Article Is Available On- D
Total Page:16
File Type:pdf, Size:1020Kb
Biogeosciences, 17, 4025–4042, 2020 https://doi.org/10.5194/bg-17-4025-2020 © Author(s) 2020. This work is distributed under the Creative Commons Attribution 4.0 License. Environmental controls on ecosystem-scale cold-season methane and carbon dioxide fluxes in an Arctic tundra ecosystem Dean Howard1,2, Yannick Agnan3, Detlev Helmig1, Yu Yang4, and Daniel Obrist2 1Institute of Arctic and Alpine Research, University of Colorado Boulder, Boulder, CO, USA 2Department of Environmental, Earth and Atmospheric Sciences, University of Massachusetts Lowell, Lowell, MA, USA 3Earth and Life Institute, Université Catholique de Louvain, Louvain-la-Neuve, Belgium 4Department of Civil and Environmental Engineering, University of Nevada, Reno, NV, USA Correspondence: Daniel Obrist ([email protected]) Received: 6 November 2019 – Discussion started: 18 November 2019 Revised: 28 June 2020 – Accepted: 2 July 2020 – Published: 10 August 2020 Abstract. Understanding the processes that influence and vide valuable insight into how a changing Arctic climate control carbon cycling in Arctic tundra ecosystems is essen- may impact methane emission, and highlight a need to focus tial for making accurate predictions about what role these on soil temperatures throughout the entire active soil profile, ecosystems will play in potential future climate change sce- rather than rely on air temperature as a proxy for modelling narios. Particularly, air–surface fluxes of methane and car- temperature–methane flux dynamics. bon dioxide are of interest as recent observations suggest that the vast stores of soil carbon found in the Arctic tundra are becoming more available to release to the atmosphere in the form of these greenhouse gases. Further, harsh winter- 1 Introduction time conditions and complex logistics have limited the num- ber of year-round and cold-season studies and hence too our Active-layer soils and permafrost soils in the Arctic per- understanding of carbon cycle processes during these peri- mafrost region contain significant stores of terrestrial or- ods. We present here a two-year micrometeorological data ganic carbon. These ecosystems account for an estimated set of methane and carbon dioxide fluxes, along with support- 1307 (1140–1476) Pg of organic carbon, with ∼ 1035 Pg ing soil pore gas profiles, that provide near-continuous data found within soils between 0 and 3 m depth (Hugelius et al., throughout the active summer and cold winter seasons. Net 2014). Recently observed increases in surface air tempera- emission of methane and carbon dioxide in one of the study ture within these regions (Polyakov et al., 2002) have sparked years totalled 3.7 and 89 g C m−2 a−1 respectively, with cold- interest in the biogeochemical cycling of this carbon store, season methane emission representing 54 % of the annual as substrate metabolic activity – shown to be positively cor- total. In the other year, net emission totals of methane and related to temperature – can break down organic compounds carbon dioxide were 4.9 and 485 g C m−2 a−1 respectively, in the soil, releasing soil organic carbon to the atmosphere in with cold-season methane emission here representing 82 % the form of carbon dioxide and methane (Lai, 2009). Further- of the annual total – a larger proportion than has been previ- more, the “active layer” horizon, within which most soil car- ously reported in the Arctic tundra. Regression tree analysis bon decomposition takes place, has been observed in places suggests that, due to relatively warmer air temperatures and to be expanding as the underlying permafrost thaws under deeper snow depths, deeper soil horizons – where most mi- the influence of a warming atmosphere, thus exposing larger crobial methanogenic activity takes place – remained warm quantities of organic carbon to decomposition (Schuur et al., enough to maintain efficient methane production whilst sur- 2009; Romanovsky et al., 2010; Schuur et al., 2015; Vonk face soil temperatures were simultaneously cold enough to and Gustafsson, 2013). limit microbial methanotrophic activity. These results pro- Production of methane in the carbon-rich soils of the Arc- tic tundra takes place as a result of microbial metabolic Published by Copernicus Publications on behalf of the European Geosciences Union. 4026 D. Howard et al.: Two years of ecosystem-level methane flux measurements in the Arctic tundra activity (Lai, 2009). The breakdown of organic carbon to change budget, suggesting a predominant role of cold-season form methane is a complex process that requires contribu- processes (Zona et al., 2016). At five Alaskan sites (four tions from various taxa of microorganisms (Whalen, 2005). on coastal plains and tundra dominated by sedges, grasses, Methanogens form the last step in this process of producing and mosses within the northern coastal region surrounding methane from organic polymers (Conrad, 1999); whilst these Utqiagvik˙ and one over tussock-sedge, dwarf-shrub, moss methanogens encounter oxygen in the breaking down of ac- tundra at Ivotuk on the north slope of the Brooks Range, etate and carbon dioxide, they cannot survive in oxic environ- Walker et al., 2005), substantial methane emission was re- ments (Whalen, 2005; Kamal and Varma, 2008; Lai, 2009). ported to have occurred throughout the cold season, and al- As such, methanogenesis in peatlands is an obligate anaero- though emission rates were lower than those measured dur- bic process that takes place largely within deeper, anoxic lay- ing the warmer summer period, prolonged wintertime activ- ers of the soil, generally below the water table level (Le Mer ity amounted to 50 % ± 9 % (mean ±95 % confidence inter- and Roger, 2001). Methane production at these depths cre- val) of annual emission. The authors suggest that this cold- ates concentration gradients that lead to upward diffusion of season emission may become more important under fore- methane through the soil to the surface (Preuss et al., 2013). casted climate conditions that include higher air temperatures Furthermore, methane can be transported to the surface via and deeper snowpacks (Hay and McCabe, 2010; Zona et al., ebullition and through aerenchymatous tissues within some 2016). vascular plants (Joabsson et al., 1999; Lai, 2009). We present here a two-year micrometeorological methane In the upper soil layers, another subset of microorganisms and carbon dioxide exchange data set, undertaken over an known as methanotrophs consume a portion of this produced acidic tussock tundra site near the Toolik Field Station, methane in the presence of oxygen, eventually oxidising it to Alaska, USA, on the north slope of the Brooks Range. Com- carbon dioxide (Lai, 2009). These methanotrophs are largely plementary to air–surface exchange measurements, we re- exposed to methane diffusing through the soil pore space, as port soil pore space methane, carbon dioxide and oxygen ebullition is too quick to allow exposure to methanotrophs concentrations, and soil water content in the upper 40 cm, (Boone, 2000) and plant vascular transport shields methane as well as soil temperature profiles at and near the site from methanotrophic activity (Schimel, 1995; King et al., to a depth of 150 cm. All measurement systems were de- 1998; Verville et al., 1998). The rate of methane consump- ployed year-round, providing near-continuous data coverage tion is usually highest immediately above the water table throughout both the summer growing seasons and the cold level, where high concentrations of methane formed from the winter seasons in both years. The goal of this study was underlying methanogens meet with sufficient oxygen levels to investigate environmental controls that significantly im- from the overlying atmosphere (Dedysh et al., 2002). Rates pact the magnitude and direction of methane fluxes in this of both methanogenesis and methanotrophy are highly de- environment – particularly over the colder winter months pendent on temperature, with optimal metabolic rates (as – as environmental-control–methane-flux relationships dur- determined in the laboratory) occurring at temperatures of ing these periods are relatively poorly understood. These re- around 25 ◦C(Dunfield et al., 1993). Of the two competing sults expand our coverage of year-round methane and car- processes, methanogenesis has shown to be more tempera- bon dioxide exchange data sets across different bioclimates ture dependent, with higher reported rate changes per unit and landscapes, as well as add to our growing understanding ◦ warming (e.g. per 10 C (Q10): 5.3–16) compared to methan- of carbon exchange dynamics in the Arctic tundra soils, and otrophy (Q10: 1.4–2.1) (Dunfield et al., 1993). provide insights into how these dynamics may evolve under Estimating the methane exchange budget in Arctic tundra forecasted changing conditions in this region. ecosystems and how it relates to temperature are challeng- ing objectives, currently subject to considerable uncertain- ties. Ambient observations in northern Alaska over 29 years 2 Methods showed no clear increase in ambient atmospheric methane concentration enhancements during this period, despite no- 2.1 Site description ticeably warmer air temperatures (Sweeney et al., 2016). Direct observations of methane exchange, however, during This study was performed over two full years from Octo- the Carbon in Permafrost Experimental Heating Research ber 2014 to September 2016 at Toolik Field Station (Alaska, (CiPEHR) project, showed significant increases in methane USA) located on the north slope of the Brooks