information on soil respiration is also needed to interpret eddy covariance measurements, which are now being acquired on a quasi-continuous basis across the global FLUXNET network (Baldocchi et al., 2001). The eddy covariance method measures ecosystem productivity (NEP), a net result of photosynthesis and respiration, but it does not provide individual information such as photosynthesis, autotrophic respiration, and heterotrophic respiration (though nighttime eddy covariance data provide information on ecosystem respiration in the dark). Since these processes have different mechanisms and environmental drivers, partitioning of eddy covariance data has received much attention (Piovesan and Adams, 2000). Continuous eddy covariance measurements of CO₂ fluxes need continuous soil CO₂ measurements at a similar frequency (per half-hour) in order to decompose NEP, understand temporal variation, and explain some unusual episodic events that are observed.

Methods of soil CO₂ efflux measurement are still in development. An early method periodically extracts soil gas samples from different depths to study CO₂ profile and diffusion (De Jong and Schapper, 1972, Wagner and Buyanovsky, 1983, Burton and Beauchamp, 1994, Davidson and Trumbore, 1995). The gas extraction method can provide information on soil CO₂ production at several depths, but it cannot provide in situ, continuous and convenient data on CO₂ efflux. Furthermore, this method will disturb the soil environment. An unavoidable bias may occur during the processes of gas extraction, storage, transport, and measurement.

Chamber-based measurements allow us to directly measure CO₂ efflux from soils on a small scale (e.g. Meyer et al., 1987, Norman et al., 1992). Fixed chambers and portable chambers have evolved into automated systems for continuous and semi-continuous measurements (Goulden and Crill, 1997, Russell et al., 1998, Scott et al., 1999, Drewitt et al., 2002, King and Harrison, 2002). Shortcomings with closed-chamber methods, however, still exist. Efflux readings may be biased by disturbing air pressure and altering CO₂ concentration in the soil (Livingston and Hutchinson, 1995, Healy et al., 1996, Davidson et al., 2002). By measuring accumulation of soil CO₂ productivity released from the soil surface, chambers are unable to provide information about soil profiles and individual contributions at certain soil depths, which is important for understanding soil carbon mechanisms. Currently, no reliable and robust automated chambers for field measurements are commercially available.

Understory eddy covariance towers provide an alternative to continuously measure soil CO₂ efflux without disturbing the soil (Baldocchi and Meyers, 1991, Law et al., 1999). As with overstory eddy covariance techniques, understory eddy covariance measurement may face difficulty in measuring respiration at night when turbulence is weak and intermittent and drainage flows dominate the transfer of CO₂ (Goulden et al., 1996, Moncrieff et al., 1997). Compared with overstory eddy covariance, the low height of understory towers corresponds with small areas of footprint, which may induce errors when large areas of ecosystems are represented. Furthermore, understory eddy covariance data cannot separate soil CO₂ efflux, bole respiration below sensors, and overlying herbaceous vegetation, when it is present.

Partitioning NEP into GPP (gross primary productivity) and NPP (net primary productivity), and partitioning soil respiration into autotrophic and heterotrophic respiration are of critical importance for building process-based models since these components respond differently to abiotic and biotic drivers. Despite the development of methods such as trenching and isotopic approaches for partitioning the source of soil CO₂ (Hanson et al., 2000), few studies have directly measured and modeled heterotrophic respiration in situ without any disturbance. As a result, studies on temperature sensitivity (Q₁₀) of soil CO₂ efflux often combine heterotrophic respiration with autotrophic respiration (e.g. Raich and Schlesinger, 1992, Lloyd and Taylor, 1994, Xu and Qi, 2001), which may vary with plant physiological and phenological factors other than temperature. Thus, correlation coefficients between soil CO₂ efflux and temperature often have low values. Savanna ecosystems with dead grasses and live but sparse trees in the summer provide a unique opportunity to measure and model heterotrophic respiration. However, publications on heterotrophic respiration in savannas are limited.

Due to the limitation of instrumentation, particularly due to the large size of commonly used infrared gas analyzers, there are very few publications on continuous measurements of CO₂ profile in the soil. Recently, an innovative CO₂ sensor was developed for air quality monitoring and control. This instrument has the potential to be buried in the soil and measure CO₂ in the soil atmosphere. Hirano et al. (2000) first used a type of these small CO₂ sensors (GMD20, Vaisala Inc., Finland) buried in the soil under a deciduous broad-leaved forest in Japan to deduce soil respiration, and therefore have demonstrated the feasibility of the instrument.

In order to develop more measurement methods in soil CO₂ efflux, this paper describes in detail the use of the new small solid-state CO₂ sensors (GMT222, Vaisala Inc., Finland) to continuously monitor soil CO₂ profiles and soil CO₂ efflux by burying these CO₂ sensors at different soil depths. Based on the measurement of the CO₂ profile and a diffusivity model, we estimated rates of soil CO₂ efflux in a dry season in a Mediterranean savanna ecosystem in California. The relationship between CO₂ efflux and soil temperature was explored. Soil CO₂ efflux measurements by chambers were used to validate this method.