Introduction

Carbon dioxide (CO2) and methane (CH4) are important greenhouse gases (GHG). The concentrations of CO2 and CH4 in atmosphere increased from 280 ppm and 715 ppb in pre-industrial times to 379 ppm and 1,774 ppb in 2005, respectively (IPCC 2007). The levels of CO2 and CH4 have a significant impact on global warming. Therefore, there is a need for quantifying the potential of an individual ecosystem as a source or sink for atmospheric CO2 and CH4 (Purvaja and Ramesh 2001).

Coastal marsh ecosystem is characterized by high temporal and spatial variations including topographic feature, environmental factors, and astronomic tidal fluctuation, and is very sensitive to global climate changes and human activities (Sun et al. 2013). Considerable efforts have been invested in the past two decades to quantify the CO2 and CH4 fluxes in different coastal wetlands (Purvaja and Ramesh 2001; Allen et al. 2007; Cheng et al. 2007; Tong et al. 2012; Sun et al. 2013; Poffenbarger et al. 2011). However, most of the research focused on the emission of GHG from natural wetlands; data of GHG emission from anthropogenic coastal wetland is insufficient. As the development of economy, human activities, such as land-use changes and introduction of invasive alien plants, have more and more impact on coastal wetlands. The phenomenon of transformation of natural coastal wetlands into a harbor, seawall, industrial complex or urban district is very common, this transformation will change the geomorphology of the coastal line and the physical processes of the coastal system permanently, which can result more negative influence on the coastal environment and ecosystem (Bi et al. 2012). Changes in land use have a profound impact on GHG flux. Inubushi et al. (2003) suggested that converting a secondary forest peatland to paddy field increased the annual emissions of CO2 and CH4 to the atmosphere, while transforming the secondary forest to upland decreased the emissions.

Human-induced invasion by exotic-species also have a profound impact on the GHG flux. Invasion by exotic plant species has been considered to be one of the most serious problems for natural ecosystems (Walker and Smith 1997). Spartina alterniflora was introduced to China in 1979, to protect the coastal banks and stabilize the sediment along the eastern coast in Fujian province, Southeast China . Currently, S. alterniflora distributs widely along the east coast of China (Wang et al. 2006a) due to its faster growth rate compared to the native species (Qin and Zhong 1992; Wang et al. 2006b). The coverage of S. alterniflora was approximately 260 ha in six counties by 1985 and increased to more than 112,000 ha by 2,000 (An et al. 2007). Therefore, information on emission of GHG from ecosystem invaded by S. alterniflora is urgently needed, however, studies in this field were mainly conducted at the estuary in the southern part of China (Tong et al. 2012; Cheng et al. 2007; Cheng et al. 2010; Zhang et al. 2010), but information for the estuary in the northern part of China is largely unknown as yet. Thus, it is very important to evaluate how the invasion by exotic-species affects GHG emission in wetlands at estuary area in Northern China.

The Yellow River is well known as a sediment-laden river. Approximately 1.05 × 107 tons of sediment is carried to the estuary and deposited in the delta each year (Cui et al. 2009) resulting in vast area of floodplain and special wetland landscape (Xu et al. 2002; Wang et al. 2004). Typical reclaimed land patterns in the Yellow River estuary included harbor, seawall, salt pans, oilfield, aquaculture ponds and industrial complex. A recent study showed that the area of natural wetlands decreased by 44.5 % in four decades (1976–2008) in the Yellow River estuary, while constructed wetland increased by 1.997 × 104 hm2 in the same period due to rapid development of coastal aquaculture and salt industry (Chen et al. 2011). Wang et al. (2013) pointed that the exploitation of tidal flats resources and construction of artificial ponds related to holothurian culture in the Yellow River delta had become an emerging industry. And the field occupied by holothurian culture covered an area of 1.5 × 104 ha in the Yellow River delta. S. alterniflora was transplanted into Yellow River estuary for three times in 1985, 1987, and 1990. The coverage of S. alterniflora has now reached up to 614.59 hm2 (Zhu et al. 2012). Human activities have more and more impact in Yellow River estuary. Therefore, evaluating the influence of human activities on the emission of GHG is a big necessary in this area.

In this paper, we quantitatively evaluated the variations in the levels of CH4 and CO2 in a typical coastal marsh which was significantly influenced by human activities. The objectives of this study were to: (i) measure the emissions of CO2 and CH4 from exotic S. alterniflora; (ii) determine the relationship between GHG emission and environmental factors; and (iii) estimate the difference of the seasonal change in CO2 and CH4 emissions under different human activities.

Materials and Methods

Site Description

This study was conducted in the Yellow River estuary in Dongying City, Shandong Province, China. The Yellow River estuary has a typical continental monsoon climate with distinct seasons; summer is warm and rainy, and winter is cold. The annual average temperature is 12.1 °C and the frost-free period is 196 d. The average temperatures for spring, summer, autumn and winter are 10.7, 27.3, 13.1, and −5.2 °C, respectively. The mean tidal range of the irregular semidiurnal tide is 0.73 to 1.77 m in the intertidal zone of the Yellow River estuary. Soils in the study area are dominated by intrazonal tide and salt soil. The dissoluble salt content in surface layer (0 to 20 cm) of salt soil is very high (>8 g/kg), and its grain composition is dominated by sand and silt (50–80 %). The average annual evaporation and precipitation are 1962 and 551.6 mm, respectively, with about 70 % of the precipitation occurs in June to August. The main types of vegetation are Sueada salsa, Phragmites australis, Triarrhena sacchariflora, Myriophyllum spicatum, Tamarix chinensis, and Limoninum sinense.

Due to its rich oil and gas resources, Wuhaozhuang region, part of the Yellow River estuary, is significantly affected by human activities. Seawalls were constructed in this region to improve oil production and other economic activities (e.g. aquaculture). S. alterniflora was introduced to Wuhaozhuang in 1990 to protect the seawall from damage. Additionally, Wuhaozhuang is one of important aquaculture farms in Yellow River estuary. We selected four sampling sites in this region, which represented the four typical human-influenced areas in this region, including (a) the west side of the seawall (WSS) (east side of the seawall is the Bohai Sea) (38°01′8.79″N, 118°58′6.84″E); (b) oil field (OF, there are lots of oil wells on the ground) (38°01′8.73″N, 118°57′44.63″E); (c) S. alterniflora coastal marsh (SCM) (38°00′24.8″N, 118°58′23.2″E); and (d) aquaculture pond (ACP) (38°00′26.16″N, 118°58′23.0″E) (Fig. 1).

Fig. 1
figure 1

Sketch of the study area and experimental plots (black star) in the Yellow River estuary

Full size image

Experimental Design

Fluxes of CO2 and CH4 from WSS, OF, and SCM were measured using static, manual stainless steel chamber and gas chromatography techniques. A stainless steel base with a water groove on the top was inserted into the ground for 20 cm depth in May 2013. A chamber was placed into the groove during measurement; meanwhile water was injected into the groove to build an open-bottom square box. The outside of the chamber was covered with an insulating layer (2 cm thick) to reduce the impact of direct radiative heating during sampling, which can cause very little difference in temperature between the inside and the outside of the chamber (Teiter and Mander 2005; Søvik and Kløve 2007; Jiang et al. 2010; Tong et al. 2010; Sun et al. 2013, 2014). Air inside the chamber was circulated with battery-driven fans installed inside the chamber to make sure that the gas sample was uniform in the chamber. CH4 and CO2 emissions from ACP were measured using floating chambers and gas chromatography techniques. The floating chambers were made of opaque PVC. A flotation gear was installed to make sure that the chamber can float on water. Insulating layer and fans also placed for floating chambers.

Measurements were made in June, August, October, and December of 2013 at the four sites. Each measurement campaign consisted of 12 chambers set up at four positions (three chambers per site). Gas samples were collected at 7:00, 9:30, 12:00, 14:30; and 17:00 h on each sampling date, which have been shown to be the optimum measurement period by Sun et al. (2013, 2014). The gas samples were withdrawn from the headspace of the chamber in a 20-min interval (totally 60 min for each measurement) using a 50 ml syringe equipped with a three-way stopcock. Samples were injected into pre-evacuated packs and taken to the laboratory for determination within 36 h.

The gas samples were determined with gas chromatography (Agilent 7890A, Agilent Co., Santa Clara, CA, USA) equipped with FID. The CH4 was separated from the other gases with a 2 m stainless-steel column, with an inner diameter of 2-mm 13XMS column (60/80 mesh). The CO2 was separated with a 2 m stainless-steel column with an inner diameter of 2 mm Porapak Q (60/80 mesh). The FID operated at 200 °C using high-pure nitrogen as a carrier gas, at a flow rate of 30 ml/min. The column temperatures were maintained at 55 °C for all separations. The greenhouse gas concentrations were quantified by comparing the peak areas of samples against standards. During the gas measurement, standards were analyzed every eight samples of determination to ensure the data quality, the relative standard deviation (RSD) for each sample should below 6 %. The gas flux was calculated using the following equation (Song et al. 2008):

$$ J=\frac{dc}{dt}\times \frac{M}{V_0}\times \frac{P}{P_0}\times \frac{T}{T_0}\times H $$

where J is the gas flux (mg m−2 h−1), dc/dt is the slope of the gas concentration curve variation, along with time, M is the mole mass of each gas, P is the atmospheric pressure at the sampling site, T is the absolute temperature during sampling, V 0 , T 0 and P 0 are respectively, the gas mole volume, air absolute temperate and atmospheric pressure under standard conditions, H is the height of chamber above the water surface. The rates of CH4 and CO2 emissions were calculated by fitting the changes in the determined concentrations of CH4 and CO2 over a 60-min period to a linear model. The regression concentration coefficients from linear regressions were rejected when R2 was less than 0.9. Positive values indicate net flux to the atmosphere (efflux), and negative values indicate consumption of atmosphere gases by the soil (influx).

Environmental data were measured at each site during sampling. Air temperature inside the chamber was measured with a thermometer inserted into the chamber. Soil temperature at 5, 10, 15, 20, and 25 cm depth was measured with five ground thermometers inserted into the corresponding depth. On each sampling date, three soil samples per layer (0 to 10, 10 to 20, and 20 to 30 cm) were collected at each site to determine soil water content. Water temperature was also measured with the thermometer during gas sampling. In August 2013, the aboveground biomass in WSS, OF and SCM was estimated. Three quadrants (50 × 50 cm for OF and SCM, 5 × 5 m for WSS) were selected for biomass measurement at each of the three sites. Biomass was oven-dried (80 °C) to a constant weight.

Statistical Analysis

Statistical analyses were conducted using SPSS 16.0 and Origin 7.5. The results were presented as a mean of replicates, with standard error (SE). Significant differences in GHG emissions and environmental factors between different sites were determined by one-way analysis of variance [ANOVA, followed by Tukey’s Honest Significant Difference (HSD) test]. Correlation analysis was conducted to examine the relationship between fluxes and the measured environmental variables. In all tests, differences were considered significant when p < 0.05.

Results

Plant Growth

Vegetation in WSS is predominated by Tamarix chinensis (>60 %), while that in OF and SCM are Suaeda salsa (>99 %) and S. alterniflora (>99 %), respectively. The coverage and maximum aboveground biomass of T. chinensis, S. salsa, and S. alterniflora are 10, 70, 95 %, and 200.11 ± 15.82 (mean ± SE) g m−2, 376.862 ± 31.50 g m−2, 1281.92 ± 176.93 g m−2, respectively. The aboveground biomass of S. alterniflora was greater than these of T. chinensis and S. salsa.

Variation in CO2 Fluxes

CO2 flux includes respiration from living aboveground and belowground plant parts as well as aerobic and anaerobic microbial activities in the soil column. This CO2 flux can be called ecosystem respiration and is associated with the overall carbon flow of the ecosystem (Nykänen et al. 1998). CO2 fluxes from the four sites ranged from −181.23 to 878.03 mg CO2 m−2 h−1 over the entire sampling period (Fig. 2). Average CO2 fluxes in WSS, OF, SCM and ACP from June to December were 125.36, 111.03, 241.97 and −39.49 mg CO2 m−2 h−1, respectively. All sites except ACP (negative value) released CO2 during the entire experimental period. The CO2 flux rates from WSS, OF, and SCM showed the similar seasonal pattern, initial increase followed by a subsequent fall. The greatest CO2 flux rates from WSS (334.69 mg CO2 m−2 h−1), OF (264.32 mg CO2 m−2 h−1), and SCM (583.07 mg CO2 m−2 h−1) were observed in August while the smallest CO2 flux rates (27.08, 13.41, and 37.22 mg CO2 m−2 h−1, respectively) were observed in December. A significantly greater CO2 flux was observed from SCM than WSS (p = 0.028), OF (p = 0.016), and ACP (p = 0.000). The CO2 flux from ACP varied significantly from June to December, and smaller than that from WSS (p = 0.002), OF (p = 0.006) and SCM (p = 0.000). The greatest consumption was observed from ACP in August (−71.14 mg CO2 m−2 h−1).

Fig. 2
figure 2

Carbon dioxide (CO2) fluxes from WSS (west side of the seawall), OF (Oil field), SCM (Spartina alterniflora coastal marsh) and ACP (aquaculture pond) in different months in the Yellow River estuary

Full size image

Variation in CH4 Fluxes

CH4 fluxes from the four sites ranged from −0.2390 to 0.5252 mg CH4 m−2 h−1 (Fig. 3). Average CH4 fluxes in WSS, OF, SCM, and ACP from June to December were −0.0110, −0.0165, 0.2012, and 0.0034 mg CH4 m−2 h−1, respectively. The flux rate of CH4 from SCM was significantly greater than that from WSS (p = 0.000), OF (p = 0.000), and ACP (p = 0.000). During the entire experimental period, SCM was a net source of CH4, while both net emission and net consumption of CH4 occurred at other sites. The greatest CH4 flux rate from SCM (0.4107 mg CH4 m−2 h−1) was observed in August, while it was observed in October from OF (0.0157 mg CH4 m−2 h−1) and ACP (0.0165 mg CH4 m−2 h−1). The CH4 flux from WSS varied significantly from month to month during the measurement period, with the greatest consumption being observed in October (−0.0258 mg CH4 m−2 h−1).

Fig. 3
figure 3

Methane (CH4) fluxes from WSS (west side of the seawall), OF (Oil field), SCM (Spartina alterniflora coastal marsh) and ACP (aquaculture pond) in different months in the Yellow River estuary

Full size image

Environmental Factors

There was no significant difference in the air temperature among the four sites (p > 0.05), and so did in soil temperature among different layers (5, 10, 15 and 20 cm depth) at WSS, OF, and SCM (p > 0.05) (Table 1). Spearman correlation analysis indicated that most of the relationships between CO2 (or CH4) fluxes and air temperature (or soil temperature) from OF and ACP were not significant (p > 0.05) (Table 2). CO2 fluxes from SCM and WSS showed significantly positive relationships with air or soil temperature (p < 0.05). And there was no significant relationship between CH4 flux and air/soil temperature (p > 0.05) at the sampling sites except SCM. In addition, CH4 flux significantly correlated with soil water content (positive, p < 0.05), but not for the CO2 flux.

Table 1 Air, ground and water temperatures, and soil water contents of the four study sites in the Yellow River estuary
Full size table
Table 2 Spearman correlation analysis between CO2, CH4 fluxes and environmental factors
Full size table

Discussions

Seasonal Variation of CO2 and CH4 Fluxes

CO2 and CH4 emissions varied markedly among different seasons at the four sites (Figs. 2 and 3). Similar variations have been reported in previous studies (Allen et al. 2007; Song et al. 2008; Cheng et al. 2010; Sun et al. 2013). Chen et al. (2010) reported a seasonal variation in CO2 flux in a subtropical mangrove swamp in Hong Kong showed that seasonal variations and the flux in warm seasons was greater than in cold seasons. We also found that CO2 flux in warm seasons (June and August) was significantly greater than that in the cold seasons (October and December) in WSS, OF, and SCM. The significant relationship between CO2 flux and air temperature and soil temperature was consistent with those reported by Cheng et al. (2010) who found that CH4 emissions from S. alterniflora and S. mariqueter soils positively correlated with soil temperature. Similarly, Whalen (2005) observed that seasonal patterns of trace gas emissions were governed by seasonal variability in temperatures affecting water availability, production of substrate precursors, and microbial activity. No significant relationship were found between CO2 flux and air or soil temperature from OF, maybe due in part to the complex interactions of temperature and other biotic/abiotic factors, such as water content.

CH4 flux was not related to temperature significantly expect SCM. Sun et al. (2013) suggested that, in coastal marsh of the Yellow River estuary, seasonal variations in CH4 emission was not affected by seasonal variability in temperatures. In our study, the greatest CH4 emission rate from OF (0.0157 mg CH4 m−2 h−1) and WSS (−0.0258 mg CH4 m−2 h−1) were both observed in October. This result indicated that the influence of temperature on CH4 emission was masked by other biotic/abiotic factors, such as vegetation or soil moisture. Factors affecting CH4 emission are diverse and controlled by the interplay of CH4 production, oxidation, and transport processes (Ding et al. 2004). Kutzbach et al. (2004) reported that the ratio between CH4 production and oxidation is controlled by soil moisture which regulates the relative extent of oxic and anoxic environment within soils. Our result showed that CH4 emissions were positively correlated with soil moisture in WSS, OF, and SCM (Table 2).

CO2 and CH4 fluxes across the air-water interface of ACP had obvious seasonal variations (Figs. 2 and 3). Xing et al. (2005) also pointed that the fluxes of CH4 and CO2 showed strong seasonal dynamics from a shallow hypereutrophic subtropical Lake in China, CH4 emission rate was the greatest in summer, whereas CO2 was adsorbed from the atmosphere in spring and summer, but underwent a large-scale emission in winter. In our study, CO2 flux over the entire sampling period from ACP ranged from −71.14 to −3.96 mg CO2 m−2 h−1, indicating that ACP was a sink for CO2, with the greatest CO2 consumption occurred in August. Previously, 13CDIC and pCO2 measurements suggested that respiration and decomposition of organic sediments were the primary sources of CO2 in the water column (Striegl et al. 2001). Del Giorgio et al. (1999) reported that in the water column of temperate lakes the baseline respiration, fueled by allochthonous C, was independent of phytoplankton production. When primary production was high, baseline respiration was insignificant and algal activity usually dominated the CO2 exchange across air-water interface. Therefore, we consider ACP in our study to be a highly autotrophic water body with high primary production, and the baseline respiration supported by external organic matter was insignificant, which corresponded to the measured CO2 flux that ranged from −71.14 to −3.96 mg CO2 m−2 h−1. Moreover, the insignificant relationship between CO2 flux and air temperature or water temperature (Table 2) confirmed the conclusion that ACP was an autotrophic water body where algal photosynthesis, rather than the mineralization of organic matter, played a more important role in CO2 flux, because an increase in air temperature favored the CO2 production derived from the mineralization of organic matter (Huttunen et al. 2003). Xing et al. (2005) observed that, in a subtropical lake (Donghu, China), exponential relationships between CH4 emission and air, water surface, and sediment surface temperature were observed in a subtropical lake (Donghu, China) (Xing et al. 2005). However, in our study, CH4 emission from ACP was not affected by seasonal variability of temperature. CH4 emission in air-water interface resulted from the balance of two opposing processes: methanogenesis in anoxic conditions and the oxidation of the generated CH4. The production of CH4 is dependent on the concentration of PO4 3− (Schrier-Uijl et al. 2011), phytoplankton primary production (Xing et al. 2005), electron donors and acceptors (Van Bodegom and Scholten 2001). In addition, the CH4 could be oxidized to CO2 during any stage of its travel from the sediment through the water column to the atmosphere (Whiting and Chanton 2001). A large fraction of the unoxidized CH4 was likely to be emitted to the atmosphere by diffusion.

Spatial Variation of CO2 and CH4 Fluxes

Vegetation type and species composition affected the carbon dynamics and the formation and emission of the GHG in wetlands (Van Der Nat and Middelburg 2000; Ström et al. 2005). The significant higher fluxes of CO2 and CH4 from SCM than other three sites indicated that the invasion by exotic S. alterniflora had resulted in increase in CO2 and CH4 fluxes sharply, which was consistent with the result of Tong et al. (2012), who found that CH4 cycling in S. alterniflora marshes was high CH4 production in Min River estuary in southern China. The same conclusion was obtained in the Yangtze River estuary (Cheng et al. 2007) and in Jiangsu province (Zhang et al. 2010).

Plants have three main functions in the regulation of CO2 and CH4 emissions. Firstly, plants act as an important source of methanogenic substrate through excreting labile carbohydrates as exudates and root debris. Minoda et al. (1996) and Watanabe et al. (1999) pointed out that substrates derived from plants contributed up to 90 % of the total CH4 emission. Roots can directly regulate most aspects of rhizosphere C flow either by regulating the exudation process itself or by directly regulating the recapture of exudate from soil (Jones et al. 2004). Zhang et al. (2010) found that increase of CH4 emissions was mainly due to a rise in porewater CH4 concentrations in the S. alterniflora mesocosm, therefore concluded that S. alterniflora could fix and allocate more organic carbon, such as root exudates and debris inputs to the soil. Therefore, when compared with S. salsa, the presence of S. alterniflora ensured enhanced CH4 production and emission in wetlands. In our study, the WSS, OF and SCM were predominated by T. chinensis, S. salsa, and S. alterniflora, respectively. Rhizospheres in these sites were different due to different vegetations and that probably affected the emission of CH4 and CO2. It is likely that the slightest change in the chemistry of the soil or physiology of the plant induced rapid shifts in the quantity and quality of the exudative flux (Jones et al. 2004).

Secondly, numerous reports demonstrated that CH4 emission was well correlated with aboveground living biomass of vegetation (Hirota et al. 2007; Tong et al. 2012). In this study, the maximum aboveground biomass of T. chinensis, S. salsa, and S. alterniflora were 200.1131 ± 15.8172, 376.8582 ± 31.5023, and 1281.9200 ± 176.9304 g m−2, respectively. The high aboveground biomass of S. alterniflora matched with the high emission of CO2 and CH4 from the SCM, therefore supporting the conclusion that aboveground live plant biomass was a key factor controlling carbon production and emission. Tong et al. (2012) also found that S. alterniflora could fix and then allocate more carbon to the soil, which in turn resulted in higher CH4 production and emission when compared with native species.

Thirdly, plants act as a conduit for CO2 and CH4 transport through the aerenchyma system, and as a source of oxygen stimulating CH4 oxidation. Using 14C labeling techniques, Christensen et al. (2003) observed that the emission of CH4 was dependent on the amount of vascular plants. Other studies also found that 39.7–90.0 and 48.8–90.0 % of CH4 emission were transported by S. alterniflora and Phragmites australis (Cheng et al. 2007). S. salsa adapts to tidal inundation because the transportation mechanism carries oxygen from aboveground parts to the roots via the aerenchyma (Han et al. 2005). But we found that CO2 and CH4 emissions from OF (dominated by S. salsa) were lower than those from SCM (dominated by S. alterniflora). That may be due to the difference of vegetations in the two sites. S. salsa is a succulent halophytic herb (Song et al. 2009) while S. alterniflora, which belongs to the perennial grass family, has visibly evident lacunae in its stems (Tong et al. 2012), which may explain why the CH4 transport potentials of S. alterniflora was higher than that of S. salsa. However, S. alterniflora has a thick stem and well-developed aerenchyma tissue, which can deliver more oxygen into the rhizosphere and lead to higher rates of CH4 oxidation under S.alterniflora stands. But CH4 emission from S. alterniflora was still higher than others because this effect was outweighed by the higher CH4 production.

Soil moisture in SCM was greater than that in WSS and OF. This was a result of the construction of seawall, which blocked the transport of sea water into the study area and the low coverage of vegetation in WSS and OF under the strong evaporation (evaporation/precipitation ratio, 3.52) condition. Significantly positive impact of moisture on CH4 emission was observed in WSS, OF, and SCM, but the relationship between CO2 fluxes and moisture was not significant (Table 2). In our study, soil moisture in SCM was greater than these in WSS and OF, and was inundated occasionally by tide on the neap tide day. The consumption of CH4 from WSS and OF and the emission from SCM in our study consisted with the conclusion that soil moisture controlled CH4 emission from sites where the water table fluctuates below the soil surface (Christensen 1993). On other hand, low soil moisture make O2 diffuse into soil to oxygen CH4 to CO2.

Comparisons with Other Studies

At present, reports which focused on GHG emission in Yellow River estuary are scarce. Sun et al. (2013) and Zhang et al. (2014) studied GHG emission in natural wetlands and found that fluxes of CH4 and CO2 were −0.0128 mg CH4 m−2 h−1 (Sun et al. 2013) and 20.86 to 45.31 mg CO2 m−2 h−1 (Zhang et al. 2014), respectively (Table 3). Compared with CO2 and CH4 fluxes from natural wetlands where is affected minimally by human mentioned in these two studies, we found that CO2 fluxes recorded from WSS and OF in our study were greater, while CH4 fluxes were close to the reported values. That might due to the construction of seawall leading to low moisture which accelerated CO2 emission. CO2 and CH4 emissions from S. alterniflora marsh (SCM) in our study were also greater compared with that from natural wetland, differences in vegetation may be the main reason. CH4 flux from S. alterniflora marshes in Yellow River estuary was smaller than these from S. alterniflora marshes in Min River estuary (Tong et al. 2012), Yangtze River estuary (Cheng et al. 2010), and Jiangsu province, China (Zhang et al. 2010), which might due to the different latitudes that study areas located. The emission of CH4 and consumption of CO2 in aquaculture pond, similar to the result of Yang et al. (2012), which different from natural water body (lakes, reservoirs) act as a source of CO2 and CH4 (Tremblay et al. 2005; Silvennoinen et al. 2008; Schrier-Uijl et al. 2011; Diem et al. 2012). Aquaculture ponds are significantly influenced by human activities, therefore nutrient substance content and physicochemical property of these will be different from natural water bodies (lakes, reservoirs), which can lead to differences in GHG emissions.

Table 3 CO2 and CH4 flux rates from this study and literature
Full size table

Conclusion

Human activities have profound impact on greenhouse gas emission in the Yellow River estuary. Exotic-specie S. alterniflora invasion significantly increased CO2 and CH4 emissions due to its strong gas transportation capacity and excreting large amounts of substrates for methanogens. The construction of seawall blocked sea water transporting into the study area leading to low soil moisture which accelerated CO2 emission. Aquaculture pond was a source of CH4 and a sink of CO2. However, the results of this study are preliminary and need to be validated with further studies. More investigations and long-term measurements (including year-to-year variations) on CO2 and CH4 exchanges between ecosystem and atmosphere are needed in order to gain a better understanding of human activities on CO2 and CH4 emissions in the Yellow River estuary.