4.1. Gas Transport Mechanism in S. Mariqueter
 Researchers had demonstrated that mechanisms of gas transport commonly employed by wetland plants include molecular diffusion, characterized by migration of gas molecules along a concentration gradient, and convective flow in response to a pressure differential [Armstrong et al., 1992; Brix et al., 1992]. Previous studies have demonstrated significant differences in plant-mediated CH4 fluxes under dark versus light conditions [Whiting and Chanton, 1996; Van der Nat and Middelburg, 2000], when wetland plants were the primary conduits facilating CH4emission via convective throughflow and stomata-controlled transport [Van der Nat and Middelburg, 1998]. Significantly attenuated CH4 emission rates have been observed when aboveground portions of the vegetation were clipped under light conditions, because the capacity for pressurized transport was eliminated [Van der Nat and Middelburg, 2000]. However, there can be significant increase in CH4 flux under dark conditions when the stem or the leaf is cut, if diffusion is the primary transport mechanism in the wetland plant, and especially diffusion had been limited by the resistance of the aboveground portion of the plant [Schimel, 1995].
 Although many studies focused on CH4 transport by wetland plants, the effects of plant on N2O emission have not been extensively studied, both of these gases, are soluble in water, so they can be transported concurrently within wetland plants by convective throughflow or diffusion. Some studies have reported that wetland plants can affect N2O fluxes by acting as a transport conduit [Reddy et al., 1989; Mosier et al., 1990], influencing nitrification-denitrification processes in the rhizosphere [Reddy et al.,1989; Bodelier et al., 1996; Kettunen et al., 2005], and/or producing N2O by photosassimilation of NO2− in the leaves [Smart and Bloom, 2001].
 Same as many other wetland plants, S. mariqueter possesses abundant intercellular gas space in the stem and leaf tissues (see Figures 2b–2d), indicating that convective throughflow is likely, particularly when an intensive pressure differential exists. In the previous research on CH4 emission, when the transpiration was higher, molecular diffusion and convective gas flow were the two main mechanisms of CH4 transport in S. mariqueter plants [Wang et al., 2009]. In July, when S. mariqueter was exuberant and PAR and temperature were relatively high, higher gas transport efficiency via convective throughflow was expected. Cutting the aboveground portion of S. mariqueter significantly decreased the CH4 fluxes in light chambers [Wang et al., 2009], because the capacity for pressurized transport has been eliminated [Van der Nat and Middelburg, 2000]. But in light chambers, clipping the aboveground portion of S. mariqueter significantly enhanced N2O flux compared with the unclipped chambers (p < 0.001) (see Figure 3), there was no observable of the transportation of N2O by convective throughflow. On the other hand, N2O fluxes in the unclipped dark were also significantly higher than those in light chambers (p < 0.001), suggesting that S. mariqueter photosynthetic activity significantly decreased the N2O emission.
 A microscopic anatomical evaluation of S. mariqueter revealed that cortical gas spaces occupied the full cortex (see Figure 2a), suggesting low resistance on transporting gas from root to stem and leaf. Such a configuration generally has little effect on gas diffusion [Schuette et al., 1994; Sorrell et al., 1997], although the potential diffusion resistance at the transition between the rhizosphere and the root aerenchyma was unknown. Based on the previous research in which molecular diffusion became the primary transport mechanism when S. mariqueter began to senesce, clipping the aboveground portion of S. mariqueter enhanced CH4 emission indicating the resistance of stems and leaves [Wang et al., 2009], suggesting that aboveground portion of vegetation was a factor regulating CH4 diffusion [Schimel, 1995]. While cutting of the aboveground portion of S. mariqueter significantly decreased the N2O flux in dark chamber (p < 0.001) especially in July and August, It was clear that gas transport by S. mariqueter was not the primary factor governing N2O emission in wetland. Therefore, the S. mariqueter transport function was to a degree influenced by its other physiological activities, such as production of root exudates and oxygenation of the rhizosphere.
4.2. Effect of S. mariqueter on Nitrous Oxide Fluxes
 Compared to CH4, there is not a large amount of research on plant-dependent N2O flux from wetland ecosystems, including information on the relationship between N2O flux and plant productivity. In the study of nitrogen loading in a freshwater marsh from Sanjiang plain, North China, Zhang et al.  proposed that Deyeuxia angustifolia had the potential to increase the production of N2O by supplying easily decomposable root exudates, which enhance microbial activity in soil. Based on a significant correlation between N2O flux and aboveground biomass, they inferred that a large fraction of N2O flux was facilitated by Deyeuxia angustifolia via the transpiration stream [Zhang et al., 2007]. However, in other studies, there was no correlation between plant characteristics and N2O flux [Chen et al., 1997; Müller, 2003]. Yan et al.  further concluded that N2O is mainly transported by diffusion through the soil surface rather than through plants [Yan et al., 2000].
 In our study, the effect of S. mariqueter on N2O flux was relatively distinct and straightforward. By comparing N2O flux in light and dark chambers, we conclude that the photosynthetic activity of S. mariqueter attenuated N2O flux significantly during the entire study period (see Table 1 and Figure 3). With the exception of June, monthly averaged N2O absorption rate varied with the growth of S. mariqueter. Significant positive correlations were found between N2O emission in dark chambers and temperature; conversely, there was a significant negative correlation between N2O flux and PAR under light conditions (p < 0.01). Furthermore, clipping the aboveground part of S. mariqueter in light chambers greatly enhanced N2O emission during each monthly measurements. On the other hand, there was no significant difference in N2O flux between clipped and unclipped dark chambers. We interpret these results as follows: PAR governed the photosynthetic rate and growth of S. mariqueter; consequently, a significant inhibitory effect on N2O emission was imposed by the S. mariqueter community. Scatterplots of diurnal N2O flux in light unclipped chambers versus NEE and NPP show the significant negative correlation between N2O flux and NEE and NPP (see Figure 5), the regression curve of N2O fluxes and NEE, NPP was described by the linear function: N2O flux (μg N2O m−2 h−1) = −0.2790 × NEE (mg CO2 m−2 h−1) − 2.0345 (R2 = 0.7758, p < 0.0001), N2O flux (μg N2O m−2 h−1) = −0.1224 × NPP (mg CO2 m−2 h−1) + 24.546 (R2 = 0.4211, p < 0.001).
Figure 5. Diurnal gas fluxes in light chamber in unclipped treatments, plotted against net ecosystem exchange (NEE) and net primary production (NPP). Bars represent standard deviation of the triple duplicates.
Download figure to PowerPoint
 N2O production in freshwater marshes was clearly limited by nitrogen deficiency [Zhang et al., 2007]. Under field conditions, concentrations of inorganic nitrogen compounds are usually low in the root zone in the growing season due to plant uptake of nitrogen [Bodelier et al., 1996; Van der Nat et al., 1997]. The growth of S. mariqueter might suppress N2O emission by taking up nitrogen in sediment, which could directly inhibit the N2O production by diminishing the substrate for nitrification and denitrification. When denitrifiers and nitrifiers in the sediment were suffering the NO3− limitation, they would consume the N2O diffusing from the atmosphere [Frasier et al., 2010]. Although this process is poorly understood, some researches found that N2O was consumed by the soil [Chapuis-Lardy et al., 2007; Peichle et al., 2010; Ullah and Moore, 2011].On the other hand, Zhang et al.  assumed that plant uptake can reduce some N2O, most likely by taking up available nitrogen. Furthermore, our explanation is also in agreement observations on Spartina alterniflora and Phragmites australis reported by Cheng et al.  in experimental mesocosms using plants and soils from the Jiuduansha salt marsh in the Yangtze River estuary.
 Beside the competition between S. mariqueter and microbes for nitrogen, O2 is transported and diffused into S. mariqueterrhizosphere during photosynthesis. It effectively increases the aerobic-anaerobic surface area and influences anaerobic metabolism in the wetland soil [Megonigal et al., 2004]. In another research report [Wang et al., 2009], CH4 flux in light chamber was higher than in dark chamber from July to September, but there were no significant difference in July and August, and in October, CH4 flux in light chamber was slightly lower than in dark chamber. Photosynthesis of the plant or the O2 transported into sediment did not inhibit methanogenesis process greatly because more organic materials would also be transported into root and exuded [Zhang et al., 2007]. Denitrification is often tightly coupled to nitrification in the high redox area [Kettunen et al., 2005] that predominates in the rhizosphere, while methanogenesis dominates at lower reducer environment only when other electron acceptors are almost exhausted. Basing on this spatial distribution, it is reasonable to assume that the denitrification is more sensitive to the rhizospheric O2 delivery.
 In sediment, because of nitrate limitation coupled nitrification-denitrification was the main process of N2O production [Moseman-Valtierra et al., 2011]. LaMontagne et al.  found that there was a higher N2O uptake rates in the opaque chamber deployed on macroalgae covered sediments, and benthic N2O sink can be explained by a close coupling of nitrification and denitrification [LaMontagne et al., 2003]. When coupled nitrification provides the nitrate for denitrifier, N2O uptake can occur [van Raaphorst et al., 1992], denitrifiers would consume N2O during hypoxic conditions [Usui et al., 2001]. When S. mariqueter photosynthesized under light conditions, more O2 was released from the root to sediment meeting the demand of nitrification, stimulated denitrifier consumed the N2O inducing to it diffusing from atmosphere into sediment. While in the dark chambers, although denitrification process would be slowed down because of the nitrate limitation, the consumption was decreased quickly and there was a net N2O production from denitrification. On the other hand, the stimulation of CH4 production by root exudates of S. mariqueter [Wang et al., 2009] indicates that the root exudates should also fuel the denitrifiers resulting in greater denitrification rate [Zhang et al., 2007] and more N2O consumption.
 From July to October, under light conditions, besides the diffusing into marsh sediment directly, N2O could diffuse with O2 from the stomata to the rhizosphere when S. mariqueter photosynthesized and the stomata were open, where it was consumed by coupled nitrify denitrification [Vieten et al., 2008; Moseman-Valtierra et al., 2011]. Under dark conditions, molecular diffusion was not the main transporting mechanism of N2O emission. Clipping the aboveground portion of S. mariqueter had no obviously effect N2O flux, facilitating N2O diffusing from rhizosphere to atmosphere. In July, the flux in clipped light and dark chambers was significantly lower than in unclipped dark chambers (p < 0.001, Figure 3), indicating that the positive effect of respiration of S. mariqueter on N2O emission exceeding the negative effect by the resistance of leaf and stem on N2O diffusion. But there was no significant difference between flux in unclipped dark chambers and clipped light and dark chambers from August to October (Figure 3). N2O diffusing out or into by molecular diffusion method through intercellular gas space was depended on its gradient between rhizosphere and atmosphere. The transport of N2O by S. mariqueter is not only simple molecular diffusion but the result associated with the complex effects of nitrogen cycling in rhizosphere. More detailed studies on N2O transport and its production and consumption in the rhizosphere via nitrification and denitrification process are needed in order to fully understand such effects.