Estuaries and Coasts DOI 10.1007/s12237-015-9997-4
Carbon Dioxide Fluxes and Their Environmental Control in a Reclaimed Coastal Wetland in the Yangtze Estuary Qicheng Zhong 1 & Kaiyun Wang 2 & Qifang Lai 1 & Chao Zhang 3 & Liang Zheng 1 & Jiangtao Wang 2
Received: 5 June 2014 / Revised: 13 May 2015 / Accepted: 27 May 2015 # Coastal and Estuarine Research Federation 2015
Abstract Large areas of natural coastal wetlands have suffered severely from human-driven damages or conversions (e.g., land reclamations), but coastal carbon flux responses in reclaimed wetlands are largely unknown. The lack of knowledge of the environmental control mechanisms of carbon fluxes also limits the carbon budget management of reclaimed wetlands. The net ecosystem exchange (NEE) in a coastal wetland at Dongtan of Chongming Island in the Yangtze estuary was monitored throughout 2012 using the eddy covariance technique more than 14 years after this wetland was reclaimed using dykes to stop tidal flooding. The driving biophysical variables of NEE were also examined. The results showed that NEE displayed marked diurnal and seasonal variations. The monthly mean NEE showed that this ecosystem functioned as a CO2 sink during 9 months of the year, with a maximum value in September (−101.2 g C m−2) and a minimum value in November (−8.2 g C m−2). The annual CO2
1
Research Center for Saline Fisheries Technology, East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, No. 300 Jungong Road, Shanghai 200090, China
2
Shanghai Key Laboratory for Urban Ecological Processes and Eco-Restoration, East China Normal University, No. 500 Dongchuan Road, Shanghai 200241, China
balance of the reclaimed coastal wetland was −558.4 g C m−2 year−1. The ratio of ecosystem respiration (ER) to gross primary production (GPP) was 0.57, which suggests that 57 % of the organic carbon assimilated by wetland plants was consumed by plant respiration and soil heterotrophic respiration. Stepwise multiple linear regressions suggested that temperature and photosynthetically active radiation (PAR) were the two dominant micrometeorological variables driving seasonal variations in NEE, while soil moisture (Ms) and soil salinity (PSs) played minor roles. For the entire year, PAR and daytime NEE were significantly correlated, as well as temperature and nighttime NEE. These nonlinear relationships varied seasonally: the maximum ecosystem photosynthetic rate (A max), apparent quantum yield (∂), and Q10 reached their peak values during summer (17.09 μmol CO2 m−2 s−1), autumn (0.13 μmol CO2 μmol−1 photon), and spring (2.16), respectively. Exceptionally high Ms or PSs values indirectly restricted ecosystem CO2 fixation capacity by reducing the PAR sensitivity of the NEE. The leaf area index (LAI) and live aboveground biomass (AGBL) were significantly correlated with NEE during the growing season. Although the annual net CO2 fixation rate of the coastal reclaimed wetland was distinctly lower than the unreclaimed coastal wetland in the same region, it was quite high relative to many inland freshwater wetlands and estuarine/coastal wetlands located at latitudes higher than this site. Thus, it is concluded that although the net CO2 fixation capacity of the coastal wetland was reduced by land reclamation, it can still perform as an important CO2 sink.
3
Key Laboratory of Geographic Information Science, Ministry of Education, East China Normal University, No. 500 Dongchuan Road, Shanghai 200241, China
Keywords Yangtze estuary . Coastal wetland . Land reclamations . CO2 sink capacity
Communicated by Rui Santos Qicheng Zhong and Chao Zhang contributed equally to this work. * Qicheng Zhong
[email protected] * Chao Zhang
[email protected]
Estuaries and Coasts
Introduction Natural coastal wetlands generally exhibit high primary productivity and a low soil organic matter decomposition rate (Odum 1971; Bouillon et al. 2008). These ecosystems are also able to rapidly sequestrate carbon associated with tide-driven sediment deposition, and their CH4 emission rate is low (Chmura et al. 2003; Donato et al. 2011; Poffenbarger et al. 2011; Hopkinson et al. 2012). Therefore, natural coastal wetlands have always been considered as significant contributors to global Bblue carbon^ resources (Bridgham et al. 2006; Laffoley and Grimsditch 2009; Livesley and Andrusiak 2012). Unfortunately, in recent years, coastal ecosystems have been subjected to an increasing number of anthropogenic activities. Thus, plenty of natural coastal wetlands have suffered severe damages or have even been converted to other land uses (Connor et al. 2001; Gedan et al. 2009; Hopkinson et al. 2012). It was recently indicated that approximately 20∼50 % of the world’s coastal tidal wetlands have been lost as a result of their direct conversion into land for agriculture and aquaculture use (Kirwan and Megonigal 2013). As a result, the carbon budgets of coastal wetlands have probably been profoundly altered by these human disturbances (Hopkinson et al. 2012; Pendleton et al. 2012). In the context of global change, in view of restoring and preserving their high efficiency of carbon sequestration (Mcleod et al. 2011; Hopkinson et al. 2012), more attention should be paid to possible changes in carbon budgets resulting from damage to or the conversions of coastal wetlands. In recent decades, large areas of natural coastal wetlands around the world have been reclaimed by human-constructed dykes to satisfy the land demands driven by economic development and population growth in coastal zones (Connor et al. 2001; An et al. 2007; Laudicina et al. 2009; Kirwan and Megonigal 2013). A distinctive human-disturbed coastal wetland, i.e., a Bcoastal reclaimed wetland^ has a high possibility of appearing at reclaimed sites where subsequent utilizations are not implemented. Compared with unreclaimed coastal wetlands, the land use change stops tidal influences on the coastal reclaimed wetland; cuts off the exchange of water, heat, and matter between these areas and the offshore sea; and profoundly changes the hydrological conditions and nutrient status. Therefore, a coastal reclaimed wetland usually has a lower water table, an altered soil salinity level, and a weaker soil anaerobic environment (Connor et al. 2001; Gedan et al. 2009; Fernandez et al. 2010). Moreover, the secondary vegetation succession in coastal reclaimed wetlands can lead to changes in primary productivity (Roman et al. 1984; Sun et al. 2003; Zhong et al. 2014). All the environmental changes (especially the changes in hydrological factors) derived from land reclamations can potentially affect the carbon budgets of coastal wetlands (Laudicina et al. 2009; Pendleton et al. 2012). At the same time, such anthropogenic
disturbances to natural coastal wetlands may also potentially influence the responses of carbon budgets to ongoing climate change (Connor et al. 2001; Zhong et al. 2013). Despite the prevalence of land reclamation in natural coastal wetlands at a global scale, until now, in situ evaluation of the impacts of reclaiming wetlands on coastal carbon fluxes at the ecosystem scale is rare. Moreover, to provide a scientific basis to perform sustainable management and to make reasonable predictions of the carbon budgets of reclaimed coastal wetlands, it is necessary to understand the temporal dynamics and the driving biophysical variables of carbon fluxes in these ecosystems. Dongtan of Chongming Island is an internationally important wetland in the Yangtze estuary that covers an area of 326 km2. In addition to its rich biodiversity-supporting capacity, the solid CO2 sink function of this natural (unreclaimed) coastal wetland was recently confirmed (Guo 2010). However, Chongming Island was subjected to frequent land reclamations for a long period of time, and most of its land area was derived from reclaiming tidal flat and vegetated coastal wetlands (GSICI, 1996). In 1998, a critical portion of the natural coastal wetland in Dongtan was reclaimed using dykes by the local authorities for potential agricultural use. However, because the reclaimed site can still provide additional and excellent habitats for migratory and resident birds (Boulord et al. 2012), no agricultural management has been conducted at this site. Thus, this site provides a suitable location for exploring the effects of land reclamations on carbon fluxes in coastal wetlands, without any further agricultural management. In recent years, the eddy covariance technique has been employed to estimate CO2 exchanges (net ecosystem exchange—NEE) in both estuarine/coastal wetlands (Kathilankal et al. 2008; Guo et al. 2009; Zemmelink et al. 2009; Zhou et al. 2009; Han et al. 2013) and inland freshwater wetlands (Frolking et al. 1998; Lafleur et al. 2001; Syed et al. 2006; Aurela et al. 2007; Jimenez et al. 2012; Schedlbauer et al. 2010). Here, eddy covariance measurements of NEE were conducted in a reclaimed coastal wetland at Dongtan throughout the entire year of 2012. The overall effects of land reclamation on the NEE of the coastal wetland were evaluated by comparing our results from the reclaimed site with the results reported by Guo (2010) for the unreclaimed site. A series of environmental variables were also measured to reveal the major biophysical variables driving NEE in the reclaimed coastal wetland.
Materials and Methods Site Description The CO2 flux-monitoring tower was placed in a reclaimed coastal wetland at Dongtan of Chongming Island in the Yangtze estuary, China (31° 38 N, 121° 58 E) (Fig. 1). The area has a northern subtropical marine climate with a mean annual air temperature of 15.3 °C and a mean annual precipitation of
Estuaries and Coasts Fig. 1 The geographical location of the CO2 flux monitoring tower in the reclaimed coastal wetland
1004 mm (GSICI, 1996). The study site was part of a welldeveloped natural coastal wetland until it was reclaimed for potential agricultural use in 1998 using dykes. During the first few years after the reclamation, the soil conditions were not favorable for crop cultivation. Because this site is adjacent to the Chongming Dongtan Birds National Nature Reserve and the mixed herbaceous community dominated by Phragmites australis can provide a good habitat for a variety of migratory and resident birds, the local authorities have not yet implemented any agricultural practices at this site. Although this site is no longer affected by regular tidal flooding, surface ponding can occur when precipitation is abundant. The mean altitude of the site is 3.8 m, and the annual mean water level was 30 cm below the ground surface in 2011. The tower was 6.0 m in height and was constructed from galvanized steel. A Phragmites australis community covered 80 % of the land area surrounding the tower within a 300-m radius. This species generally germinates in mid-March and withers in midOctober; the maximum height of the vegetation canopy can reach more than 2 m during the peak growing season. The remaining surrounding land area was covered by water bodies (15 %) and artificial vegetation (5 %). CO2 Flux Monitoring From April 2011, the CO2 fluxes between the atmosphere and the reclaimed coastal wetland were continuously monitored
using the eddy covariance technique. The open-path eddy covariance system was installed on the CO2 flux-monitoring tower. The system consisted of a three-dimensional sonic anemometer (GILL-windmaster pro, GILL corporation, London, England); an open-path infrared gas analyzer (Li7500a, Li-cor, Lincoln, NE); and a high-frequency data logger (Li-7550, Li-cor, Lincoln, NE). The sonic anemometer and infrared gas analyzer were placed at the top of the tower above the vegetation canopy. The vertical distance between the two sensors and the ground surface was 6.5 m, while the horizontal distance between the two sensors was 20 cm. A data logger was placed in the middle of the tower (avoiding direct sun exposure), and the raw flux data were collected and stored on a 4-GB disk with a 10-Hz recording frequency in the data logger. Both the eddy covariance system and the micrometeorological monitoring system were powered by lead-acid batteries (12 V, 1200 Ah). The battery was charged by a solar panel (60 W) and could also be charged using AC in case of weak solar radiation. The chemicals tubes of the open-path infrared gas analyzer were replaced every year. The zero point of the CO2 concentration was calibrated using N2 gas, which had been purified by desiccant and soda lime. The span of the CO2 concentration was calibrated using standard CO2 gas (±1 % 450 ppm), and the span of the H2O concentration was calibrated using a dewpoint meter (LI-610, Li-cor Inc., Lincoln, NE). The half-hourly horizontal wind components of the sonic anemometer (i.e., u and v directions) were regularly
Estuaries and Coasts
compared to the corresponding variables measured using a wind speed meter and a wind direction meter. If the error tolerance exceeded 5 %, the sonic anemometer was calibrated. Micrometeorological Monitoring The micrometeorological system was used to continuously monitor a series of micrometeorological variables, including mainly air temperature (Ta, °C) and relative air humidity (RH, %) (HMP45C, Vaisala,Helsinki), photosynthetically active radiation (PAR, μmol m−2 s−1) (LI-190SA, Li-cor, Lincoln, NE), total solar radiation (TSR, W m−2) (LI-200SA, Li-cor, Lincoln, NE), wind speed (WS, ms−1), and wind direction (WD, °) (05103–10, RM Young, Traverse City, MI), precipitation (PRE, mm) (TE525 M, Texas Electronics, Dallas, TX) at a height of 4 m, and soil temperature (Ts , °C), moisture (M s, %), salinity (PS s , g NaCl L−1) at depths of 10, 20, and 30 cm (Hydra Probe II, Stevens Water Monitoring Systems Inc., Portland, OR). A data logger (DT-80, Datataker, Thermo Fisher Scientific, Scoresby, VIC, Australia) was used to collect micrometeorological data at a 1-min frequency; these data were converted into half-hourly data using the built-in automatic data program and were stored in the memory of the data logger. The daily mean evapotranspiration (E) of the vegetation was calculated using the PenmanMonteith equation according to Allen et al. (1998). Calculation, Calibration, and Quality Control of CO2 Flux
where ω' and ρc ' are the variances of the vertical speed wind and atmospheric CO2 concentration, respectively, and the over bar represents the average value. The half-hourly CO2 flux data outputted by EddyPro were further filtered according to a series of standards before they were used for later analysis. The excluded data mainly included the following: (1) The CO2 flux data when precipitation or fog occurred; (2) The CO2 flux data whose absolute value (|FCO2|) exceeded 30 μmol CO2 m−2 s−1 (does not meet biology regularity); (3) The CO2 flux data when the air turbulence was weak, especially when friction velocity (Ustar)
