Wetlands DOI 10.1007/s13157-014-0522-5
SHORT COMMUNICATION
Carbon Dioxide and Methane Emissions from a Wet-Dry Tropical Floodplain in Northern Australia A. M. Bass & D. O’Grady & M. Leblanc & S. Tweed & P. N. Nelson & M. I. Bird
Received: 8 November 2013 / Accepted: 7 February 2014 # Society of Wetland Scientists 2014
Abstract In order to better understand how global climate change will affect greenhouse gas dynamics in wetland systems, an accurate quantification of global greenhouse gas emissions from these areas is essential. Despite a large proportion of wetlands occurring in tropical areas, data on greenhouse gas fluxes from these areas is limited. This study aimed to quantify carbon dioxide (CO2) and methane (CH4) fluxes from an undisturbed tropical wetland environment in northern Australia, and to evaluate the influence of different habitat types on net emission rates. Fluxes were measured at seven sites, representing different habitat types during the inundation season of 2012. Highest CO2 fluxes, with a maximum at 199.4 mg CO2-C m−2 h−1, were measured in open water areas. This likely corresponded to increased mobilisation of sediment organic matter and high water turbulence, as inferred from turbidity measurement. CH4 fluxes, however, were greatest in densely vegetated areas and peaked at 153.2 mg CH4-C m−2 h−1, a possible result of increased transport through plant stems. Carbon dioxide and methane fluxes over the whole wetland averaged 86.0 mg CO2-C m−2 h−1 and 25.3 mg CH4-C m−2 h−1, respectively.
Keywords Carbon dioxide . Methane . Floodplain . Wetland . Tropical . Australia
A. M. Bass (*) : P. N. Nelson : M. I. Bird Centre for Tropical Environmental and Sustainability Science and School of Earth and Environmental Sciences, James Cook University, Cairns, Australia e-mail:
[email protected] D. O’Grady : M. Leblanc : S. Tweed Centre for Tropical Water and Aquatic Ecosystem Research, School of Earth and Environmental Sciences, James Cook University, Cairns, Australia
Introduction Wetlands are a relatively small surface feature, occupying 4– 6 % of the terrestrial land area (Mitsch and Gosselink 2000; Kuehn et al. 2004), but have a disproportionately important role in both carbon fluxes and storage (e.g. Zhou et al. 2009; Miller 2011). As the build-up of carbon dioxide (CO2), methane (CH4) and other greenhouse gases (GHGs) in the atmosphere continues to increase the global average temperature, understanding and quantifying the fluxes to and from various environments has become paramount. The majority of wetland carbon research to date has been conducted in boreal and temperate regions in the northern hemisphere and these studies indicated that wetlands generally act as a sink of atmospheric CO2 and a source of atmospheric CH4 (e.g. Lund et al. 2010; Petrescu et al. 2010; Livesley and Andrusiak 2012). However, there is a small, but increasing body of evidence analysing trace gas fluxes from tropical wetland systems, where rates of autotrophic and heterotrophic processing are expected to be elevated due to higher temperatures (Kemenes et al. 2011). These studies suggest that such wetlands are sites of significant CO2 and CH4 emissions (e.g. Richey et al. 2002; Melack et al. 2004; Belger et al. 2011). The paucity of data means that our level of understanding of the biogeochemical functioning of tropical wetlands is still low and global estimates of wetland GHG fluxes are confounded by the majority of tropical / subtropical studies being undertaken in the Amazon basin, Florida or in rice paddies (e.g. Banker et al. 1995; Husin et al. 1995; Adhya et al. 2000). CO2 and CH4 are the dominant GHG’s emitted from tropical wetlands. Emission rates are governed by numerous factors including water depth, temperature, dissolved gas concentration and vegetative cover (e.g. MacIntyre et al. 1995; Whalen 2005). Under aerobic conditions, CO2 is produced by the breakdown of organic material via heterotrophic processes
Wetlands
and concurrently consumed via autotrophic pathways, with the balance between the two being a primary dictator of fluxes. Aerobic conditions in the soil lead to the consumption of methane via methanotrophic bacteria. Both CO2 and CH4 are also produced under anaerobic conditions via anaerobic respiration and methanogenesis, respectively. CO2 emissions are dominated by diffusive fluxes while both ebullitive (bubbling) and diffusive fluxes are significant for methane. Ebullitive flux limits the amount of time methane is present in the aerobic water layers, reducing the chance for oxidation. Thus, ebullitive fluxes can be quantitatively significant (e.g. Devol et al. 1988; Engle and Melack 2000). Emergent vegetation can further enhance CH4 emissions, allowing a pathway to bypass the aerated water column. Indeed, estimates suggest methane flux through tree stems may account for 20 % of the CH4 flux from terrestrial ecosystems (Gauci et al. 2010). Determining the dominant controls on GHG fluxes in tropical wetlands is increasingly important when considering future climate change; particularly with regard to positive feedback loops. Current projections suggest the rates of GHG emissions from wetlands will increase as the global average temperature continues to rise, and this is of particular significance in temperate and tropical systems (e.g. Cao et al. 1998; Shindell et al. 2004). More recently evidence suggests that both temperature and the magnitude of wet-dry cycles may have a significant effect of GHG emissions (Nahlik and Mitsch 2011), both of which are predicted to rise in some areas with climate change. In this study, we measured CO2 and CH4 emissions from a range of different settings across a wet-dry tropical floodplain of northern Australia during an inundation season. Our objective was to determine how physio-chemical factors such as water temperature, depth and habitat type modulated emission rates. Additionally we examine how the emission rates change over the course of a flood season in our wetland site.
Methods Study Site The Mary River catchment, in tropical north Australia, drains an area of approximately 8090 km 2 of which roughly 1080 km2 is seasonally inundated (Fig. 1). Inundation occurs during a pronounced wet season (approximately November to July) with the majority of the 915 mm annual average rainfall taking place between November and April. Shallow silt and clay-dominated soils dominate the floodplain and are underlain by gleyed marine sediments (Wassen 1992). Aquatic grasslands dominate the wetland, with forested areas also significant. Within a wetland, vegetation communities tend to be related to water depth (Finlayson 2005; Petit et al. 2011).
Seven sites encompassing a range of habitat types were selected for sampling. A transect extending from the river channel to the floodplain edge (Fig. 1) included an area of open flood water (site 1), two areas of thick aquatic grass cover (Urochloa mutica) (sites 2, 4), an area of deep water vegetated by large lillies (Nelumbo nucifera) (site 3), and an area of partially open water with submerged macrophytes (site 5). This transect covered a distance of ca. 3 km. Two additional sites were located in a submerged forest area (dominated by Melaleuca leucadeudia) (W) and in a permanently inundated billabong (BB) that due to its relatively large depth (>3 m) retains water year round. Data Collection A total of eight sampling events were conducted throughout the inundation period, on the 2nd/6th/9th and 27th of February, 2nd and 30th of March, and 5th and 8th of May 2012. Sites were accessed using an airboat. Once stationary, the site was measured on either side of the vessel where no vegetative disturbance had occurred. Due to vegetation damage from previous sampling trips, each sample location was very slightly different to the last to avoid measuring previously disturbed areas. Water Quality Each site was instrumented with a Schlumberger Mini-Diver depth/temperature probe, securely tied to a metal picket, encased in a porous PVC tube casing at the bottom of the water column. Depth and temperature were logged at 1-h intervals with an accuracy of ±1 cm and ±0.1 °C respectively. Water temperature, pH, specific conductance and dissolved oxygen concentration at the time of gas sampling were measured using an In-Situ Troll 9500 multi-parameter probe accurate to ±0.1 °C, ±0.1 pH units, ±1 % saturation and ±2 μS cm−1, respectively. Dissolved organic carbon (DOC) and turbidity were measured using a S::can Spectrolyser, a portable UV-Vis spectrometer capable of recording DOC and turbidity in-situ (Bass et al. 2011). The Spectrolyser unit was calibrated against known concentration solutions of DOC and turbidity with an accuracy of ±0.20 mg L−1 and ±0.10 NTU respectively. Measurements were taken throughout the water column at 20 cm intervals in triplicate, though unless otherwise stated only surficial values are used when considering the results reported in this study. Chlorophyll a in the top 10 cm was measured using an AquaFluor™ handheld fluorometer. The in-vivo fluorescence data was correlated with extracted chlorophyll a data from samples measured spectrophotometrically at the Cairns Regional Council laboratory. Due to unavailability in the early stages, only the last four sampling dates have chlorophyll a measurements.
Wetlands
Fig. 1 Location of the Mary River floodplain area in northern Australia showing the spatial distribution of different inundated land cover types measured in this study. The magnified area shows the locations of the seven study sites
Gas Flux Measurements
Fig. 2 Time series of daily hydro-meteorological data. Deepest and shallowest depth data (from sites 3 and 1 respectively) shows depth range of the measured sites. Black circles indicate dates where gas flux measurements were made
CO2 and CH4 fluxes were measured using a floating chamber methodology based on that described in Rosenqvist et al. (2002). Chambers were 15 cm in diameter with a headspace volume of 2 L. Gas concentrations were recorded at approximately 40 s intervals using an INNOVA Photoacoustic gas analyser with each measurement taking approximately 10– 15 min depending on the flux rate. Duplicate measurements were made at each of the seven sample sites per sampling date. Ebullitive flux was calculated when a sudden increase in gas concentration was observed. While questions have been raised over the accuracy of chamber made gas flux measurements in lentic systems (Vachon et al. 2010), the methodology used in this study is widely applied and as such the data comparable to the existing literature.
Wetlands Table 1 Summary statistics for the Mary River floodplain for sample dates between 02/01/2012 and 08/05/2012, including sites specific values (value ±1 standard deviation). DOC (dissolved organic carbon), NTU (Nephelometric turbidity unit) Min Max Depth (cm) Temperature (°C) pH Turbidity (NTU)
Ave
SD
26.1 276.2 162.3 64.3 25.5 34.4 30.0 2.4 4.9 8.7 5.9 0.8 6.5 35.5 15.3 5.8
DOC (mg L−1) 7.0 16.7 CO2 flux −9.7 603.8 (mg CO2-C m−2 h−1) −3.3 153.2 CH4 flux (mg CH4-C m−2 h−1)
Site 1 98.4±55.2 28.6±1.3 5.7±0.2 11.9±1.5
Site 2
Site 3
Site 4
Site 5
W
151.8±52.2 198.7±53.2 148.7±55.9 92.1±46.0 28.9±2.2 28.8±1.8 30.6±2.4 30.8±2.8 31.3±3.0 5.3±0.3 5.4±0.3 5.6±0.2 5.9±0.2 5.8±0.3 18.8±8.0 13.4±4.3 16.7±5.3 17.5±6.0 15.8±5.2
11.4 2.3 13.7±1.7 86.0 95.2 199.4±195.7
9.6±1.5 9.7±1.4 75.3±38.4 166.4±43.4
10.1±1.7 10.8±1.6 11.3±2.4 27.9±11.6 65.7±18.7 52.5±52.4
25.3 33.3
68.9±48.1
31.9±20.4
33.0±27.3
39.4±27.0
3.0±2.8
1.8±2.7
BB 206.1±55.1 32.0±1.6 7.6±0.9 13.2±6.5 14.0±1.4 29.1±28.4 0.1±1.6
Statistical Analysis
Results
Comparison of average parameter values, encompassing all measurement dates, between sites was carried out using one way analysis of variance (ANOVA) with Tukeys range posthoc test used to deduce individual differences. Where requirements for parametric tests were not met Kruskal-Wallis ANOVA and Dunns method for post-hoc analysis was carried out. Differences were deemed significant if the 95 % confidence level was met.
Peak rainfall occurred over two main periods, from mid to late January, and from late February to mid-March in 2012. Two significant peaks in the floodwater depth were observed as a result (Fig. 2). Sampling dates corresponded to the peak of the initial, smaller flood, and the rising and receding limbs of the largest flood. Water depth varied from 26 to 276 cm depending on the site and date. The deepest site on the floodplain was site 3 which averaged approximately 2 m.
Fig. 3 a surface water temperature, b pH, c dissolved organic carbon (DOC) and (d) chlorophyll a concentration time series for each sampling site
Wetlands
Summary statistics for all sites is presented in Table 1. Water temperature over the flood season averaged 30.0± 2.4 °C (n=55) with the highest temperatures recorded in the permanently inundated billabong (BB) (Fig. 3a). Highest water temperatures corresponded to the periods of deepest inundation for all sample sites. Temporal variation in pH was generally less than ±0.3 (Fig. 3b), except in the billabong where it was both significantly higher (average 7.6 compared to 5.3–5.8 for other sites) and more variable (±0.9). Significant variation was observed in site average DOC concentration between sites (ANOVA, DF=6, p0.05). Significant variation in CH4 emissions (Fig. 4)
Fig. 4 Average CO2 and CH4 fluxes for each study site. Values represent the average of all data collected over the study period for each site and error bars represent ±1 SD (n=16). Letters represent significant differences between groups calculated by Tukeys post-hoc analysis
between sites was observed (Kruskal-Wallis, DF = 6, p80 % of CH4 emissions from wetlands occur in this way (e.g. Kreuzwieser et al. 2003; Cheng et al. 2006). Dense grasslands cover much of the Mary River wetland and accounted for approximately 80 % of the CH4 flux measured during the inundation season. Plant mediated transport of CH4 through the grasses likely accounts for the high CH4 fluxes. The lower
values at site 4 than at site 2 (both grassland) are attributed to a general change in substrate type from the wetland edge to the river channel, which has been observed in the neighbouring Fogg Dam wetland system (Beringer et al. 2013). The scarcity of GHG flux data from tropical floodplains outside Amazonia is an essential knowledge gap that needs addressing. The data presented in this study reveals significant fluxes and emphasises the necessity for their inclusion in global carbon budgets. The significance of tropical wetlands as GHG sources mean elucidating their flux rates and controlling mechanisms is vital at a time when significant climatic change is occurring. Acknowledgments The staff of the Wildman Ranger station are thanked for their assistance and willingness to provide logistical advice. We thank owners and staff of the Wildman Wilderness lodge for providing access to the study site. This manuscript was significantly improved by the insightful recommendations of two anonymous reviewers. This work was funded by ARC Discovery project ’Greenhouse gas emissions from tropical floodplains’ (DP110103364).
References Adhya TK, Mishra SR, Rath AK et al (2000) Methane efflux from ricebased cropping systems under humid tropical conditions of eastern India. Agric Ecosyst Environ 79:85–90 Ballester MVR, Santos JE (2001) Biogenic gases in a tropical floodplain river. Braz Arch Biol Technol 44:141–147 Banker BC, Hludze HK, Alford DP, DeLaune RD, Lindau CW (1995) Methane sources and sinks in rice paddy soils: relationship to emissions. Agric Ecosyst Environ 53:243–251 Bass AM, Bird MI, Liddell MJ, Nelson PN (2011) Fluvial dynamics of dissolved and particulate organic carbon during periodic discharge events in a steep tropical rainforest catchment. Limnol Oceanogr 56: 2282–2292 Bass AM, O’Grady D, Berkin C, Leblanc M, Tweed S, Nelson PN, Bird MI (2013) High diurnal variation in dissolved inorganic C, δ13C values and surface efflux of CO2 in a seasonal tropical floodplain. Environ Chem Lett. doi:10.1007/s10311-013-0421-7 Belger L, Forsberg BR, Melack JM (2011) Carbon dioxide and methane emissions from interfluvial wetlands in the upper Negro River basin, Brazil. Biogeochemistry 105:171–183 Beringer J, Livesley SJ, Randle J, Hutley L (2013) Carbon dioxide fluxes dominate the greenhouse gas exchanges of a seasonal wetland in the wet-dry tropics of northern Australia. Agric For Meteorol. doi:10. 1016/j.agrformet.2013.06.008 Cao M, Gregson K, Marshall S (1998) Global methane emissions from wetlands and its sensitivity to climate change. Atmos Environ 12: 3291–3299 Cheng W, Yagi K, Sakai H, Kobayashi K (2006) Effects of elevated atmospheric CO2 concentrations on CH4 and N2O emission from rice soil, An experiment in controlled-environment chambers. Biogeochemistry 77:351–373 Devol AH, Richey JE, Clark WA, King SL, Martinelli LA (1988) Methane emissions to the troposphere from the Amazon floodplain. J Geophys Res 93:1583–1592 Devol AH, Richey JE, Forsberg BR, Martinelli LA (1994) Environmental methane in the Amazon River floodplain. In: Mitsch WJ (ed) Global wetlands: old world and new. Elsevier Science, Amsterdam, pp 151–165
Wetlands Engle D, Melack JM (2000) Methane emissions from an Amazon floodplain lake: enhanced release during episodic mixing and during falling water. Biogeochemistry 51:71–90 Finlayson CM (2005) Plant ecology of Australia’s tropical floodplain wetlands: a review. Ann Bot-Lond 96:541–555 Garnet KN, Megonigal P, Litchfield C (2005) Physiological control of leaf methane emissions from wetland plants. Aquat Bot 81:141–155 Gauci V, Gowing DJG, Hornibrook ERC, Davis JM, Dise MB (2010) Woody stem methane emission in mature wetland alder trees. Atmos Environ 44:2157–2160 Husin YA, Murdiyarso D, Khali MAK et al (1995) Methane flux from Indonesian wetland rice: the effects of water management and rice variety. Chemosphere 31:3153–3180 Kemenes A, Forsberg BR, Melack JM (2011) CO2 emissions from a tropical hydroelectric reservoir (Balbina, Brazil). J Geophys Res 116. doi:10.1029/2010JG001465 King JY, Reeburgh WS, Regli SK (1998) Methane emissions and transport by arctic sedges in Alaska: results of a vegetation removal experiment. J Geophys Res 103:29083–29092 Kreuzwieser J, Buchholz J, Rennenberg H (2003) Emission of methane and nitrous oxide by Australian mangrove ecosystems. Plant Biol 5: 423–431 Kuehn KA, Steiner D, Gessner MO (2004) Diurnal mineralization patterns of standing-dead plant litter: implications for CO2 flux from wetlands. Ecology 85:2504–2518 Livesley SJ, Andrusiak SM (2012) Temperate mangrove and salt marsh sediments are a small methane and nitrous oxide source but important carbon store. Estuar Coast Shelf Sci 97:19–27 Lund M, Lafleur PM, Roulet NT et al (2010) Variability in exchange of CO2 across 12 northern peatland and tundra sites. Glob Chang Biol 16:2436–2448 MacIntyre S, Wanninkhof R, Chanton J (1995) Trace gas exchange across the air-water interface in freshwater and coastal marine environments. In: Matson P, Harris R (eds) Biogenic trace gases: measuring emissions from soil and water. Blackwell, Cambridge, pp 52–97 Melack JM, Hess LL, Gastil M, Forsberg BR, Hamiltons SK, Lima IBT, Novo EMLM (2004) Regionalization of methane emissions in the Amazon Basin with microwave remote sensing. Glob Chang Biol 10:530–544
Miller RL (2011) Carbon gas fluxes in re-established wetlands on organic soils differ relative to plant community and hydrology. Wetlands 31: 1055–1066 Mitsch WJ, Gosselink JG (2000) Wetlands third edition. John Wiley and Sons, New York Nahlik AM, Mitsch WJ (2011) Methane emissions from tropical freshwater wetlands located in different climatic zones of Costa Rica. Glob Chang Biol 17:1321–1334 Petit NE, Bayliss P, Davies PM, Hamilton SK, Warfe DM, Bunn SE, Douglas MM (2011) Seasonal contrasts in carbon resources and ecological processes on a tropical floodplain. Freshw Biol 56:1047– 1064 Petrescu AMR, van Beek LPH, van Huissteden J, Prigent C, Sachs T, Corradi CAR, Parmentier FJW, Dolman AJ (2010) Modeling regional to global CH4 emissions of boreal and arctic wetlands. Global Biogeochem Cycles 24. doi:10.1029/2009GB003610 Richey JE, Melack JM, Aufdenkampe AK, Ballester MVR, Hess LL (2002) Outgassing from Amazonian rivers and wetlands as a large tropical source of atmospheric CO2. Nature 416:617–620 Rosenqvist A, Forsberg BR, Pimentel T, Rauste YA, Richey JE (2002) The use of spaceborne radar data to model inundation patterns and trace gas emissions in the central Amazon floodplain. Int J Remote Sens 33:1303–1328 Shindell DT, Walter BP, Faluvegi G (2004) Impacts of climate change on methane emissions from wetlands. Geophys Res Lett. 31. doi:10. 1029/2004GL021009 Vachon D, Prairie YT, Cole JG (2010) The relationship between nearsurface turbulence and gas transfer velocity in freshwater systems and its implications for floating chamber measurements of gas exchange. Limnol Oceanogr 55:1723–1732 Wasson RJ (1992) Modern sedimentation and late quaternary evolution of the Magela Creek plain. Supervising scientist for Alligator Rivers Region Report 6. Australian Government Publishing Service, Canberra Whalen SC (2005) Biochemistry of methane exchange between natural wetlands and the atmosphere. Environ Eng Sci 22:73–93 Zhou L, Zhou G, Jia Q (2009) Annual cycle of CO2 exchange over a reed (Phragmites australis) wetland in Northwest China. Aquat Bot 91: 91–98
