Soil Moisture Effects on Leaf Litter Decomposition and Soil Carbon ...

Published September 26, 2014

Wetland Soils

Soil Moisture Effects on Leaf Litter Decomposition and Soil Carbon Dioxide Efflux in Wetland and Upland Forests Tae Kyung Yoon

Dep. of Environmental Science and Ecological Engineering Graduate School Korea Univ. Seoul 136-713 Korea

Nam Jin Noh

River Basin Research Center Gifu Univ. Gifu 501-1193 Japan

Saerom Han Jongyeol Lee

Dep. of Environmental Science and Ecological Engineering Graduate School Korea Univ. Seoul 136-713 Korea

Yowhan Son*

Dep. of Environmental Science and Ecological Engineering Graduate School Korea Univ. Seoul 136-713 Korea and River Basin Research Center Gifu Univ. Gifu 501-1193 Japan and Dep. of Biological and Environmental Sciences Qatar Univ. PO Box 2713 Doha Qatar

This study examined, first, the response of litter decomposition and soil CO2 efflux (RS) to different soil moisture conditions and, second, the application of various litter decomposition functions in a wetland and upland forest dominated by Japanese alder. One upland site (US) and three wetland sites—a drained site (DS), poorly drained site (PDS), and surface saturated site (SSS)— were selected based on their variation in soil moisture conditions. Litter mass loss, as determined by a 4-yr litter bag incubation, was applied to Olson’s simple exponential function, Berg’s asymptotic function, and the rational function. The litter decomposition rate constant (yr−1), which was commonly obtained by the simple exponential function, was highest in PDS (1.181), followed by SSS (0.950), DS (0.922), and US (0.528). The limit value of the litter mass loss, as determined by the asymptotic function was higher in DS (91.7%) and PDS (89.0%) than in SSS (76.9%) and US (70.5%). The rational function provided the most precise fitting of the litter mass loss pattern with few parameters. Periodic saturation and the higher leaf N content in PDS may enhance litter decomposition compared to constant saturation or drained conditions. The RS (mg C m−2 h−1) values, periodically measured using a portable infrared gas analyzer, were ranked in the order US (12.6–355.1) = DS (7.1–324.0) > PDS (5.5–220.9) > SSS (0.0–153.8). More hydric conditions probably reduced the vegetation biomass (in contribution to autotrophic RS) and aerobic microbial activities (in contribution to heterotrophic RS). The RS temperature dependency (Q10) was little affected by soil moisture conditions, ranging from 2.48 to 2.69. It is concluded that the litter decomposition rate and RS were highest under periodic saturation and under lower soil moisture conditions, respectively. Abbreviations: AIC, Akaike information criterion; AICc, corrected Akaike information criterion; DAMM, dual Arrhenius and Michaelis–Menten; DS, drained site; HELCA, Heonilleung Ecosystem and Landscape Conservation Area; PDS, poorly drained site; SSS, surface saturated site; SWCV, volumetric soil water content; US, upland site.

E

cologists have dedicated themselves to the elucidation of decomposition processes in ecosystems. Among various decomposition processes, litter decomposition and soil CO2 efflux (i.e., soil respiration; RS) have been the most popular topics of study since the middle of the 20th century. Litter decomposition is defined as “physical, chemical, and biological mechanisms that transform organic matter into increasingly stable forms” in plant detritus (Berg and McClaugherty, 2008, p. 1), and it contributes to soil organic matter formation, soil fauna composition, nutrient availability, and vegetation growth (Hattenschwiler et al., 2005; Prescott, 2005). Soil respiration is the primary process by which an ecosystem releases sequestered C back to the atmosphere (Schlesinger and Andrews,

Soil Sci. Soc. Am. J. 78:1804–1816 doi:10.2136/sssaj2014.03.0094 Received 7 Mar. 2014. *Corresponding author ([email protected]). © Soil Science Society of America, 5585 Guilford Rd., Madison WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Soil Science Society of America Journal

2000; Ryan and Law, 2005; Davidson et al., 2006). Therefore, the importance of these studies is currently highlighted in the interest of ecosystem C cycling under the changing global climate. However, litter decomposition and RS in wetlands have received less attention than those in other ecosystems, despite the important contribution of wetlands to the global C cycle (Eswaran et al., 1993; Erwin, 2009). For example, the litter decomposition and RS records for wetlands comprised only 5 and 3% of the global database compiled by Zhang et al. (2008) and Bond-Lamberty and Thomson (2012), respectively. Moreover, wetland soil properties have a large spatial variability that is strongly related to soil moisture conditions (Yoon et al., 2013); therefore, litter decomposition and RS in wetland are expected to vary along a soil moisture gradient, even within a single plot. Several studies have attempted to compare the influence of soil moisture conditions on litter decomposition (Neckles and Neill, 1994; Atkinson and Cairns, 2001; Bedford, 2005) and RS (Davidson et al., 1998; Savage and Davidson, 2001; Webster et al., 2008, 2009; Fissore et al., 2009). The soil moisture effects on various biogeochemical responses are undeniable (RodriguezIturbe, 2000; Porporato et al., 2003); however, our knowledge of the soil moisture effects remains insufficient for comprehending complex response patterns (e.g., linear or nonlinear and positive or negative) and the varied optimum conditions (Western et al., 2002; Porporato et al., 2003; Robinson et al., 2008). Here, empirical reports for the response of litter decomposition and RS to soil moisture would be valuable, not only for elucidating these processes in wetland soils, but also by providing algorithms for ecohydrologic modeling (Porporato et al., 2003; Manzoni and Porporato 2009; Davidson et al., 2012; Moyano et al., 2013). Moreover, there remain unsolved questions regarding the interpretation of the key parameters of litter decomposition and RS, despite the abundance of reports and long history of observations. For instance, Prescott (2005) questioned whether litter decomposition rates are a valuable and informative measure for representing the litter decomposition process. Instead, other measures and functions that would be more informative have been recently suggested, such as a limit value of the litter decomposition concept (Berg et al., 1996; Prescott, 2005), a litter quality compartment model (Zhang et al., 2008; Manzoni et al., 2012), or a process-based model that includes the kinetic and stoichiometric laws between substrates and decomposers and between C and N (Manzoni and Porporato, 2009). Various newly-suggested litter mass loss functions could provide an opportunity for elaborating on the preexisting knowledge of litter decomposition processes (Manzoni et al., 2012); however, the previous case studies that have applied, validated, and compared those functions are insufficient. Meanwhile, RS studies have challenged the temperature dependency of RS. The regulating factors on the temperature dependency of RS such as temperature (Chen and Tian, 2005; Davidson and Janssens, 2006), substrate quality (Sierra, 2012), and soil moisture (Silvola et al., 1996; Wickland et al., 2010; Taggart et al., 2012; Moyano et al., 2013) have been tested. For www.soils.org/publications/sssaj

instance, Chen and Tian (2005) suggested that the temperature dependency of RS would differ among global biomes and decrease with temperature based on a meta-analysis. Taggart et al. (2012) reported that the temperature dependency of RS was highest for higher soil moisture conditions that probably did not limit RS. In addition, the general temperature dependency of RS has been recently suggested based on either an empirical synthesis (Chen and Tian, 2005) or a mechanistic approach (Davidson et al., 2006). Because the temperature dependency of RS is assumed to indicate the response of soil C emission to global warming and to change the C balance between the soil and the atmosphere (Davidson and Janssens, 2006), its mechanism and interactions with environmental factors need to be elucidated. Moreover, understanding the temperature dependency of RS would be critical for wetland soils, which are vulnerable to changes in the global climate (Davidson and Janssens, 2006; Mitsch et al., 2013). This study aimed to determine the soil moisture effect on litter decomposition and RS under different soil moisture conditions in a wetland and an adjacent upland. We hypothesized that, first, litter decomposition and RS would be lower in more hydric conditions (Hypothesis 1), and second, the temperature dependency of RS would decrease as volumetric soil water content (SWCV ) increased (Hypothesis 2). In addition, we aimed to compare different litter mass loss functions for describing respective characteristics of litter mass loss patterns at each site and to fit an RS function based on the environmental variables for analyzing controlling factors to the spatiotemporal variability of RS.

Materials and methods Study Site

This study was performed in a pure Japanese alder [Alnus japonica (Thunb.) Steud.] forest in the Heonilleung Ecosystem and Landscape Conservation Area (HELCA) in a suburban area of Seoul, central Korea (37°27¢52² N, 127°04¢53² E, 40 m asl) (Fig. 1). Japanese alder is widely found from the wetland area to the upland area on the bottomland (2 ha) of the southern slope of the HELCA watershed. A wetland in the Japanese alder forest of HELCA has developed during the last 25 yr due to landuse change in the surrounding area. In the mid-1980s, a parking lot and a management office building were constructed at the downstream of HELCA, altering the hydrology of the Japanese alder forest. The poor drainage due to the construction of the parking lot, the building, and greenhouses resulted in a wetland developing in the Japanese alder forest (Fig. 1a). In addition, to reduce the poor drainage conditions, several drainage channels were constructed through the wetland in the mid-2000s. Consequently, the landscape position, parking lot, and drainage channels have interactively affected the soil moisture conditions in the wetland and are thus highly heterogeneous, from well drained to saturated, on a small scale. Therefore, this site could be ideal for determining the response of soil C processes to soil moisture manipulation with identical species and climate. The mean temperature was 13.5°C (monthly range: −1.9 to 27.2°C), and the annual precipitation was 1473 mm (annual range: 1057– 1805

Fig. 1. Location of (a) the study site in (b) central Korea. The study was conducted in a Japanese alder forest, where the hydrology had been manipulated. One upland site (US) and three wetland sites (US: upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site) were selected, following the hydrologic gradient that had originated from the construction of a parking lot, a management office building, and greenhouses since the 1980s. At each site, soil CO2 efflux was measured at the five points (closed circles) on a 30-m transect (dashed lines) crossing a 400-m2 plot. Litter bags were placed on nearby soil CO2 efflux measurement points.

1791 mm) from 2007 to 2012 (http://www.kma.go.kr, accessed 29 Nov. 2013) (Fig. 2). In 2005, HELCA was established by the Seoul Metropolitan Government to protect the function of this area as a habitat for native species. A detailed description of the site was provided in a previous study (Yoon et al., 2013). One US (Alfisol) and three wetland sites (Histosol)—a DS, PDS, and SSS—were selected based on the natural drainage classes of the USDA (Soil Survey Staff, 1999) and on hydrologic indicators, such as gravimetric soil water content and the water table (Table 1). Following the hydrologic gradient from dry to wet, DS, PDS, and SSS have experienced unsaturated, seasonally saturated, and constantly saturated conditions, respectively. Especially, PDS was saturated during days after rainfall events; then, unsaturated. Each site has different tree densities and soil properties due to the soil moisture differences (Table 1). Litter bag incubation and soil RS measurement were conducted at five points along an approximately 30-m transect that crossed a 400-m2 plot at each site (Fig. 1a).

Litter Bag Incubation Experiment Litter decomposition was determined using a litter bag incubation experiment. Fresh alder leaf litter on the forest floor was collected at each site in November 2008, and 10 g of airdried litter was inserted into 1-mm mesh litter bags (30 by 30 cm). A total of 200 litter bags (4 sites ´ 5 replicates ´ 10 retrieval times) were placed along the transects in February 2009. Five bags were retrieved from each site 3, 6, 10, 12, 15, 26, and 50 mo from the beginning. Because litter bags were under repeated saturated and flooded conditions, the litter was tightly attached to sediment, unlike the upland litter. Therefore, sediment-free litter mass was measured by the following procedure. The litter was removed from the litter bag and gently washed on a 1-mm 1806

mesh sieve, which was the same mesh size of the litter bags; the litter fragments smaller than 1 mm were assumed to be fine particulate organic matter, which should be excluded from remaining litter mass. Then, the litter was oven-dried at 65°C and weighed. Nevertheless, we were not confident that all sediments were completely removed; thus, the dry mass of litter was then corrected using the ash-free dry mass measured by loss on ignition at 550°C for 4 h. In the correction procedure, a calibration curve, which was preliminarily developed between the ash-free mass to dry mass ratio vs. the mixing ratio of litter and soil using the standard samples (an artificial mixture of pure litter and soil samples), was applied to eliminate the effect of organic matter in the sediment on the ash-free dry mass. The C and N contents in the ground litter were determined using an elemental analyzer (Vario Macro CN analyzer, Elementar Analysensysteme GmbH). Manganese, which may affect longterm litter decomposition (Berg et al., 1996; Virzo De Santo et al., 2009), P (data not presented in the article), and exchangeable cation (Ca, K, Mg, Na; data not presented in the article) contents were determined using inductively coupled plasma–optical emission spectrometry (Vista PRO, Varian, USA) following wet-digestion (BD-46 Block Digester, Lachat Instruments). The change in litter mass remaining was applied to three different litter decomposition functions: Olson’s simple exponential function (Olson, 1963), Berg’s asymptotic function (Berg et al., 1996), and the rational function specified by Rovira and Rovira (2010). Table 2 describes the formula and parameters of each function. The differences between these functions depend on the assumptions of litter mass loss patterns. Here we briefly introduce the key feature of these functions, the details of which will be discussed later (see Discussion). Olson’s simple exponential function assumes that the litter mass is lost with a Soil Science Society of America Journal

Table 1. Description of the upland site (US), drained site (DS), poorly drained site (PDS), and surface saturated site (SSS). Parameter Japanese alder tree Density, trees ha−1 Basal area, m2 ha−1 Diameter at breast height, cm*

US

DS

PDS

SSS

900 104.9

575 45.0

550 35.3

300 21.6

32.6 (9.4)† 27.1 (7.1)

24.2 (8.4) 25.6 (9.4)

Japanese alder leaf litter C content, % 50.0 (0.9) 50.1 (0.2) 49.2 (0.7) N content, %* 2.3 (0.0) 2.4 (0.0) 2.6 (0.0) C/N ratio* 21.8 (0.7) 20.8 (0.4) 18.6 (0.4) Mn content, g kg−1* 1.50 (0.08) 1.52 (0.02) 1.63 (0.02) Surface soil (0–10 cm)

49.5 (0.4) 2.4 (0.0) 20.7 (0.4) 0.78 (0.04)

sandy loam

silt loam

Soil texture Soil pH * Water table, cm

sandy loam

silt loam

4.45 (0.07) 5.04 (0.16) 5.14 (0.08) 5.37 (0.12) < −100 < −15 −15 to 0 >0

Gravimetric soil water content, kg kg−1

2.0

NA‡

d

in reference to decomposition rates estimated by the rational function (kR) is integrated into the mass remaining function Fig. 2. Monthly (a) air temperature and (b) precipitation at the study site from 2007 to 2012, with annual precipitation levels are presented in the inset; (c) volumetric soil water content was determined using time-domain reflectometry measurement under different soil moisture conditions from 2007 to 2010 (US: upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site). Vertical bars indicate standard deviations of means (N = 5).

constant decomposition rate (kE), and reaches zero at the end. On the other hand, Berg’s asymptotic function assumes that the limit value for litter mass loss (m) rarely occurs due to stabilized residues of litter decomposition and humus accumulation. The asymptotic function describes the decrease of litter mass with a constant decomposition rate (kA), which is the same as kE in the simple exponential function, to the specific limit value at which the litter mass loss stops, in contrast to the assumption of complete decomposition in the simple exponential function. Finally, Rovira and Rovira’s (2010) rational function originates from the assumption that the decomposition rate changes with respect to litter quality and environment changes through the decomposition stages. The temporal change in the decomposition rate www.soils.org/publications/sssaj

 -a 3t 4   X t = X 0 exp  2  4b ( t 2 + b )    See Rovira and Rovira (2010) and Table 2 for the definition of each parameter. Because the rational function depends on four parameters, Parameters c and d can be assumed to be constants in practice (Rovira and Rovira, 2010). Assuming c = 0 indicates that kR reaches zero at the end of the kR decrease. When the mass loss curve is close to an S-shape, d = 3 can be assumed.

Soil Respiration Soil respiration was measured at the five points along the transect in each site every 1 to 2 mo from May 2009 to October 2010 using a portable infrared gas analyzer (EGM-4, PP Systems) attached with a closed dynamic chamber (SRC-1, PP Systems; 10-cm diameter by 15 cm). The EGM-4 determines the change of CO2 concentration (DCO2) inside the chamber with an accuracy of 1% of the calibration range (0–2000 mmol mol−1) every 4.8 s for 2 min, and then automatically 1807

Table 2. Description of three litter decomposition functions. Function

Equation†

Simple exponential Xt = X0exp(−kEt) Asymptotic

Xt = X0 − m[1 − exp(−kAt/m)]

Parameter

Reference

kE, decomposition rate constant for simple exponential function

Olson (1963)

kA, decomposition rate constant for asymptotic function m, limit value of litter mass loss

Berg et al. (1996)

a, b, site-dependent parameters kR, instantaneous decomposition rate (dX/X)/dt = c + [at/(t2 + b)]d Rovira and Rovira (2010) c, d, assumed constant parameters (c = 0, d = 3) † Xt, litter mass remaining at time t; X0, initial litter mass.

Rational

Xt = exp[−a3t4/4b(t2 + b2)

calculated the RS (PP Systems, 2004). For minimizing the potential artifacts and biases from RS measurement using a closed dynamic chamber (Davidson et al., 2002), several procedures were performed as follows: (i) the measurements were repeatedly conducted at the permanent points on the preinstalled PVC soil collars (10-cm diameter by 8 cm) to reduce the spatial variability (Rodeghiero and Cescatti, 2008); (ii) the measurements were conducted as close to noon as possible (at approximately 1100–1500 h) to minimize the diurnal variation of RS (Davidson et al., 1998; Noh et al., 2010); (iii) the chamber was manually flushed before every measurement to keep the air inside the chamber corresponding to ambient air; (iv) the DCO2 during the first 5 to 10 s after chamber installation on the soil collar were automatically excluded by the default program of EGM-4 to equilibrate the unstable DCO2 in the chamber; (v) the DCO2 was fitted by quadratic functions to reduce the pressure artifacts (Davidson et al., 2002), following the recommendation of the manufacturer (PP Systems, 2005); (vi) the in-line gas passed a hydrophobic filter that prevented condensation and water vapor effects on CO2 determination; and (vii) a soda lime absorber column, which supported the stability and accuracy of the CO2 detector, were replaced, when necessary (PP Systems, 2004). Simultaneously with the measurement of RS, we measured soil temperature (TS), using a soil temperature sensor (Digi-Sense Type K, Cole-Parmer Instrument Company), and SWCV, using a time-domain-reflectometry sensor (Hydrosense, Campbell Scientific), at a 12.5-cm depth. Soil respiration at each site was determined with simple exponential functions dependent on the TS for each site

RS = a exp (bTS ) (a and b are site-dependent parameters). Temperature dependency, defined as the rate of RS change as a consequence of an increase in TS of 10°C, was calculated based on the exponent parameter b of the simple exponential function [Q10 = exp(10b)]. The site-specific seasonal TS values in 2010, which were continuously logged using independent temperature loggers (TidbiT v2, Onset), were applied to the temperature-dependent functions for estimating annual soil RS. In addition, an RS function using TS and SWCV across all sites was developed based on the nonlinear regression function of Mielnick and Dugas (2000) e

RS = a exp (bTS )(SWCV + c )(d - SWCV )

1808

(a–e are parameters across all sites). Various RS functions using multiple variables (Davidson et al., 1998; Mielnick and Dugas, 2000; Lee et al., 2002; Tang and Baldocchi, 2005; Webster et al., 2008, 2009) were preliminarily tested, and the function suggested by Mielnick and Dugas (2000) was determined to be most appropriate for describing our data.

Statistical Analysis Differences in litter mass remaining, litter chemical properties, and RS among the sites and sampling times were tested using two-way analysis of variance with Duncan’s multiple range test. A nonlinear regression analysis was performed for fitting litter mass remaining and RS. A two-tailed Z-test was pairwisely performed to determine significant differences in parameters driven by nonlinear regression fitting such as the decomposition rate constant and Q10. The performance of the different litter decomposition functions and RS functions were evaluated using the Akaike information criterion (AIC), which was able to present a measure for both the goodness and the simplicity of the model fitting at the same time. Litter decomposition studies generally collected relatively small sample numbers; therefore, a corrected AIC (AICC) was applied (Rovira and Rovira, 2010; Manzoni et al., 2012). From either an AIC or an AICC value of each function, the lowest value of the functions was subtracted for the ease of comparison. All statistical analyses were conducted using SAS 9.2 software (SAS Institute, 2009).

Results

Litter Mass Remaining After 50 mo, the mass remaining of leaf litter was lowest at DS (9.6%), followed by PDS (19.3%), SSS (27.6%), and US (31.9%) (P < 0.01; Fig. 3). At PDS and SSS, most of the mass loss occurred during the first 18 mo and the mass loss was reduced after that time. However, the mass loss occurred continuously during the entire study period at both US and DS. Each decomposition function distinctively described the mass loss patterns at each site (Fig. 3). First, Olson’s simple exponential function determined the decomposition rate constant (kE; yr−1) as follows: PDS (1.181) > SSS (0.950) = DS (0.911) > US (0.528) (P < 0.05; Table 3). The decomposition rate constants represented the general pattern of the litter mass loss during the whole decomposition period. However, large underestimates of the remaining mass, ranging from 7.3% (DS) to 25.7% (SSS) were observed after 4 yr (Fig. 3). Soil Science Society of America Journal

predicted the change in the decomposition rate as time passed (Fig. 3). The decomposition rate (kR ; yr−1) steeply increased during the first 6 mo after decomposition, reached its highest value (1.14–2.16), and then steadily decreased, nearly to zero (Fig. 4). The decomposition rates estimated by the rational function (kR) could adequately describe the temporal change of the observed decomposition rates

x x  - ln  t +1 t   t +1 t  among the litter bag retrieval intervals (R2 = 0.711; inset in Fig. 4).

Carbon and Nitrogen Contents in Leaf Litter The C and N contents and the C/N ratio by mass in the leaf litter were all significantly different among the sites (P < 0.001), the sampling times (P < 0.001), and their interactions (P < 0.001). In particular, the mean C content in the initial leaf litter was 49.3% without significant differences among the sites (P = 0.11; Fig. 5a). However, the C content in the leaf litter after 4 yr was higher in US (52.6%) than PDS (47.5%) and DS (47.1%); that in SSS (50.0%) did not differ from those in the others (P < 0.01; Fig. 5a). The N content and C/N ratio of the initial leaf were in the order of PDS (2.6%) > DS (2.4%) = SSS (2.4%) > US (2.3%) (P < 0.001; Fig. 5b) and US (21.8) > DS (20.8) = SSS (20.6) > PDS (18.6) (P < 0.001; Fig. 5c), respectively. In stark contrast, there was no significant difference in the N content or C/N ratio in the leaf litter after 2 and 4 yr, respectively (P > 0.05). The C content was relatively constant at 50% (Fig. 5a), while the N content increased from 2.4% at the beginning to 3.3% (Fig. 5b) after 4 yr. In addition, the mean C/N ratio changed from 20.5 to 15.0 concomitant with the changes in C and N contents (Fig. 5c).

Fig. 3. The litter mass of Japanese alder leaves remaining under different soil moisture conditions during the 4 yr of litter bag incubation. The pattern of litter mass loss at each site (US: upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site) was fitted into three different litter decomposition functions: the simple exponential function (Olson, 1963), asymptotic function (Berg et al., 1996), and rational function (Rovira and Rovira, 2010). Vertical bars indicate standard errors of means (N = 5).

Second, Berg’s asymptotic function described the limit value of litter mass loss (Fig. 3). The limit value (m) was determined to be 70.5% for US, 91.7% for DS, 89.0% for PDS, and 76.9% for SSS (Table 3). In comparison with Olson’s simple exponential function, this function notably reduced the difference between measured and estimated values, from −0.2% (DS) to −8.2% (PDS), after 4 yr. Third, the rational function from Rovira and Rovira (2010) described the most realistic and efficient estimation of the mass remaining. This function (R2 = 0.97–0.99) presented the best fit among the functions (R2 = 0.94–0.99) (Table 3). In addition, AICC values were lowest in the rational function (Table 3). The difference between the measured and estimated values was relatively small, not only after 4 yr (mean difference: −0.4%) but also after 3 mo (mean difference: 1.4%). Moreover, it accurately

Table 3. The parameters and model performance statistics of the three different litter decomposition functions for Japanese alder leaves under different soil moisture conditions as shown in Fig. 3. From a corrected Akaike information criterion (AICC) value of each function, the lowest AICC value (min) was subtracted to evaluate the quality of the model. Simple exponential function

Site† kE

R2

AICC − min(AICC)

Asymptotic function kA

m

R2

Rational function AICC − min(AICC)

US 0.528 (0.037)‡ 0.971 34.33 0.751 (0.060) 0.705 (0.034) 0.987 5.69 DS 0.911 (0.052) 0.980 11.10 0.988 (0.072) 0.917 (0.040) 0.982 9.60 PDS 1.181 (0.087) 0.956 36.54 1.316 (0.124) 0.890 (0.041) 0.964 31.75 SSS 0.950 (0.084) 0.938 31.46 1.269 (0.144) 0.769 (0.040) 0.965 11.06 † US, upland site; DS, drained site; PDS, poorly drained site; SSS, surface saturated site. ‡ Numbers in parentheses indicate standard errors.

www.soils.org/publications/sssaj

a

b

R2

AICC − min(AICC)

0.955 (0.066) 1.161 (0.081) 1.175 (0.077) 0.948 (0.083)

0.208 (0.033) 0.227 (0.034) 0.205 (0.028) 0.151 (0.030)

0.988 0.986 0.984 0.974

0.000 0.000 0.000 0.000

1809

Soil Respiration The RS (mg C m−2 h−1) ranged seasonally from 12.6 to 355.1 for US, from 7.1 to 324.0 for DS, from 5.5 to 220.9 for PDS, and from 0.0 to 153.8 for SSS (Fig. 6a). The RS was significantly different among sites (P < 0.001) and sampling times (P < 0.001). The mean RS was highest in the order of US (158.1) = DS (137.4) > PDS (98.9) > SSS (70.8) (P < 0.001). The RS exponentially responded to the seasonal change in TS at each site (Fig. 6a). The Q10 values of US, DS, PDS, and SSS were determined to be 2.69, 2.61, 2.48, and 2.56, respectively; there was no significant difference among the sites (P > 0.05). In addition, the continuous RS (Fig. 6b) was predicted by the site-specific RS models using TS (Fig. 6a). The annual RS (Mg C ha−1 yr−1) was estimated to be 12.52 for US, 12.26 for DS, 8.64 for PDS, and Fig. 4. The change in the rate of decomposition of Japanese alder leaf litter under 6.45 for SSS. Regression models estimating RS across all sites different soil moisture conditions estimated by the rational function (Rovira and Rovira, 2010; Tables 2 and 3). The estimated decomposition rates were validated by were developed using the environmental variables the observed decomposition rates between the litter bag retrieval intervals (inset) (US: TS and SWCV (Table 4). The value of R2 was high- upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site). est and AIC was lowest in the TS and SWCV-based higher than other wetland types such as marshes, fens, bogs, regression model. The TS and SWCV-based regression model and even artificial wetlands (Atkinson and Cairns, 2001; (0.866) increased the R2 by 0.053 from the TS-based regression Kang and Freeman, 2009). model (0.813). The R2 of SWCV-based regression model was only 0.295. The current results indicate that the litter decomposition rate in wetlands is faster than that in uplands (Fig. 7; Atkinson and Discussion Cairns, 2001; Kang and Freeman, 2009; this study), even though Litter Decomposition Rate Constant litter decomposition is generally expected to be slower under anThe litter decomposition rate constant is a measure generally aerobic conditions. Higher decomposition rate constants under used in studies for describing how fast litter is decomposed. We more hydric conditions have also been observed in other alder compiled several meta-datasets to present litter decomposition forests (Dilly and Munch, 1996; Edmonds and Tuttle, 2010). rate constants in global ecosystems (Zhang et al., 2008), Korean Moreover, within the wetland sites, the decomposition rate was forests (Kim, 2012), alder forests (Edmonds and Tuttle, 2010), highest in PDS (under periodic saturation) than SSS (under conand wetlands (Atkinson and Cairns, 2001; Kang and Freeman, stant saturation) and DS (under unsaturated condition). 2009) (Fig. 7). A couple of significant findings were observed on The mechanism involved in the faster litter decomposicomparing the litter decomposition rate constants obtained in tion process under more hydric conditions is not clear; we have our study and those of other studies as follows: speculated that the physical environment and litter chemical properties may favor litter decomposition under hydric condi1. The decomposition rate constants at our wetland sites tions. First, temporal or periodic saturation may enhance litter (kE = 0.911–1.181) were quite high. For example, the litter decomposition compared to unsaturation or constant saturation decomposition rate constant in DS, PDS, and SSS ranked (Neckles and Neill, 1994; Atkinson and Cairns, 2001; Bedford, in the 75th, 87th, and 78th percentiles, respectively, of the 2005). Both O2 and moisture availabilities for litter decomposiglobal decomposition rate constant dataset synthesized by Zhang et al. (2008). Moreover, the decomposition rate tion might be optimized in PDS under periodic saturation. On constants in our wetland sites were much higher than those the other hand, lower SWCV in US (Fig. 2c) might limit the in Korean forests (kE = 0.268–0.488; Kim, 2012). moisture availability of the litter. Second, the N content (Fig. 5b) and C/N ratio (Fig. 5c) in initial leaf litter may contribute 2. Our litter decomposition rate constants corresponded to to the early-stage litter decomposition rate (Fig. 8). The litter dethose for alder forests and for swamps. Our median values composition rate constant was probably correlated with the N (kE; yr−1) of the upland (0.528) and the wetland sites (0.950) were similar to those for upland (0.600) and riparian (0.995) content (R = 0.899; P < 0.10; Fig. 8a) and with the C/N ratio alder forests, respectively, reported by Edmonds and Tuttle (R = −0.917; P < 0.10; Fig. 8b) in the initial leaf litter. Nitrogen (2010). In addition, the values of the wetland sites were is generally determined to be one of the most important factors similar to those of swamps, ranging from 0.639 (the 25th affecting litter decomposition in the early stages (Berg et al., percentile) to 1.427 (the 75th percentile), and were much 1996; Berg, 2000; Vestgarden, 2001; Zhang et al., 2008), and dif1810


tent and the litter decomposition rate constant was highest (Fig. 8a). In addition, fast litter decomposition has generally been observed in alder species (Fig. 7; Edmonds and Tuttle, 2010), which have a higher N leaf content (greater than ~2%; Kjøller et al., 1985; Dilly and Munch, 1996; this study) than other species.

Limit Value of Litter Mass Loss

Fig. 5. (a) Carbon and (b) N contents and (c) the C/N ratio of Japanese alder leaf litter under different soil moisture conditions during the 4 yr of litter bag incubation (US: upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site). Vertical bars indicate standard errors of means (N = 5). Asterisks indicate significant differences among the sites at the same sampling time (P < 0.05).

ferent site N availabilities (Ågren et al., 2001; Vestgarden, 2001) or species that have different leaf chemical properties (Canhoto and Graça, 1996; Moro and Domingo, 2000) could affect litter decomposition rates. Indeed, the leaf litter N content even within same species could vary, depending on a site N availability (Sariyildiz and Anderson, 2003; Tibbets and Molles, 2005) and flooding regime (Tibbets and Molles, 2005; Lecerf and Chauvet, 2008). Therefore, we speculate that the different soil N statuses due to different soil moisture conditions in this study may affect the initial leaf N status and, thus, the litter decomposition rate. For instance, the highest inorganic soil N content was observed in PDS (23.8 mg kg−1; Table 1), where the initial leaf N conwww.soils.org/publications/sssaj

The limit values of litter mass loss at each site were clearly observed (Fig. 3; Table 3) and fitted using asymptotic functions. Unlike the assumption of Olson’s simple exponential function, numerous field studies have reported that litter mass loss neither occurred with a steady decomposition rate nor proceeded to complete decomposition (Berg et al., 1996; Ågren et al., 2001; Prescott, 2005; Edmonds and Tuttle, 2010). Applying the simple exponential function, the underestimation of mass remaining after 4 yr was quite significant (7.3–25.7%; Fig. 3). Instead, litter mass loss seemed to stop at some point and reach a constant value at the late decomposition stage. Berg et al. (1993) first hypothesized this limit value of litter mass loss and Berg et al. (1996) described the limit value concept using an asymptotic function. Several studies have supported and adopted the limit value concept centered around Berg’s group (e.g., Coûteaux et al., 1995; Prescott, 2005; Virzo De Santo et al., 2009; Berg et al., 2010; Edmonds and Tuttle, 2010), yet uncertainties remain about what the limit value really describes, represents, or implies. The limit value is assumed to have originated from (i) the change in the litter chemical properties (e.g., lignin, Mn, N, C) to the point where what remains is resistant to decomposition and (ii) the lack of soil fauna infiltration for complete decomposition due to the artifact of litter bag (Berg et al., 1996; Prescott, 2005). However, investigating the process by which the litter mass loss reaches the limit value is still challenging. Nevertheless, the lower limit of litter mass loss was actually observed in field studies included in this study, and the asymptotic function could provide the degree of the limit value. In addition, the limit value can imply the potentially accumulated humus converted from the annual litter input mass (Berg and McClaugherty, 2008). Lower limit values at US (70.5%) and SSS (76.9%) were observed compared to those at DS and PDS (~90%); however, we have no explanation for why 29.5 and 23.1% of the litter mass at US and SSS, respectively, were undecomposed and deposited on the soil surface as a stable humified residue. Two possible processes can be speculated, although the lack of lignin measurement in this study limits the clear understanding of the process. First, the Mn content in the initial leaf litter was lower at SSS (0.78 g kg−1) than the other sites (1.50–1.63 g kg−1) (P < 0.001; Table 1); the lower Mn availability may reduce ligninase synthesis at SSS. On the other hand, the higher fresh litter N content in PDS (Table 1) could promote ligninase synthesis and may increase the limit value (Berg, 2000). Second, we observed that most litter bags in SSS were found under approximately 5 to 20 cm of soil after 4 yr. Saturation and floods may inhibit the decomposition process in litter that is covered by 1811

Fig. 6. (a) The soil CO2 efflux (RS) response to the seasonal soil temperature change under different soil moisture conditions in a Japanese alder forest, with simple exponential functions applied for fitting the response of the RS to the soil temperature change (N = 51–55, P < 0.001); vertical bars indicate standard errors of means (N = 5); and (b) predicted seasonal RS using the simple exponential functions and climate data in 2010; soil temperatures at each sampling time were averaged (US: upland site, DS: drained site, PDS: poorly drained site, SSS: surface saturated site).

highest R2 and the lowest AICC values (Table 3) supported that this function could precisely describe the litter mass loss patterns with few parameters as assuming that two out of total four parameters were constant (Rovira and Rovira, 2010). Therefore, the rational function may be the best for realistically and efficiently estimating mass loss. However, the parameters in this function may fail to represent or imply significant ecological information concerning the litter decomposition process, unlike the former functions (e.g., decomposition rate constant and limit value of decomposition).

soil. In addition, the potentially accumulated humus mass (Mg C ha−1 yr−1) was estimated as 1.19 for US, 0.20 for DS, 0.24 for PDS, and 0.23 for SSS using the limit values and the annual litter input (unpublished data, 2013).

Applicability of Different Litter Decomposition Functions These current results could support the conclusion that all these litter decomposition functions have respective advantages in terms of explaining the characteristics of the mass loss pattern: 1. Olson’s simple exponential function primarily informs the basic status of litter decomposition. The litter decomposition rate constant can easily be obtained and was comparable to that determined in other studies (Fig. 7). However, the lowest R2 and the highest AICC values (Table 3) indicated that the simple exponential function was least accurate at describing the mass loss patterns. It could not explain the inhibited mass loss in the later decomposition stages (Fig. 3). Moreover, Prescott (2005) questioned whether the litter decomposition rate may be less informative than the decomposition limit value for studying C fluxes and nutrient cycling. 2. Berg’s asymptotic function describes the inhibited mass loss at the late decomposition stage (Fig. 3). It could provide the limit value of the litter mass loss (Table 3). The model performance was also slightly improved compared to the simple exponential function (Table 3). Determining the limit value could be useful for revealing the contribution of the litter decomposition process to soil organic matter. However, knowledge of its definition, processes, and implications is insufficient. 3. The rational function from Rovira and Rovira (2010) could simulate the change in the decomposition rate constant consistently throughout decomposition stages (Fig. 4). It is supported by the fact that litter quality and environmental change could alter decomposition rates. For example, the function successfully described the inhibited mass loss of the first 3 mo due to low temperatures in the winter (Fig. 3). The 1812

Soil Respiration

Soil respiration was strongly suppressed by high SWCV (Fig. 6). We speculate that soil moisture may influence both autotrophic and heterotrophic RS. First, under higher soil moisture, the lower vegetation biomass that could be assumed from the lower tree density (Table 1) would reduce autotrophic RS (von Arnold et al., 2005). Second, anaerobic conditions could reduce soil microbial activities. Although heterotrophic RS was not directly measured, reduced enzyme activities and net N transformation rates (unpublished data, 2013) would support the reduced heterotrophic RS under hydric conditions. Lower RS in wetlands has generally been reported in other studies (Savage and Davidson, 2001; Webster et al., 2008, 2009; Fissore et al., 2009). Although a lack of measurements of autotrophic and heterotrophic RS Table 4. Regression models for estimating soil CO2 efflux (RS) based on soil temperature (TS) and volumetric soil water content (SWCV) in a Japanese alder forest. From an Akaike information criterion (AIC) value of each function, the lowest AIC value (min) was subtracted to evaluate the quality of the model. Parameter

Equation

R2

AIC − min(AIC)

TS

RS = 30.68exp(0.1043TS)

0.813

64.1

SWCV

RS = 0.000376(SWCV + 2997.3) ´ (153.4 − SWCV)0.7507

0.295

349.5

TS ´ SWCV

RS = 0.000466exp(0.0959TS) ´ (SWCV + 2152.6) ´ (154.5 − SWCV)0.772

0.866

0.0


to 2.7, which corresponded to the mean Q10 value of 2.7 in a global temperate dataset (Chen and Tian, 2005) and in a global wetland dataset (Bond-Lamberty and Thomson, 2012). The fact that no difference in Q10 values was found in our study demonstrates that the temperature dependency was independent from soil moisture conditions. Other wetland studies have reported that soil moisture may (Silvola et al., 1996; Inglett et al., 2012; Taggart et al., 2012) or may not (Fissore et al., 2009; Vicca et al., 2009) affect Q10 values. One possibility is that soil moisture may not directly affect temperature sensitivity but may instead influence it through changing the soil temperature or substrate quality. This may be more reasonable because soil temperatures and substrate quality are the factors whose effects on temperature dependency have been supported by theoretical modeling (Davidson et al., 2006; Sierra, 2012) and meta-analysis (Chen and Tian, 2005). For example, Wickland et al. (2010) reported a large difference in Q10 values (3.8–6.6) among boreal forests with different soil moisture levels and simultaneously different soil temperature conditions. However, the soil temperatures in Fig. 7. Comparison of litter decomposition rate constants between this this study did not significantly differ among the sites (P > 0.05; study and other studies integrating metadata: global synthesis (Zhang data not shown), and it could be one possible reason for there et al., 2008), Korean forests (Kim, 2012), alder forests (Edmonds and Tuttle, 2010), and wetlands (Atkinson and Cairns, 2001; Kang and being no difference in the Q10 values. The limited number of Freeman, 2009). Boxes indicate the range of values between the 25th studies that have reported on Q10 values in wetlands makes it difand 75th percentiles. The black and white lines in the boxes indicate ficult to determine the general response of Q10 to different soil the median and the mean values, respectively. Error bars above and moisture conditions. below a box indicate the 10th and 90th percentiles. Points above and below a box indicate 5th and 95th percentiles. Our RS regression model using TS and SWCV performed successfully, supported by the lowest AIC value among the other would make it difficult to understand detailed processes, the reducing effect of soil moisture on RS was obvious. models (Table 4). The coefficient of determination (R2) of our Soil respiration increased with increasing TS (Fig. 6), as model (0.867) was higher than those of Mielnick and Dugas most RS studies have found (Chen and Tian, 2005; Hibbard (2000) (0.52), Webster et al. (2008) (0.723), Li et al. (2008) et al., 2005; Ryan and Law, 2005; Davidson et al., 2006). (0.67), and Webster et al. (2009) (0.51–0.71). Homogenous Temperature dependency values ranged from approximately 2.5 vegetation composition dominated by Japanese alder might have reduced the variability of RS, which was not explained by TS and SWCV. In addition, our model described a nearly linear decrease of RS alongside an SWCV increase, even though the quadratic function was applied. However, the effect of SWCV on RS was generally modeled as a quadratic pattern that described a maximum RS at an optimum SWCV of approximately 10 to 30% (Davidson et al., 1998; Mielnick and Dugas, 2000; Lee et al., 2002; Webster et al., 2008; Taggart et al., 2012). This is due to the abundant moisture in our study sites, which may not meet dry conditions limiting RS; for instance, SWCV was barely below 10% even in US (Fig. 2c). Soil respiration regression models using TS and soil moisture have contributed to precisely estimating site-specific RS with temporal and spatial variability (Davidson et al., 1998; Mielnick and Dugas, 2000; Lee et al., 2002; Tang and Baldocchi, 2005; Webster et al., 2008, 2009). However, the addition Fig. 8. The linear relationships between (a) the N content and (b) the C/N ratio in of a SWC variable into the T -dependent model V S initial leaf litter and the litter decomposition rate constant (kE) of Japanese alder leaves. may not considerably increase the explanation of Bidirectional bars indicate standard errors of means (N = 5) (US: upland site, DS: the variance. The TS and SWCV-dependent model drained site, PDS: poorly drained site, SSS: surface saturated site). www.soils.org/publications/sssaj

1813

increased the explanation of variance by only 5% compared to the TS dependent model (Table 4). Webster et al. (2009) also reported that the SWCV-dependent models (R2 = 0.51–0.58) increased the explanation of variance by 2 to 9% the TS-dependent model (R2 = 0.48–0.49). Moreover, the limited mechanistic basis of these models restricts their applicability, implication, and harmonization. Responses of the RS to soil moisture changes are generally modeled from empirical fitting. Even though TS and soil moisture would dominantly contribute to the temporal variability and spatial variability in RS, respectively (Graf et al., 2012), the addition of a soil moisture variable into the RS model may not simply satisfy modeling and understanding of the soil moisture effects on RS. Here, the DAMM model, which was recently suggested by Davidson et al. (2012), could open a new opportunity based on a mechanistic understanding of dual Arrhenius and Michaelis–Menten (i.e., DAMM) kinetics. In this model, the soil moisture effect on heterotrophic RS was modeled, first via substrate diffusion, then O2 availability using the Michaelis–Menten function. Unfortunately, unfractionated RS observations in this study could not be applied to the DAMM model; nevertheless, further field studies should consider adopting the DAMM model.

Conclusions Differences in litter decomposition and RS along a soil moisture gradient were presented. The significant findings are summarized as follows: 1. The wetter site in general had a higher level of litter decomposition (Hypothesis 1 partly rejected) and a lower RS (Hypothesis 1 partly supported). However, soil moisture may or may not be linearly related to litter decomposition and RS. For example, the highest litter decomposition rates occurred at PDS, not SSS (the most hydric), and RS did not differ between US and DS. The current results indicate that the influence of soil moisture on litter decomposition and RS may be considerably more complex and interactively related to various environmental variables, such as the wet–dry regime, water potential, O2 and nutrient availability, and/or sediment accumulation (Xiong and Nilsson, 1997; Luo and Zhou, 2006; Cook and Orchard, 2008). 2. Each of these litter decomposition functions may have respective advantages for describing the characteristics of mass loss patterns. Olson’s simple exponential function can be easily obtained and is comparable to that determined in other studies; however, it is not appropriate for explaining the inhibited mass loss in the late decomposition stage. On the other hand, Berg’s asymptotic function and the rational function proposed by Rovira and Rovira (2010) may realistically estimate the mass loss pattern, especially during the later decomposition stages. The former function is good for estimating the limit value of the litter mass loss, while the latter function has the most realistic fitting for predicting the change in the decomposition rate constant across decomposition stages. 3. The temperature dependency of RS did not differ among 1814

the sites (Hypothesis 2 rejected). Nevertheless, we conjecture that this report could be valuable for supporting the response of temperature dependency of RS to soil moisture, for which there has been insufficient evidence to date. The temporal variation of RS can be explained by understanding that temperature was dominant, while spatial variation explained by the level of soil moisture was relatively small.

Acknowledgments

This work was supported by the National Research Foundation of Korea (2010-0020227), the Korea Environmental Industry & Technology Institute (C314-00131-0408, G214-00181-0401), and the Korea University Grant (2013). We appreciate the cooperation of Heolleung Office, Culture Heritage Administration of Korea for conserving this precious ecosystem and allowing us to access HELCA. We also appreciate Ah Reum Lee, Joomi Kim, Seongjun Kim, and Yujin Noh for their assistance in the laboratory and field.

References

Ågren, G., E. Bosatta, and A. Magill. 2001. Combining theory and experiment to understand effects of inorganic nitrogen on litter decomposition. Oecologia 128:94–98. doi:10.1007/s004420100646 Atkinson, R., and J. Cairns. 2001. Plant decomposition and litter accumulation in depressional wetlands: Functional performance of two wetland age classes that were created via excavation. Wetlands 21:354–362. doi:10.1672/0277-5212(2001)021[0354:PDALAI]2.0.CO;2 Bedford, A. 2005. Decomposition of Phragmites australis litter in seasonally flooded and exposed areas of a managed reedbed. Wetlands 25:713–720. doi:10.1672/0277-5212(2005)025[0713:DOPALI]2.0.CO;2 Berg, B. 2000. Litter decomposition and organic matter turnover in northern forest soils. For. Ecol. Manage. 133:13–22. doi:10.1016/S0378-1127(99)00294-7 Berg, B., M.P. Berg, P. Bottner, E. Box, A. Breymeyer, R.C. Anta, et al. 1993. Litter mass loss rates in pine forests of Europe and Eastern United States: Some relationships with climate and litter quality. Biogeochemistry 20:127–159. doi:10.1007/BF00000785 Berg, B., M.P. Davey, A. De Marco, B. Emmett, M. Faituri, S.E. Hobbie, et al. 2010. Factors influencing limit values for pine needle litter decomposition: A synthesis for boreal and temperate pine forest systems. Biogeochemistry 100:57–73. doi:10.1007/s10533-009-9404-y Berg, B., M.-B. Johansson, G. Ekbohm, C. McClaugherty, F. Rutigliano, and A. Virzo De Santo. 1996. Maximum decomposition limits of forest litter types: A synthesis. Can. J. Bot. 74:659–672. doi:10.1139/b96-084 Berg, B., and C. McClaugherty. 2008. Plant litter: Decomposition, humus formation, carbon sequestration. Springer-Verlag, Berlin. Bond-Lamberty, B.P., and A.M. Thomson. 2012. A global database of soil respiration data, version 2.0. [Data set.] Oak Ridge Natl. Lab. Distributed Active Archive Ctr., Oak Ridge, TN. Canhoto, C., and M.S. Graça. 1996. Decomposition of Eucalyptus globulus leaves and three native leaf species (Alnus glutinosa, Castanea sativa and Quercus faginea) in a Portuguese low order stream. Hydrobiologia 333:79–85. doi:10.1007/BF00017570 Chen, H., and H.-Q. Tian. 2005. Does a general temperature-dependent Q10 model of soil respiration exist at biome and global scale? J. Integr. Plant Biol. 47:1288–1302. doi:10.1111/j.1744-7909.2005.00211.x Cook, F.J., and V.A. Orchard. 2008. Relationships between soil respiration and soil moisture. Soil Biol. Biochem. 40:1013–1018. doi:10.1016/j.soilbio.2007.12.012 Coûteaux, M.-M., P. Bottner, and B. Berg. 1995. Litter decomposition, climate and litter quality. Trends Ecol. Evol. 10:63–66. doi:10.1016/S0169-5347(00)88978-8 Davidson, E.A., E. Belk, and R.D. Boone. 1998. Soil water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest. Global Change Biol. 4:217–227. doi:10.1046/j.1365-2486.1998.00128.x Davidson, E.A., and I.A. Janssens. 2006. Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440:165–173.


doi:10.1038/nature04514 Davidson, E.A., I.A. Janssens, and Y. Luo. 2006. On the variability of respiration in terrestrial ecosystems: Moving beyond Q10. Global Change Biol. 12:154–164. doi:10.1111/j.1365-2486.2005.01065.x Davidson, E.A., S. Samanta, S.S. Caramori, and K. Savage. 2012. The Dual Arrhenius and Michaelis–Menten kinetics model for decomposition of soil organic matter at hourly to seasonal time scales. Global Change Biol. 18:371–384. doi:10.1111/j.1365-2486.2011.02546.x Davidson, E.A., K. Savage, L.V. Verchot, and R. Navarro. 2002. Minimizing artifacts and biases in chamber-based measurements of soil respiration. Agric. For. Meteorol. 113:21–37. doi:10.1016/S0168-1923(02)00100-4 Dilly, O., and J.-C. Munch. 1996. Microbial biomass content, basal respiration and enzyme activities during the course of decomposition of leaf litter in a black alder (Alnus glutinosa (L.) Gaertn.) forest. Soil Biol. Biochem. 28:1073–1081. doi:10.1016/0038-0717(96)00075-2 Edmonds, R.L., and K.M. Tuttle. 2010. Red alder leaf decomposition and nutrient release in alder and conifer riparian patches in western Washington, USA. For. Ecol. Manage. 259:2375–2381. doi:10.1016/j.foreco.2010.03.011 Erwin, K. 2009. Wetlands and global climate change: The role of wetland restoration in a changing world. Wetlands Ecol. Manage. 17:71–84. doi:10.1007/s11273-008-9119-1 Eswaran, H., E. Van Den Berg, and P. Reich. 1993. Organic carbon in soils of the world. Soil Sci. Soc. Am. J. 57:192–194. doi:10.2136/sssaj1993.03615995005700010034x Fissore, C., C.P. Giardina, R.K. Kolka, and C.C. Trettin. 2009. Soil organic carbon quality in forested mineral wetlands at different mean annual temperature. Soil Biol. Biochem. 41:458–466. doi:10.1016/j.soilbio.2008.11.004 Graf, A., M. Herbst, L. Weihermüller, J.A. Huisman, N. Prolingheuer, L. Bornemann, and H. Vereecken. 2012. Analyzing spatiotemporal variability of heterotrophic soil respiration at the field scale using orthogonal functions. Geoderma 181–182:91–101. doi:10.1016/j.geoderma.2012.02.016 Hattenschwiler, S., A.V. Tiunov, and S. Scheu. 2005. Biodiversity and litter decomposition in terrestrial ecosystems. Annu. Rev. Ecol. Evol. Syst. 36:191–218. doi:10.1146/annurev.ecolsys.36.112904.151932 Hibbard, K.A., B.E. Law, M. Reichstein, and J. Sulzman. 2005. An analysis of soil respiration across northern hemisphere temperate ecosystems. Biogeochemistry 73:29–70. doi:10.1007/s10533-004-2946-0 Inglett, K.S., P.W. Inglett, K.R. Reddy, and T.Z. Osborne. 2012. Temperature sensitivity of greenhouse gas production in wetland soils of different vegetation. Biogeochemistry 108:77–90. doi:10.1007/s10533-011-9573-3 Kang, H., and C. Freeman. 2009. Soil enzyme analysis for leaf litter decomposition in global wetlands. Commun. Soil Sci. Plant. Anal. 40:3323–3334. doi:10.1080/00103620903325935 Kim, J.G. 2012. Accumulated organic matter, litterfall production, and decomposition tell us the status of litter dynamics in forests. J. Ecol. Field Biol. 35:99–109. doi:10.5141/JEFB.2012.014 Kjøller, A., S. Struwe, and K. Vestberg. 1985. Bacterial dynamics during decomposition of alder litter. Soil Biol. Biochem. 17:463–468. doi:10.1016/0038-0717(85)90010-0 Lee, M.-S., K. Nakane, T. Nakatsubo, W.-H. Mo, and H. Koizumi. 2002. Effects of rainfall events on soil CO2 flux in a cool temperate deciduous broad-leaved forest. Ecol. Res. 17:401–409. doi:10.1046/j.1440-1703.2002.00498.x Lecerf, A., and E. Chauvet. 2008. Intraspecific variability in leaf traits strongly affects alder leaf decomposition in a stream. Basic Appl. Ecol. 9:598–605. doi:10.1016/j.baae.2007.11.003 Li, H.-L., J.-X. Yan, X.-F. Yue, and M.-B. Wang. 2008. Significance of soil temperature and moisture for soil respiration in a Chinese mountain area. Agric. For. Meteorol. 148:490–503. doi:10.1016/j.agrformet.2007.10.009 Luo, Y., and X. Zhou. 2006. Soil respiration and the environment. Academic Press, Burlington, MA. Manzoni, S., G. Piñeiro, R.B. Jackson, E.G. Jobbágy, J.H. Kim, and A. Porporato. 2012. Analytical models of soil and litter decomposition: Solutions for mass loss and time-dependent decay rates. Soil Biol. Biochem. 50:66–76. doi:10.1016/j.soilbio.2012.02.029 Manzoni, S., and A. Porporato. 2009. Soil carbon and nitrogen mineralization: Theory and models across scales. Soil Biol. Biochem. 41:1355–1379. doi:10.1016/j.soilbio.2009.02.031 Mielnick, P.C., and W.A. Dugas. 2000. Soil CO2 flux in a tallgrass prairie. Soil Biol. Biochem. 32:221–228. doi:10.1016/S0038-0717(99)00150-9 Mitsch, W., B. Bernal, A. Nahlik, Ü. Mander, L. Zhang, C. Anderson, et al. 2013.

www.soils.org/publications/sssaj

Wetlands, carbon, and climate change. Landscape Ecol. 28:583–597. doi:10.1007/s10980-012-9758-8 Moro, M.J., and F. Domingo. 2000. Litter decomposition in four woody species in a Mediterranean climate: Weight loss, N and P dynamics. Ann. Bot. 86:1065–1071. doi:10.1006/anbo.2000.1269 Moyano, F.E., S. Manzoni, and C. Chenu. 2013. Responses of soil heterotrophic respiration to moisture availability: An exploration of processes and models. Soil Biol. Biochem. 59:72–85. doi:10.1016/j.soilbio.2013.01.002 Neckles, H., and C. Neill. 1994. Hydrologic control of litter decomposition in seasonally flooded prairie marshes. Hydrobiologia 286:155–165. doi:10.1007/BF00006247 Noh, N.J., Y. Son, S.K. Lee, T.K. Yoon, K.W. Seo, C. Kim, et al. 2010. Influence of stand density on soil CO2 efflux for a Pinus densiflora forest in Korea. J. Plant Res. 123:411–419. doi:10.1007/s10265-010-0331-8 Olson, J.S. 1963. Energy storage and the balance of producers and decomposers in ecological systems. Ecology 44:322–331. doi:10.2307/1932179 Porporato, A., P. D’Odorico, F. Laio, and I. Rodriguez-Iturbe. 2003. Hydrologic controls on soil carbon and nitrogen cycles. I. Modeling scheme. Adv. Water Resour. 26:45–58. doi:10.1016/S0309-1708(02)00094-5 PP Systems. 2004. EGM-4 environmental gas monitor for CO2: Operator’s manual version 4.14. PP Systems, Hitchin, UK. PP Systems. 2005. SRC-1/CPY-2 closed system chambers for use with all EGM’s (1/2/3/4) and CIRAS-1: Operator’s manual version 3.31. PP Systems, Hitchin, UK. Prescott, C.E. 2005. Do rates of litter decomposition tell us anything we really need to know? For. Ecol. Manage. 220:66–74. doi:10.1016/j.foreco.2005.08.005 Robinson, D.A., C.S. Campbell, J.W. Hopmans, B.K. Hornbuckle, S.B. Jones, R. Knight, et al. 2008. Soil moisture measurement for ecological and hydrological watershed-scale observatories: A review. Vadose Zone J. 7:358–389. doi:10.2136/vzj2007.0143 Rodeghiero, M., and A. Cescatti. 2008. Spatial variability and optimal sampling strategy of soil respiration. For. Ecol. Manage. 255:106–112. doi:10.1016/j.foreco.2007.08.025 Rodriguez-Iturbe, I. 2000. Ecohydrology: A hydrologic perspective of climate-soil-vegetation dynamics. Water Resour. Res. 36:3–9. doi:10.1029/1999WR900210 Rovira, P., and R. Rovira. 2010. Fitting litter decomposition datasets to mathematical curves: Towards a generalised exponential approach. Geoderma 155:329–343. doi:10.1016/j.geoderma.2009.11.033 Ryan, M.G., and B.E. Law. 2005. Interpreting, measuring, and modeling soil respiration. Biogeochemistry 73:3–27. doi:10.1007/s10533-004-5167-7 Sariyildiz, T., and J.M. Anderson. 2003. Interactions between litter quality, decomposition and soil fertility: A laboratory study. Soil Biol. Biochem. 35:391–399. doi:10.1016/S0038-0717(02)00290-0 SAS Institute. 2009. SAS/STAT 9.2 user’s guide. SAS Inst., Cary, NC. Savage, K.E., and E.A. Davidson. 2001. Interannual variation of soil respiration in two New England forests. Glob. Biogeochem. Cycles 15:337–350. doi:10.1029/1999GB001248 Schlesinger, W.H., and J.A. Andrews. 2000. Soil respiration and the global carbon cycle. Biogeochemistry 48:7–20. doi:10.1023/A:1006247623877 Sierra, C. 2012. Temperature sensitivity of organic matter decomposition in the Arrhenius equation: Some theoretical considerations. Biogeochemistry 108:1–15. doi:10.1007/s10533-011-9596-9 Silvola, J., J. Alm, U. Ahlholm, H. Nykanen, and P.J. Martikainen. 1996. CO2 fluxes from peat in boreal mires under varying temperature and moisture conditions. J. Ecol. 84:219–228. doi:10.2307/2261357 Soil Survey Staff. 1999. Soil taxonomy: A basic system of soil classification for making and interpreting soil surveys. 2nd ed. Agric. Handbk. 436. U.S. Gov. Print. Office, Washington, DC. Taggart, M., J.L. Heitman, W. Shi, and M. Vepraskas. 2012. Temperature and water content effects on carbon mineralization for sapric soil material. Wetlands 32:939–944. doi:10.1007/s13157-012-0327-3 Tang, J., and D.D. Baldocchi. 2005. Spatial–temporal variation in soil respiration in an oak–grass savanna ecosystem in California and its partitioning into autotrophic and heterotrophic components. Biogeochemistry 73:183– 207. doi:10.1007/s10533-004-5889-6 Tibbets, T.M., and M.C. Molles. 2005. C: N: P stoichiometry of dominant riparian trees and arthropods along the Middle Rio Grande. Freshwater Biol. 50:1882–1894. doi:10.1111/j.1365-2427.2005.01465.x Vestgarden, L.S. 2001. Carbon and nitrogen turnover in the early stage of

1815

Scots pine (Pinus sylvestris L.) needle litter decomposition: Effects of internal and external nitrogen. Soil Biol. Biochem. 33:465–474. doi:10.1016/S0038-0717(00)00187-5 Vicca, S., I.A. Janssens, H. Flessa, S. Fiedler, and H.F. Jungkunst. 2009. Temperature dependence of greenhouse gas emissions from three hydromorphic soils at different groundwater levels. Geobiology 7:465– 476. doi:10.1111/j.1472-4669.2009.00205.x Virzo De Santo, A., A. Marco, A. Fierro, B. Berg, and F. Rutigliano. 2009. Factors regulating litter mass loss and lignin degradation in late decomposition stages. Plant Soil 318:217–228. doi:10.1007/s11104-008-9831-x von Arnold, K., M. Nilsson, B. Hånell, P. Weslien, and L. Klemedtsson. 2005. Fluxes of CO2, CH4 and N2O from drained organic soils in deciduous forests. Soil Biol. Biochem. 37:1059–1071. doi:10.1016/j.soilbio.2004.11.004 Webster, K.L., I.F. Creed, R.A. Bourbonnière, and F.D. Beall. 2008. Controls on the heterogeneity of soil respiration in a tolerant hardwood forest. J. Geophys. Res. 113:G03018. Webster, K.L., I.F. Creed, M.D. Skowronski, and Y.H. Kaheil. 2009. Comparison of the performance of statistical models that predict soil respiration from

1816

forests. Soil Sci. Soc. Am. J. 73:1157–1167. doi:10.2136/sssaj2008.0310 Western, A.W., R.B. Grayson, and G. Bloschl. 2002. Scaling of soil moisture: A hydrologic perspective. Annu. Rev. Earth Planet. Sci. 30:149–180. doi:10.1146/annurev.earth.30.091201.140434 Wickland, K.P., J.C. Neff, and J.W. Harden. 2010. The role of soil drainage class in carbon dioxide exchange and decomposition in boreal black spruce (Picea mariana) forest stands. Can. J. For. Res. 40:2123–2134. doi:10.1139/X10-163 Xiong, S., and C. Nilsson. 1997. Dynamics of leaf litter accumulation and its effects on riparian vegetation: A review. Bot. Rev. 63:240–264. doi:10.1007/BF02857951 Yoon, T.K., N.J. Noh, S. Han, H. Kwak, W.-K. Lee, and Y. Son. 2013. Small-scale spatial variability of soil properties in a Korean swamp. Lands. Ecol. Eng. doi:10.1007/s11355-013-0236-5 Zhang, D., D. Hui, Y. Luo, and G. Zhou. 2008. Rates of litter decomposition in terrestrial ecosystems: Global patterns and controlling factors. J. Plant Ecol. 1:85–93. doi:10.1093/jpe/rtn002
