Greenhouse gases in an urban river: Trend, isotopic evidence for underlying processes, and the impact of in-river structures

Highlights

• Urban river impact by in-river structures is an important GHGs source.

• WWTP effluent contributes large amount of dissolved N₂O and CO₂ to river.

• WWTP effluent and dam create favorable condition for N₂O consumption.

• Dam can stimulate development of anoxic condition and production of CO₂ and CH₄.

• Aeration device is beneficial to reduce CH₄ production.

Abstract

Urban rivers are important sources of greenhouse gases (GHGs) to the atmosphere; however, these high dissolved GHGs are not well studied. We report the first detailed investigation on an urban river influenced by wastewater treatment plant (WWTP), dam and aeration device, with a focus on the concentration of GHGs and their underlying microbial processes basing on isotope approaches. The result revealed that these infrastructures influenced carbon and/or nitrogen input and redox condition of the river system, leading to impaction of GHGs concentrations and spatiotemporal pattern. The WWTP effluent was the point-source of N₂O loading to rivers, and also contained high NO₃^– and dissolved organic carbon, creating favorable conditions for denitrification by microbes. We measured isotopic compositions (δ¹⁵N-N₂O and δ¹⁸O-N₂O) and site preference (SP) of dissolved N₂O to differentiate the relative contribution of N₂O production processes and calculate the degree of N₂O reduction (Fr). The relative contribution of denitrification increased at downstream and Fr suggested that N₂O was produced and consumed by heterotrophic denitrification, resulting in a sink of N₂O downstream. The closure of dam increased water residence time, trapped the sediment, accelerated oxygen consumption, and development of anoxic conditions that stimulating the production of CO₂ and CH₄, as well as consumption of N₂O. Aeration devices maintained oxic condition, that slowing the development of the anoxic stage and eliminating the methanic stage at downstream. As a result, CO₂ concentrations increased along the river due to production of CO₂ via anaerobic and aerobic respiration of aqueous organic matter. CH₄ concentrations increased at downstream from organic matter biodegradation under anoxic conditions. The measured dissolved GHGs, especially the upper ranges, correspond to a large oversaturation of surface water with respect to atmospheric equilibrium, leading high fluxes to the atmosphere. Our values are on the high end of estimates reported in the literature among different climate regions, indicating that urban river influenced by various of infrastructures are comparable to those in wetlands and intensively agricultural rivers in terms of GHGs emissions. Urban river should be urgently included in the determination of global carbon and nitrogen budgets from rivers.

Introduction

Carbon oxide (CO₂), methane (CH₄) and nitrous oxide (N₂O) are important greenhouse gases (GHGs), contributing to global climate change (IPCC, 2013). Rivers are important GHGs emission sources because receive water from sewage discharge, agricultural soil, wetland and groundwater, which contain GHGs and its precursors. The sources of these potent GHGs are poorly constrained and quantified (Sutton et al., 2007), especially those related to release from rivers and streams (Marzadri et al., 2017). The CO₂ emissions from rivers and streams have been updated to 1.8 Pg C/y (Raymond et al., 2013), while other recent work has provided an even lower estimate of 0.65 Pg C/y from global rivers and streams (Lauerwald et al., 2013). There is also a lack of consensus on both the magnitude of the anthropogenic contribution and the general importance of lotic systems to N₂O emissions. The United States Environmental Protection Agency (EPA) (Anderson et al., 2010) estimates that the natural nitrous oxide emissions from rivers are 0.1 Tg N − N₂O yr⁻¹, while a recent large-scale tracer study suggests that rivers account for at least 0.68 Tg N − N₂O yr⁻¹, representing up to 10% of global anthropogenic N₂O emissions (Hu et al., 2016, Beaulieu et al., 2010). Annual CH₄ evasion from fluvial environments is estimated to be 26.8 Tg CH₄, which is comparable to the values for both lakes and wetlands (Stanley et al., 2015). Methane dynamics have been well documented in inland water, such as lakes and reservoirs. However, less is known about riverine CH₄ dynamics (Atkins et al., 2017). The prevailing large uncertainty involved in GHG flux estimates for rivers is essentially due to two reasons. This first reason is the paucity of available data, particularly that related to low-order rivers. Although small rivers are individually trivial GHG sources, they are likely to be more supersaturated in GHGs than large rivers (Zeng and Masiello, 2010, Li et al., 2018). The second reason is an incomplete understanding of the underlying processes leading to these emissions (Quick et al., 2019).

Riverine N₂O production and consumption processes include nitrification (the oxidation of NH₃ to NO₃^–, release N₂O as a by-product), nitrifier denitrification (the reduction of NO₂^– to N₂O) and heterotrophic denitrification (the reduction of NO₃^– to N₂, can either produce N₂O and reduce N₂O to N₂) (Toyoda et al., 2015, Li et al., 2019). Production processes of riverine CO₂ include biodegradation of either from terrestrial organic matter (OM) and aqueous OM, as well as respiration of phytoplankton, while riverine CO₂ can be deceased by processes of evasion to atmosphere and uptake by phytoplankton (Hosen et al., 2014, Crawford et al., 2016, Borges et al., 2019). Riverine CH₄ is produced by methanogenesis that via reduction of CO₂ or decomposition of acetate under anaerobic condition, while is consumed by oxidation (the oxidation of CH₄ to CO₂) (Atkins et al., 2017). Thereby, GHGs concentrations are supposed to related with environmental factors such as temperature, organic matter, dissolved oxygen (DO), nitrate, ammonia and salinity, which can be attempted with a correlation analysis with underlying process (Borges et al., 2015). However, the relation between GHGs and those factors are not always simple. Thus, underlying processes of the GHGs were generally identified by incubating sediment and water in laboratory conditions, despite that do not truly reflect GHGs production in the field setting of rivers (Thuan et al., 2017). Stable isotopic technique have been recommended as a powerful tool for addressing this difficult issue (Li et al., 2019).

Nitrogen and oxygen isotopic ratios (δ¹⁵N-N₂O and δ¹⁸O-N₂O) and site preference (SP, the difference in ¹⁵N abundance between central and outer N position of N₂O) have been adopted to differentiate N₂O production mechanism in agricultural soils, partial nitrification reactors, and wastewater treatment plants (WWTP) (Koba et al., 2009, Lewicka-Szczebak et al., 2017, Zou et al., 2014). In particular, SP value is only determined by production processes and not dependent on δ¹⁵N the substrate (NO₃^– or NH₄⁺) isotopic composition, which makes it universally applicable (Sutka et al., 2006). Generally, for NH₂OH oxidation (nitrification), the SP value is expected to be positive (33 ± 4 ‰) because ¹⁴N binds preferentially to hydroxylamine oxidoreductase (Toyoda et al., 2015). For NO reduction (denitrification and nitrifier denitrification), the SP is expected to be close to zero because two NO molecules simultaneously bond to the Fe centers of nitric oxide reductase (Denk et al., 2017). Additionally, δ¹⁵N^bulk-N₂O, δ¹⁸O-N₂O and SP will be enriched during the process of N₂O reduction by complete denitrification, which can be described by Rayleigh fractionation model (Wunderlin et al., 2013, Ishii et al., 2014, Thuan et al., 2017). The sable isotopic ratio also can provide insight into origin of CO₂ through the determination of δ¹³C in dissolved inorganic carbon (DIC) (Polsenaere and Abril, 2012). The δ¹³C of atmospheric CO₂ is about −7.5 ‰, whereas CO₂ produced by respiration has a δ¹³C value close to terrestrial and aqueous OM (i.e., −27 ‰ in case of C₃ plants and −13‰ in case of C₄ plants) (Miyajima et al., 2009). HCO₃^– is another major component of DIC, and the δ¹³C of HCO₃^– is controlled by type of weathering rock (carbonate and silicate) and origin of the CO₂ involved in rock dissolution (Polsenaere and Abril, 2012). Since HCO₃^– is isotopically enrichment relative to CO₂ from OM respiration, the change of δ¹³C-DIC is supposed to follow the relative abundance of CO₂ to HCO₃^– which through the processes including physical (gas exchange with atmosphere, weathering, ice cover) and biological (respiration and photosynthesis) (Roach et al., 2016). CH₄ from biogenic methanogenesis usually has δ¹³C values lower than −50 ‰, whereas oxidation extends from about −50 to −20 ‰ because of enrichment of oxidation process (Teodoru et al., 2015, Mackensen and Schmiedl, 2019; Andreas et al., 2019). The fraction of CH₄ removed by methane oxidation can be calculated with a closed-system Rayleigh fractionation model instead of open-system models (Liptay et al., 1998). Identification of these underlying processes will illuminate potential role of river on nutrient cycling, water quality, and GHGs budgets (Smith et al., 2017).

Urban river is highly human activities modified system, with various of in-river structures (i.e. dam, aeration device, WWTP, wetland, and sanitary sewer). The closure of dam can increase water residence time, trap the sediment, accelerate oxygen consumption, and development of anoxic conditions which can affect the production and consumption of GHGs (Maeck et al., 2013, Maavara et al., 2018). WWTP is known to be a source of N₂O, CH₄ and CO₂ in urban areas and contribute point-source GHGs loading to rivers (Strokal and Kroeze, 2014, Alshboul et al., 2016). It is believed that DO can be increased significantly by aeration device, resulting decrease of nitrogen, phosphorus and chlorophyll α (Wu et al., 2016). The overall result suggested that these in-river structures influenced carbon and/or nitrogen input and redox condition of the river system, leading to change of GHGs concentrations and spatiotemporal pattern (Maeck et al., 2013, Smith et al., 2017, Jin et al., 2018). In addition, dissolved GHGs were supersaturated in urban rivers which were significantly higher than those in the natural waters, creating GHGs emission hot spot (Alshboul et al., 2016, Smith et al., 2017, Hu et al., 2018, Abril and Borges, 2019). Therefore, further research is required to quantified the GHGs concentrations, identified the underlying processes related with GHGs and the controlling factors which responsible for the observed patterns of urban river.