Temperature sensitivity of anaerobic labile soil organic carbon decomposition in brackish marsh

ABSTRACT

The global warming has a potential for acceleration of labile soil organic carbon decomposition. Arrhenius equation is one of the useful equation for predicting temperature sensitivity of carbon decomposition, with the activation energy of rate constant being a key factor. The purpose of this study is the evaluation of temperature sensitivity of labile soil organic carbon decomposition under anaerobic condition in wetland soil using the activation energy of rate constant among different vegetation types. The soil samples were incubated at three different temperatures (10, 20, and 30°C) under anaerobic condition and carbon decomposition rates (sum of CO₂ and CH₄ production) were measured by gas chromatography. The first-order kinetic model with Arrhenius equation was used for approximate of anaerobic carbon decomposition. For determination of activation energy of rate constant, non-linear least-squares method was conducted between observed carbon decomposition rate and predicted carbon decomposition rate which calculated by Arrhenius equation. The activation energy of rate constant of anaerobic labile soil organic carbon decomposition was different among vegetation types. We successfully determined the activation energy of rate constant of CO2 or CH4 production from Phragites, Juncus, and Miscanthus+Cirsium-dominated vegetation soil with Arrhenius equation. Hence, this study suggests that Arrhenius equation was useful for evaluation of temperature sensitivity of labile soil organic carbon decomposition not only aerobic condition, but also anaerobic condition among several vegetation types in the wetland ecosystem. Moreover, gaseous carbon production from soil under Juncus yocoscensis dominated soil appeared higher activation energy and temperature sensitivity than that from soil under other vegetation types.

KEY WORDS:

• Activation energy of rate constant
• Arrhenius equation
• decomposition rate constant
• minimizing residual sum of square
• vegetation types

1. Introduction

Global carbon dynamics of terrestrial ecosystems affects the concentration of greenhouse gases (e.g. CO₂ and CH₄) in the atmosphere. Rising temperatures are expected to stimulate soil respiration, both by accelerating rates of carbon cycling via autotrophic respiration and by providing a potentially powerful positive feedback to climatic warming via heterotrophic decomposition of organic matter (Carey et al. 2016). Labile soil organic carbon (LSOC) is the most active fraction of soil organic carbon with rapid turnover rates less than a few years (Parton et al. 1987). The temperature sensitivity of LSOC decomposition is an intensively discussed because it has important feedback on the global carbon dynamics (Davidson and Janssens 2006). Conventionally, the increase in reaction rate per 10°C rising, i.e. the Q₁₀ coefficient is been used to describe the temperature response of soil carbon mineralization, likewise is the Arrhenius equation (Wang et al. 2015).

It is one of the most reasonable equations of thermodynamic process. It showed the activation energy of rate constant as follows:

where k is rate constant of the chemical reaction, A is the pre-exponential factor, T is the absolute temperature (K), R is the universal gas constant (8.31 J K⁻¹ mol⁻¹), and E_a is the activation energy of the reaction (J mol⁻¹). According to this equation, activation energy is a key factor for evaluating the temperature sensitivity of chemical reaction. However, most of the study about the evaluation of the effect of temperature on soil carbon mineralization using the Arrhenius equation has been conducted under oxic condition (Knorr et al. 2005; Fierer et al. 2005; Craine et al. 2010).

Soil organic carbon decomposition under anoxic conditions is different to that under oxic condition because not only CO₂ but also CH₄ is produced. Moreover, decomposition rate is slower than under oxic conditions (Boon et al. 2006). Inubushi et al. (1997) showed the usefulness of Arrhenius equation for evaluation of activation energy of carbon decomposition from paddy soil under anaerobic condition. However, the activation energy of anaerobic carbon decomposition has not been yet determined in cool-template wetland soil. Wetlands represent the largest component of the terrestrial biological carbon pool, and thus play an important role in global carbon cycles (Dixon and Krankina 1995; Sahagian and Melack 1996). The positive global warming feedback is more likely occurred at cool-temperature field compared with at warming field. Moreover, the above vegetation types are having also big an effect on the temperature sensitivity of soil organic carbon decomposition (Inglett et al. 2012). The objective of this study is the evaluation of the temperature sensitivity of anaerobic LSOC decomposition using activation energy calculated by Arrhenius equation in cool-temperate brackish marsh among three vegetation types.

2. Materials and methods

2.1. Study site

The study site was located in a cool-temperate brackish marsh around Lake Obuchi in Aomori Prefecture (40°57´N, 141°22´E; Figure 1). The average annual precipitation and temperature is 1488 mm and 10.8°C, respectively (Japan Meteorological Agency,2005; Available from: http://www.jma.go.jp/jma/menu/report.html).

Figure 1. Location of the cool-temperate brackish marsh around Lake Obuchi (40°57ʹ N, 141°22ʹ E).

The brackish marsh was classified into three typical vegetation types: Phragmites vegetation, where Phragmites. australis (Cav.) Trin. ex Steudel represented 72.5% of the vegetation, Juncus vegetation which Juncus yokoscensis (Franch. & Sav.) Satake covering in 54.3%of the vegetation, and Miscanthus+Cirsium vegetation, with Miscanthus sinensis Andersson and Cirsium inundatum Makino, which covered 37.8% and 26.0%, respectively (Suzuki et al. 2010).

Each vegetation boundaries were distinct, and the order of vegetation arrangement, Phragmites vegetation, Juncus vegetation, and Miscanthus+Cirsium vegetation, corresponded to the distance from the shore of Lake Obuchi (Figure 1). Phragmites vegetation is the nearest vegetation from the lake shore. The average water table level was around the soil surface level for Phragmites and Juncus vegetation, and approximately 15 cm below the soil surface for Miscanthus+Cirsium vegetation. The soil chemical properties in each vegetation types are described in Table 1.

Table 1. Soil chemical properties among three vegetation types.

	pH	EC (dS m^−1)	T–C (g kg^−1)	T–N (g kg^−1)	CN ratio
Phragmites vegetation	4.54 ± 0.94 b	63.9 ± 17.3 a	39.7 ± 2.23 a	3.00 ± 0.14 a	13.2
Juncus vegetation	6.12 ± 0.34 a	14.1 ± 7.0 b	53.5 ± 1.70 a	3.92 ± 0.01 a	13.7
Miscanthus+Cirsium vegetation	5.49 ± 0.14 b	3.2 ± 1.0 c	47.2 ± 1.57 a	3.83 ± 0.11 a	12.4

Values are means ± standard deviation (n = 5).

EC, electric conductivity; T–C, total carbon content in soil; T–N: total nitrogen content in soil; CN ratio: total carbon and nitrogen ratio.Significant (p < 0.05) differences among vegetation types are labeled with alphabet.

2.2. Incubation experiments and gas production

Five soil cores (30 cm in depth and 5.5 cm in diameter) were collected on late June 2008 using a core sampler (Hand sampler HS-30S, Fujiwara Scientific CO., LTD, Tokyo, Japan) from each of three vegetation types (in total 15 cores). Surface layer soils (0–5 cm) were collected from each core. Each surface layer soil in each vegetation samples was immediately sieved (2 mm) for use in the incubation experiment. The soil sample was flooded with distilled water and preincubated under O₂ free nitrogen condition for 14 days. In addition, 10 g of soil samples were put into 30-mL serum vials and their soil water content was adjusted to simulate a waterlogged condition with distilled water prior to the incubations. Headspace gas in the bottle was then replaced with nitrogen gas. After that, the bottle was sealed tightly with a butyl stopper. Five tubes of soils from each vegetation were incubated at each three temperatures (10, 20, and 30°C) in the dark for 168 days, thus representing the LSOC as it is decomposed within such short time (Parton et al. 1987). At seven-day intervals, CO₂ and CH₄ were quantified by gas chromatography with thermal conductivity detector (TCD) and flame ionization detector (FID), respectively (Inubushi et al. 1997). The sum of CO₂+CH₄ production was determined as the observed gaseous carbon production (G_obs: g–C g soil⁻¹). After each sampling, the headspace gas was replaced with fresh N₂.

2.3. Calculation of activation energy of rate constant

Soil organic carbon (C) mathematically separated into two components is represented as follows:

(1)

where C₁ (g–C g soil⁻¹) represents the LSOC fraction and C₂ (g–C g soil⁻¹) represents the recalcitrant organic carbon fraction, which was hardly decomposed during the incubation period (168 days).

In C₁ fraction, the first-order kinetic model proposed changes of organic carbon content in time (t) as follows:

(2)

where k₁ is the decomposition rate constant of C₁ fraction.

The LSOC decomposition rate is described with first-order model as follows:

(3)

where C₁,_t is LSOC decomposition rate in the time t, t is incubation days (d), and C_1,0 is LSOC fraction at day 0. Therefore, cumulative LSOC decomposition during incubation period (G_cal) is indicated as follows:

(4)

The apparent activation energy of the anaerobic LSOC decomposition rate constant is described with Arrhenius equation as follows:

(5)

where k₁₀ is the decomposition rate constant at the soil temperature of 10°C, E₁ (J mol⁻¹) is the activation energy in the Arrhenius equation, R is the universal gas constant (8.31 J K⁻¹ mol⁻¹), T (K) is soil temperature on incubation days, and T₀ (K) is the reference temperature (283 K in this study). Therefore, Equation (4) can be written as

(6)

For fitting G_cal at Equation (6), C_1,0, k₁₀, and E₁ were calculated to minimize the residual sum of squares between G_cal and G_obs using a non-linear least-squares method under all incubation temperature and incubation days with Microsoft Excel Solver program. The suitable C_1,0, k₁₀, and E₁ were determined on each vegetation types (n = 5). The determination coefficient was obtained by linear regression analysis between G_obs and G_cal.

The Q₁₀ values of k₁ are calculated as follows:

where k_1t is k₁ in temperature t (K) and k_1(t+10) is k₁ in temperature t + 10 (K).

3. Statistical analysis

Statistical analysis was performed by using SPSS software (SPSS Inc., Chicago, IL, USA). The significant differences in CO₂, CH₄ and CO₂+CH₄ production among vegetation types were estimated by the Tukey method at p < 0.05 level.

4. Results and discussion

CO₂+CH₄ production rates from soil under Miscanthus+Cirsium vegetation were the highest among the three vegetation types, while that from soil under Juncus vegetation had the lowest rates under all temperatures (Figure 2). However, CO₂+CH₄ production from soil under Juncus increased more than four times and twice between 10 and 20°C and 20–30°C, respectively (see Table 1 in Supplementary Material). Soil under Phragmites vegetation and Miscanthus+Cirsium vegetation experienced doubled carbon production rates between the ranges 10–20°C and 20–30°C, respectively. In all samples, the temperature dependency of CO₂ or CH₄ production was differed (Figure 1 in Supplementary Material). The CO₂ production from soil under Juncus vegetation increased about 4.7 times between 10 and 20°C while that from other vegetation and temperature indicates around twice (Table 1 in Supplementary Material). The CH₄ production was increased about four and six times between 10 and 20°C from soil under Juncus and Miscanthus+Cirsium vegetation, respectively. The Q₁₀ values of CO₂+CH₄ production from wetland and paddy field are ranged from 1.2 to 9.13 (Kirwan et al. 2014; Wang et al. 2015; Leroy et al. 2017). The Q₁₀ values of CH₄ production were higher in Miscanthus+Cirsium vegetation, which was the farthest vegetation from the lake shore, than other vegetation types. In this site, EC was drastically changed along with the distance from the lake shore. Neubauer (2013) indicated that temperature sensitivity of CO₂ emission did not vary with salinity, while that of CH₄ emission decreased with increasing salinity. However, this did not explain fully the difference of temperature sensitivity of anaerobic LSOC decomposition among vegetation types.

Figure 2. The carbon mineralization from anaerobic soil among each vegetation types under 10, 20, and 30°C during incubation period. Opened circle; Phragmites vegetation, opened square; Juncus vegetation and opened triangle; Miscanthus+Cirsium vegetation.

Error bars indicate the standard variation of each sample (n = 5).

The residual sum of square between G_obs and G_cal of CO₂+CH₄ was <0.05 in each vegetation types (Table 2). Under this condition, E₁, k₁₀, and C_1,0 were determined for CO₂+CH₄ prediction from soil under each vegetation (Table 3). The determination coefficients of linear regression between G_obs and G_cal of CO₂+CH₄ were >0.98 in each vegetation types and they were positively correlated (Table 2 and Figure 3). Therefore, anaerobic LSOC decomposition was well fitted for first-order models with Arrhenius equation, like it was done for aerobic carbon decomposition as Knorr et al. (2005). These high correlations indicated that the E₁, k₁₀, and C_1,0 were successfully estimated by non-linear least-squares method. E₁ for CO₂+CH₄ production was determined as 7.65 × 10⁴, 1.05 × 10⁵, and 7.98 × 10⁴ J mol⁻¹ from soil under Phragmites, Juncus and Miscanthus+Cirsium vegetation, respectively (Table 3). E₁ of CO₂+CH₄ production from soil under in each vegetation in the brackish marsh was higher than previous studies, anaerobic carbon decomposition in field soil and aerobic carbon decomposition (3.9 × 10⁴ to 5.3 × 10⁴ J mol⁻¹, Inubushi et al. 1997; Knorr et al. 2005). Therefore, temperature effects of LSOC decomposition in wetland were more sensitive than that in paddy field and under aerobic condition. Several studies also showed that the labile carbon pool was less temperature sensitive than recalcitrant organic carbon fraction, because recalcitrant carbon pool has a higher activation energy (Knorr et al. 2005; Davidson and Janssens 2006). Before plantation, paddy fields were plowed and soil was exposed to oxygen. In this process, soil organic carbon might be drastically decomposed and LSOC fraction might be increased than that in the wetland ecosystem.

Table 2. Residual sum of square (RSS) and determination coefficient (R²) between the calculated and the observed CO₂, CH₄ and CO₂ + CH₄ production.

		RSS	R^2
CO2	Phragmites vegetation	3.16 × 10^–2	0.975
	Juncus vegetation	5.71 × 10^–3	0.986
	Miscanthus+Cirsium vegetation	6.16 × 10^–2	0.978
CH4	Phragmites vegetation	4.69 × 10^–4	0.982
	Juncus vegetation	8.05 × 10^–4	0.944
	Miscanthus+Cirsium vegetation	8.80 × 10^–4	0.971
CO2+CH4	Phragmites vegetation	2.47 × 10^–2	0.981
	Juncus vegetation	4.31 × 10^–3	0.991
	Miscanthus+Cirsium vegetation	5.14 × 10^–2	0.98

Table 3. The labile soil organic carbon fraction on day 0 (C_1,0), rate constant at 10°C (k₁₀), and activation energy of rate constant (E₁) in each vegetation types.

	k10 (d^−1)	E1 (J mol^−1)	C1,0 (g kg soil^−1)
Phragmites vegetation	1.80 × 10^–3 ± 1.39 × 10^–6 a	7.65 × 10^4 ± 3.84 × 10^1 b	6.70 × 10^–1 ± 3.04 × 10^–4 b
Juncus vegetation	0.50 × 10^–3 ± 4.36 × 10^–5 b	10.3 × 10^4 ± 3.82 × 10^3 a	5.03 × 10^–1 ± 3.13 × 10^–2 b
Miscanthus+Circium vegetation	1.06 × 10^–3 ± 1.11 × 10^–4 a	7.98 × 10^4 ± 4.38 × 10^3 b	11.1 × 10^–1 ± 8.91 × 10^–2 a

Values are means ± standard deviation (n = 5). Significant (p < 0.05) differences among vegetation types are labeled with alphabet.

Figure 3. Relationships between the calculated gaseous carbon production (G_cal) and the observed gaseous carbon production (G_obs).

Dotted line indicated a line with a slope of 1.

Temperature sensitivity of anaerobic labile soil carbon decomposition was higher under Juncus vegetation than under the other vegetation types. As CO₂+CH₄ production, these parameters were also successfully determined in CO₂ or CH₄ production from each vegetation soil, respectively (Table 2 in Supplementary Material). E₁ was the highest in case of Juncus vegetation, and Miscanthus+Cirsium vegetation for CO₂ and CH₄ production, respectively. Q₁₀ of k₁ for CO₂ or CO₂+CH₄ production from soil under Juncus vegetation was higher than other vegetation (Table 4 and Table 3 in Supplementary Material). Therefore, soil under Juncus vegetation indicated higher temperature sensitivity of anaerobic labile carbon decomposition than under the other vegetation types. Moreover, CO₂ production was more strongly affecting than CH₄ production from soil under Juncus vegetation. Bray et al. (2012) showed that plant residue of Juncus related species has lower lignin content and faster decomposition rate than other wetland plants. This study suggested that soil chemical properties and LSOC molecular construction might complexly correlated temperature sensitivity of anaerobic LSOC decomposition.

Table 4. The Q₁₀ values of decomposition rate constant (k₁) between for CO₂, CH₄, and CO₂+CH₄ production among different vegetation types.

		283–293 K	293 – 303 K
CO2	Phragmites vegetation	2.78	2.59
	Juncus vegetation	4.54	4.11
	Miscanthus+Cirsium vegetation	2.99	2.78
CH4	Phragmites vegetation	2.16	2.05
	Juncus vegetation	2.37	2.24
	Miscanthus+Cirsium vegetation	3.04	3.32
CO2 + CH4	Phragmites vegetation	3.04	2.82
	Juncus vegetation	4.44	4.04
	Miscanthus+Cirsium vegetation	3.20	2.94

Future research will focus on the interaction between the detailed chemical composition of labile carbon fraction, soil chemical properties and temperature sensitivity of carbon decomposition under anaerobic condition.

Supplementary material

Supplementary materials can be accessed here