| Description: | Three field campaigns with three replicate sampling dates within each campaign were conducted in July (winter), September (spring), and December (summer) 2018. In each stream, a 60 to 100 m reach of the stream channel was selected for characterizing the physico-chemical properties and quantifying methane concentrations of stream water at three sub-stations and for conducting tracer experiments to quantify gas exchange velocity across the air-atmosphere interface using five sub-stations.
- Physical and chemical characteristics of stream water
In-situ physico-chemical characterization of basic water quality parameters were conducted using a YSI Professional Plus handheld multiparameter meter (YSI, Yellow Springs, OH, USA) to determine dissolved oxygen (DO - mg L-1), water temperature (Temp – oC), pH, electrical conductivity (EC - µS cm-2) and oxidation-reduction potential (ORP - mV). Water samples were collected in 50ml Falcon tubes for laboratory analysis and transported at low temperature to the laboratory. In the laboratory, total dissolved carbon, dissolved inorganic carbon, dissolved organic carbon, and total dissolved nitrogen (nitrogen compounds in filtered samples were converted to NO at 720ºC) were determined using a TOC-L Shimadzu TOC analyzer (Shimadzu Co., Kyoto, Japan), coupled with a TNM-L total nitrogen measuring unit (Shimadzu Co., Kyoto, Japan). The conservative tracer method was used to estimate mean stream water velocity and streamflow (Stream Solute Workshop, 1990). For this, NaCl pulses were released upstream and a YSI Professional Plus conductivity meter (YSI, Yellow Springs, OH, USA) recorded downstream conductivity every 15 s starting just prior to the addition and continued until conductivity returned to baseline conditions that had been observed before the NaCl addition (Webster and Valett, 2007).
- Methane sampling and analysis
Methane samples were collected in triplicate at each sub-station of each stream (9 samples per stream, 54 samples per field sampling campaign) using the headspace extraction technique (Schade et al., 2016). We were unable to collect samples in “sugarcane 3” stream during the spring and summer 2018, because of authorization issues. 60-mL acid-washed syringes fitted with sealed three-way stopcocks were filled with 30 mL ultra-pure nitrogen (5.0) in the lab. In the field, one syringe was filled with 30 mL stream water in each station. The stopcock was closed underwater to avoid any bubbles. Syringes were then shaken for five minutes to equilibrate gases between water and atmosphere, and the entire headspace gas was injected into a pre-evacuated gas-tight vial for methane analysis. Gas samples were stored under positive pressure until analysis.
Methane analysis was carried out using a Shimadzu GC-2014 gas chromatograph (Shimadzu Co., Kyoto, Japan) equipped with a flame ionizer detector (FID – detection limit: 3 pgC s-1) operating at 325 ºC (Bowden et al., 1991). Gas concentrations were calculated by comparing peak areas for samples with standards (Scott-Marrin – Riverside, CA, USA - 0.968 ppm, 1.842 ppm and 3.582 ppm) calibrated against standards prepared by the National Oceanic and Atmospheric Administration/Climate Monitoring and Diagnostic Laboratory (NOAA/CMDL – Boulder, CO, USA). The concentrations of methane in the headspace were converted to partial pressures of methane in the initial water samples using Bunsen solubility coefficients (Mulholland et al., 2004).
- Methane flux
Methane flux (F) was calculated multiplying gas transfer velocities (K) by the difference between the measured dissolved concentration (Cw) and the predicted methane concentration at equilibrium with the atmosphere (Ceq) (Equation 1; Beaulieu et al., 2016; Raymond et al., 2012; Schade et al., 2016).
Equation 1:
F=K(Cw-Ceq)
Where F = methane flux; K= gas transfer velocity; Cw-Ceq = difference between the measured and predicted equilibrium concentrations .
Gas transfer velocities were estimated for two field campaigns (July and December) using the gas tracer method (Tsivoglou and Neal, 1976) in all studied streams . SF6 and NaCl were employed as conservative gas and solute tracers, respectively. SF6 was continuously bubbled at an upstream station, and the NaCl pulses were used to indicate the time needed for the stream channel to become saturated with SF6. Once saturation was reached, gas samples were collected at five sub-stations using the methods described earlier and analyzed using a gas chromatograph Shimadzu GC-2014 (Shimadzu Co., Kyoto, Japan). We used the downstream decline in SF6 concentrations to estimate dissolved oxygen transfer at the air-water interface (i.e., reaeration coefficient). Reaeration coefficients were calculated and converted to gas transfer velocities using published methods and protocols (Canale et al., 1995; Raymond et al., 2012; Wanninkhof et al., 1990). |
