Seasonal Variation in Soil Greenhouse Gas Emissions at Three Age-Stages of Dawn Redwood (Metasequoia glyptostroboides) Stands in an Alluvial Island, Eastern China

Greenhouse gas (GHG) emissions are an important part of the carbon (C) and nitrogen (N) cycle in forest soil. However, soil greenhouse gas emissions in dawn redwood (Metasequoia glyptostroboides) stands of different ages are poorly understood. To elucidate the effect of plantation age and environmental factors on soil GHG emissions, we used static chamber/gas chromatography (GC) system to measure soil GHG emissions in an alluvial island in eastern China for two consecutive years. The soil was a source of CO2 and N2O and a sink of CH4 with annual emissions of 5.5–7.1 Mg C ha−1 year−1, 0.15–0.36 kg N ha−1 year−1, and 1.7–4.5 kg C ha−1 year−1, respectively. A clear exponential correlation was found between soil temperature and CO2 emission, but a negative linear correlation was found between soil water content and CO2 emission. Soil temperature had a significantly positive effect on CH4 uptake and N2O emission, whereas no significant correlation was found between CH4 uptake and soil water content, and N2O emission and soil water content. These results implied that older forest stands might cause more GHG emissions from the soil into the atmosphere because of higher litter/root biomass and soil carbon/nitrogen content compared with younger stands.


Introduction
Establishment and management of forest plantations play an increasingly important role in sequestrating carbon from the atmosphere as one of the major strategies for mitigating global warming. The emissions of greenhouse gases (GHGs) are mostly related to the carbon (C) and nitrogen (N) cycle from forest soils. Forest soils are the sink of carbon in the world and contain about 704 Pg C, with varying C densities under different environmental conditions [1]. On the contrary, they are also the source of N 2 O [1,2]. In some countries (e.g., China, India, Russian Fedration, US, Japan, etc.), plantations represent an important part of the national forested areas, and are increasing at the Dongping National Forest Park is the largest forest farm in eastern China, with 70% of the total area covered by dawn redwood plantations. Since the 1960s, plantations have been established to form different aged stands. In order to examine the effects of stand age on soil GHG emission, three different aged stands of 10, 17, and 32 years old were selected. In each stand, three plots (20 m × 20 m) were set up in August 2011 (Table 1).
Biomass carbon storage. In 2011, all trees were counted at all sites. The height of every single tree was determined by using a Haglöf Vertex III Ultrasonic Hypsometer. The diameter at breast height (1.3 m above the ground) (DBH) was measured using a measuring tape. The whole tree dry biomass was calculated by Becuwe's allometric functions (M = 0.06291 DBH 2.4841 ), and carbon stock in the stands was estimated by considering the carbon contents of tree dry biomasses (around 50%) [25].
Soil properties. To determine the bulk density, pH, total carbon (C), and nitrogen (N) concentrations of the soil in the stand, three soil samples were collected from each plot. Soil bulk density was obtained by the volumetric ring method [26]. Soil pH was measured by 1:5 dry soil: CaCl2 solution (0.01 M) [27]. The total soil C and N concentrations were determined by using an elemental analysis-stable isotope ratio mass spectrometer (Vario ELIII Elementar, Hessen Langenselbold, Germany).  Dongping National Forest Park is the largest forest farm in eastern China, with 70% of the total area covered by dawn redwood plantations. Since the 1960s, plantations have been established to form different aged stands. In order to examine the effects of stand age on soil GHG emission, three different aged stands of 10, 17, and 32 years old were selected. In each stand, three plots (20 m × 20 m) were set up in August 2011 (Table 1).
Biomass carbon storage. In 2011, all trees were counted at all sites. The height of every single tree was determined by using a Haglöf Vertex III Ultrasonic Hypsometer. The diameter at breast height (1.3 m above the ground) (DBH) was measured using a measuring tape. The whole tree dry biomass was calculated by Becuwe's allometric functions (M = 0.06291 DBH 2.4841 ), and carbon stock in the stands was estimated by considering the carbon contents of tree dry biomasses (around 50%) [27].
Soil properties. To determine the bulk density, pH, total carbon (C), and nitrogen (N) concentrations of the soil in the stand, three soil samples were collected from each plot. Soil bulk density was obtained by the volumetric ring method [28]. Soil pH was measured by 1:5 dry soil: CaCl 2 solution (0.01 M) [29]. The total soil C and N concentrations were determined by using an elemental analysis-stable isotope ratio mass spectrometer (Vario ELIII Elementar, Hessen Langenselbold, Germany). Note: (a) The source of litter data was Xiao's dissertation [30]. DBH, diameter at breast height; SOC, soil organic carbon.

Soil Gas Emissions
Gas emission measurements were based on Forestry Standards "Observation Methodology for Long-term Forest Ecosystem Research" of PR China (LY/T 1952-2011). Because the forest sites were relatively homogeneous, three observation points were systematically arranged in each stand. The static chamber method was employed to measure soil CO 2 , CH 4 , and N 2 O emissions. Gas emissions were measured every two weeks (September 2011-September 2013).
The static chamber consisted of two parts. First, the stainless steel based part (0.5 m × 0.5 m × 0.2 m) was permanently inserted at a 10 cm depth in the soil for each observation point of the plots, and the second upper part was made of a polyvinyl chloride plate with a size of 0.5 m × 0.5 m × 0.5 m. A fan was installed in each upper chamber for air mixing. Next, 30 min after closing the chamber, gas samples were collected with a gastight syringe (100 mL) every 10 min for the next 40 min (0, 10 min, 20 min, 30 min, and 40 min). Five gas samples at each observation point were taken between 9:00 a.m. and 12:00 a.m. and analyzed by gas chromatography (6890N, Agilent, Santa Clara, CA, USA) with an Electron Capture Detector (ECD) for N 2 O detection and an Flame Ionization Detector (FID) for CH 4 and CO 2 detection [31,32]. The minimum detectable limit of CO 2 , CH 4 , and N 2 O fluxes were 0.3 mg C m −2 h −1 , 4.4 µg C m −2 h −1 , and 0.3 µg N m −2 h −1 , respectively [33]. The gas emissions were calculated by the rate of gas concentration change during sampling. The calculation details were as follows.
where F is the gas emissions (mg m −2 h −1 for CO 2 and CH 4 , and µg m −2 h −1 for N 2 O), and dC dt (µL L −1 min −1 for CO 2 and CH 4 , and nL L −1 min −1 for N 2 O) is the emission rate of CO 2 , CH 4 , or N 2 O concentration in the chamber. A linear regression is used to calculate the emission rate. The m (g mol −1 ) is the molecular weight of trace gas. P indicates the atmospheric pressure (P = 1.013 × 10 5 Pa). R is the gas constant (R = 8.314 J mol −1 K −1 ). T (K) is the air temperature in the chamber. V (cm 3 ), H (cm), and A (cm 2 ) are the volume, height, and area of the static chamber, respectively.

Soil Temperature and Soil Water Content
The probe of digital thermometer JM 624 (Jinming Insturment Co., LTD, Tianjin, China) was inserted at 5 and 10 cm soil depth to detect the soil temperature on the outside of each chamber when we collected the gas samples. Soil samples were taken by soil auger from 0 cm to 10 cm and 10 cm to 20 cm depths to determine soil water contents gravimetrically by measuring the fresh and dry weights after drying in an oven at 105 • C for two days.

Data Analysis
Generally, the growing season of dawn redwood in Shanghai is from May to November, and the non-growing season is from December to April. We split our observed data into two parts according to the growing or non-growing season to determine whether soil respiration increases simultaneously with increasing photosynthesis.

Q 10 Values
The temperature sensitivity of the soil respiration rate at the three stands was calculated by a non-linear regression model with the van't Hoff function, as follows: where R S is the soil respiration (mg CO 2 m −2 h −1 ), α and β are fitted constants, and T is soil temperature, which was measured at 5 cm and 10 cm depths in the soil [34,35]. Q 10 is the factor explaining the temperature sensitivity of soil respiration, and it is calculated as follows: [36,37].

The Relationship between GHG Emissions and Environmental Factors
One-sample Kolmogorov-Smirnov testing was used to determine whether the GHG emissions, soil temperature, and soil moisture were normally distributed. Soil temperature and soil moisture were normally distributed. Data variation among the sites was tested for significance by using the Duncan test following ANOVA. Pearson correlation analyses were used to analyze the relationship between greenhouse gas and the environment factors. Statistical analysis was conducted using IBM SPSS Statistics 21 software.
Canonical correspondence analysis (CCA) was conducting by using the CCA procedure in PAST 3 to detect the relationship between soil GHG emissions and environmental factors, such as soil temperature, soil water content, soil C and N concentration, and foliage C and N concentrations. A plot of the first two canonical variables (Can 1 and Can 2) was made to visually show the correlation among gases and environmental variables.

Soil CH 4 Uptake
The soil was a sink of CH 4 in all three stands, with the highest uptake of CH 4 occurring in the summer (Figure 2). During 2011-2013, the mean soil CH 4 uptake rates were 0.026 mg m −2 h −1 , 0.032 mg m −2 h −1 , and 0.069 mg m −2 h −1 in the 10, 17, and 32-year-old stands, respectively ( Table 2). The CH 4 uptake rates were significantly higher in the older stand compared to the younger stands (p < 0.05). The highest CH 4 uptakes were measured in the growing season ( Figure 3).

Soil N 2 O Emission
There were large differences in N 2

Soil CH4 Uptake
The soil was a sink of CH4 in all three stands, with the highest uptake of CH4 occurring in the summer ( Figure 2). During 2011-2013, the mean soil CH4 uptake rates were 0.026 mg m −2 h −1 , 0.032 mg m −2 h −1 , and 0.069 mg m −2 h −1 in the 10, 17, and 32-year-old stands, respectively ( Table 2). The CH4 uptake rates were significantly higher in the older stand compared to the younger stands (p < 0.05). The highest CH4 uptakes were measured in the growing season ( Figure 3).

Soil N2O Emission
There were large differences in N2O emissions among the three stands, ranging from −19.

Annual GHG Emissions
The annual CO 2 emissions were significantly higher in the 32-year-old stand compared to the other two younger stands (p < 0.05) (Figure 4). The emissions were 23.3% and 20.0% higher in the 32-year-old stand than those in the 10 and 17-year-old stands, respectively. Moreover, the annual soil CH 4 uptake had significant differences among the three stands. The annual CH 4 uptake was highest in the 32-year-old stand and lowest in the 10-year-old stand.
The highest annual soil N 2 O emission was observed in the 32-year-old stand and we noted that the 32-year-old stand had a 56.8% higher annual N 2 O emission than the 10-year-old stand and a 17.7% higher annual emission than the 17-year-old stand. However, the N 2 O emissions among the three stands were not significantly different. 32-year-old stand than those in the 10 and 17-year-old stands, respectively. Moreover, the annual soil CH4 uptake had significant differences among the three stands. The annual CH4 uptake was highest in the 32-year-old stand and lowest in the 10-year-old stand.
The highest annual soil N2O emission was observed in the 32-year-old stand and we noted that the 32-year-old stand had a 56.8% higher annual N2O emission than the 10-year-old stand and a 17.7% higher annual emission than the 17-year-old stand. However, the N2O emissions among the three stands were not significantly different.

The Effect of Soil Temperature on GHG Emissions
In this research, soil CO2 emissions increased exponentially with soil temperature both at 5 cm and at 10 cm soil depths (RS = 62.78e 0.075T at 5 cm soil depth, and RS = 61.89e 0.077T at 10 cm soil depth). The exponential model could explain 68% or 69% (p < 0.001) of the seasonal variation in soil CO2 emissions ( Table 3). The Q10 values were calculated to be 2.12 and 2.15 at 5 cm and at 10 cm soil depths, respectively (Table 3). Usually, Q10-values were almost 3%-51% higher in the non-growing season than in the growing season. Table 3. Parameters of the exponential model for soil CO2 emissions as a function of soil temperature at 5 and 10 cm depths in the three stands.

The Effect of Soil Temperature on GHG Emissions
In this research, soil CO 2 emissions increased exponentially with soil temperature both at 5 cm and at 10 cm soil depths (R S = 62.78e 0.075T at 5 cm soil depth, and R S = 61.89e 0.077T at 10 cm soil depth). The exponential model could explain 68% or 69% (p < 0.001) of the seasonal variation in soil CO 2 emissions ( Table 3). The Q 10 values were calculated to be 2.12 and 2.15 at 5 cm and at 10 cm soil depths, respectively (Table 3). Usually, Q 10 -values were almost 3%-51% higher in the non-growing season than in the growing season. Table 3. Parameters of the exponential model for soil CO 2 emissions as a function of soil temperature at 5 and 10 cm depths in the three stands.    Table 4). Note: ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). T 5 cm and T 10 cm mean soil temperature at 5 cm soil depth and at 10 cm soil depth, respectively. SWC 0-10 cm and SWC 10-20 cm mean soil water content at 0-10 cm soil depth and at 10-20 cm soil depth, respectively.

Effects of Soil Water Content on GHG Emissions
Soil water content contributed substantially to the GHG emissions. The relationship between soil CO 2 emissions and soil water content at both 0-10 cm and 10-20 cm depths was negative. However, no significant relationship was found between CH 4 emission and soil water content, or N 2 O emission and soil water content. (Table 4).

The Main Influencing Factors of Soil Greenhouse Gas Emissions
The variations in vegetation carbon, nitrogen, and soil properties were described by two significant canonical components (explaining 100% of the variance) ( Figure 5). The first, Can 1, accounted for 98.65% of the total variance and was highly related to the trees' biomass, and C and N content in soil and foliage. Can 2 accounted for 1.21% of the total variance with close correlation among soil water content and soil temperature. The CO 2 and N 2 O emissions, and CH 4 uptake all have positive correlations with Can 1 and negative correlations with Can 2.

Soil Carbon Dynamic in Different-Age Stands
The soil was a source of CO2 and sink of CH4 in the three stands in both growing and non-growing seasons. The annual soil CO2 emissions (5.5-7.1 Mg C ha −1 year −1 ) were within the same range observed in other subtropical forests. For instance, annual soil CO2 emission was 3.1-7.3 Mg C ha −1 year −1 in the seasonal tropical primary forests in Xishuangbanna region, southwest China, and from 3.1-7.3 to 11.1-12.0 Mg C ha −1 year −1 in the subtropical forests [9,36]. In subtropical and tropical forests, annual soil CH4 uptake rates ranged from 0.8 kg C ha −1 year −1 to 4.3 kg C ha −1 year −1 [12,16,37]. Our study showed a similar uptake (1.7 kg C ha −1 year −1 to 4.5 kg C ha −1 year −1 ) in plantations located in northern subtropical areas, thereby suggesting that annual CH4 uptake does not significantly vary with subtropical or tropical biomes.

Soil Carbon Dynamic in Different-Age Stands
The soil was a source of CO 2 and sink of CH 4 in the three stands in both growing and non-growing seasons. The annual soil CO 2 emissions (5.5-7.1 Mg C ha −1 year −1 ) were within the same range observed in other subtropical forests. For instance, annual soil CO 2 emission was 3.1-7.3 Mg C ha −1 year −1 in the seasonal tropical primary forests in Xishuangbanna region, southwest China, and from 3.1-7.3 to 11.1-12.0 Mg C ha −1 year −1 in the subtropical forests [9,38]. In subtropical and tropical forests, annual soil CH 4 uptake rates ranged from 0.8 kg C ha −1 year −1 to 4.3 kg C ha −1 year −1 [12,16,39]. Our study showed a similar uptake (1.7 kg C ha −1 year −1 to 4.5 kg C ha −1 year −1 ) in plantations located in northern subtropical areas, thereby suggesting that annual CH 4 uptake does not significantly vary with subtropical or tropical biomes.
Soil CO 2 significantly varied with soil temperature and water content in the three stands in both growing and non-growing seasons. A positive relationship existed between soil temperature and CO 2 emission in these three stands, and a negative relationship was found between soil water content and CO 2 emission. The effects of soil temperature and soil water content on CO 2 emissions were statistically confounded. As such, we excluded the soil temperature effect through normalizing the soil respiration values with R S = 62.78e 0.075T at 5 cm soil depth and R S = 61.89e 0.077T at 10 cm soil depth, and found that the effect of soil water content on CO 2 emissions was not significantly negative (with Pearson correlation from −0.18 to −0.19). Respiration rates generally decreased with decreasing water content. Soil temperature was probably the key factor regulating soil respiration. However, soil water content also restricted soil respiration [40]. Both soil CO 2 emission and CH 4 uptake peaked in the period of May-November because of the wet-hot climate. The laboratory and field studies have verified that soil temperature and soil water content could account for most of the seasonal variation in soil CO 2 emission and CH 4 uptake [40][41][42].
Soil temperature and water content explained 76%-87% of soil CO 2 emission and 67%-75% of total annual emission in the wet season (April to September) of lower subtropical forests [6]. Q 10 , an exponential relationship, has been commonly used to estimate soil respiration rates from soil temperature [36]. In previous literature, the mean Q 10 values were 2.14 for tropical regions and 2.26 for temperate regions [43]. In our study, Q 10 ranged from 1.9 to 2.4 during the whole year, and soil respiration in the non-growing season was more sensitive to soil temperature. The higher Q 10 in the non-growing season could be associated with the phonological cycle of photosynthesis as compared to the growing season, which has consequences on the belowground carbon allocation. In the summer, about 50% or more of the soil CO 2 emissions could be originated from recently assimilated C, which trees allocate to the belowground system (root and rhizosphere) [44]. The values of Q 10 increased with soil depth, and this result was the same as that obtained by Pavelka [45]. The seasonal variation in soil temperature was lower in the deeper layers and soil respiration rate was relatively more sensitive to temperature fluctuations [46]. During the growing and non-growing seasons the different values of Q 10 were noted with different R 2 values, and the lower R 2 values were calculated in the growing season. During the growing season, soil temperature causes little changes in soil CO 2 emissions. The primary reason might be the low temperature amplitude during the growing season. Second, the other factors (except soil temperature) could explain the soil CO 2 emission such as the changes in photosynthesis and precipitation.
The soil temperature positively affected CH 4 uptake, and no significant relationship existed between CH 4 uptake and soil water content. Kiese and Werner observed that CH 4 uptake was negatively correlated with soil temperature and soil water content [38,39]. In mid-subtropical China, the highest CH 4 uptake (17.12 g C ha −1 day −1 ) occurred in the summer-autumn season with increasing soil temperature and water content, but the relationships between CH 4 uptake and soil temperature and CH 4 uptake and soil water content were not significant [47]. In earlier studies, CH 4 uptake had decreased with increasing soil water content during the summer season [48,49]. Maximum CH 4 uptake rate was clearly associated with the lowest soil moisture and the highest soil temperature both in temperate and tropical forests [50]. Before oxidization by methanotrophs, the soil CH 4 was emitted from anaerobic environments to the atmosphere. In the forest's soil, a certain amount of CH 4 from the atmosphere was consumed by methanotrophs [51]. The optimum conditions for growth of methanotrophic bacteria and induction of methane oxidation activity were 20%-35% water contents and 25 • C-35 • C temperatures [52]. In our study, the water content ranged from 11% to 33%, which was almost in the optimum range, and temperatures showed a larger range from 1.4 • C to 30 • C. Soil temperature could be more important than water content in regulating CH 4 consumption in this study, which is in agreement with the results of previous reports [53].

Soil Nitrogen Dynamic in Different-Aged Stands
We observed highly dynamic N 2 O emissions with low values in our study (i.e., 0.81-1.87 g N ha −1 day −1 ), which were lower than some previously reported emissions. For example, our results are similar to the N 2 O emissions from undrained forests in southern Sweden (i.e., 1.62 g N ha −1 day −1 ) [54], but they are substantially lower than the 8.77 g N ha −1 day −1 previously recorded in the subtropical forest in southern China [12].
A seasonal variation in N 2 O emissions has been reported in tropical and subtropical forests. For instance, the highest N 2 O emissions have been observed during the spring and summer months with mean values of 2-5 g N ha −1 day −1 . The lowest emissions were obtained during winter seasons, with less than 0.5 g N ha −1 day −1 [9]. The higher N 2 O emissions were emitted from temperate and tropical forest ecosystems during the wet and hot season [50]. The magnitude of N 2 O emissions was very closely linked to rainfall events [55]. The soil N 2 O was produced by microbes through nitrification in aerobic conditions and through denitrification under anaerobic conditions [56]. Factor, such as precipitation, was observed to exert some influence on the soil aeration, but soil aeration could affect N 2 O production. In our study, the highest soil N 2 O emissions were observed between May and November when higher rainfall occurred with a mean value of 2.04 g N ha −1 day −1 . The lowest soil N 2 O emissions were recorded between December and April with a mean value of 0.75 g N ha −1 day −1 . N 2 O emissions showed a positive correlation with soil temperature; no significant correlation with soil water content was observed, which was similar to a previous study in Japan [57]. However, some previous reports have shown that N 2 O emissions have a positive correlation with soil temperature and soil water content [42].

Factors Affecting Soil Greenhouse Gas Emissions
The present study showed that soil GHG emissions differed among the three stands. The 32-year-old stand had significantly higher CO 2 emissions, CH 4 uptake, and N 2 O emissions than the 10 and 17-year-old stands. Basically, these three stands differed in biomass/litter carbon storage, nitrogen content, and soil properties. The soil CO 2 , N 2 O, and CH 4 were produced by microbial activity, and these processes were controlled by environmental factors [58,59].
Forest soil CO 2 emissions were the sum of heterotrophic (microbes) and autotrophic respiration (roots), and the contribution of root respiration rates which were higher during the growing season [60]. The soil CO 2 emissions were a good indicator of total below-ground allocation of carbon and of ecosystem productivity. Among these stands, the older stands maintained higher productivity than the younger stands; it was not surprising that the older stand had the highest rates of soil respiration. Older stands released higher CO 2 , and the major difference was that the older stand had higher soil carbon, which could probably reflect higher root and litter carbon storage [61]. The research in Loess Plateau of China [62] indicated that 48% of the variations in annual soil CO 2 emissions were explained by the combined carbon stock in top soil and litter, 77% by the root carbon stock, and 63% by the combined carbon stock in roots, litter, and top soil. The aboveground litter mineralization and decomposition contributed to about 8% of the soil respiration in a subtropical Montane cloud forest in Taiwan [63]. In our study, the total carbon storage of litter, soil, and roots in the older stands was higher than the two younger stands, which indicated higher annual CO 2 emissions in the older stands. Based on the principal component analyses, the litter composition was an important stimulator for soil CO 2 emissions because of the simultaneous effects on production and consumption of the soil surface organic matter [64].
Methane emissions of soils were correlated with microbial activities, and the upper soil layer were generally CH 4 sinks [65]. The rate of CH 4 uptake was regulated by the soil C and N levels as well as soil water content, and there was a close link between labile C, N, and CH 4 uptake in forest soils [66,67]. This research has shown that carbon and nitrogen contents of litter, soil, and root in older stands were higher than in younger stands, which indicates higher annual CH 4 uptake in older stands.
In contrast to the pattern of soil CO 2 and CH 4 emissions, no distinctly different trend in N 2 O emission was observed among differently aged stands. According to the reported study, soil N 2 O production and consumption were mainly influenced by the amount of mineral N in soils, and low N availability was linked with N 2 O emissions [2]. Highly dynamic emissions of N 2 O were found among different forest soil types [68]. The primary controlling factors of N 2 O production were found to be soil pH and C/N ratio, and these soil properties could explain most of the variability of N 2 O emissions [9,69]. However, we used three stands in our study but the results indicating similar annual N 2 O emissions despite the different soil properties.

Conclusions
Soil respiration in each of the stands was strongly and positively related to soil temperature, and negatively related to soil water content. The soil CH 4 uptake was positively related to soil temperature, and soil N 2 O emission had a positive relation with soil temperature. Affected by the annual climatic conditions (e.g., temperature and precipitation), soil respiration showed a clear seasonal variation, with high emissions in the wet-hot season (from May to November) and low emissions in the dry-cool season (from December to April).
Different stages of forest stands strongly affected soil respiration and CH 4 emission rates through root respiration and/or microbial activities, but had no significant relationship with soil N 2 O emission. Carbon storage, nitrogen, and C/N ratio (soil, litter, and root) were the main factors affecting CH 4 uptake and N 2 O emission. Soil properties such as soil water content and soil pH were important indicators for soil respiration.