GHG collection and measurement
Surface water samples were collected with 120 mL glass serum bottles
at wrist depth for the determination of CH4 concentrations; samples
were collected in triplicate at each site. After poisoning with a
saturated solution of NaCl for CH4, the serum bottle was sealed with
butyl stoppers below the water surface, crimped with aluminum caps,
and stored at ambient temperature in the dark. CH4 concentrations were
measured using the headspace equilibration method44 on a gas
chromatography equipped with a flame ionization detector for CH4
(Agilent 7890B GC-FID). The partial pressure of CO2 (pCO2) in surface
water was measured in-situ using a Qubit Dissolved CO2 System equipped
with an internal pump and a pCO2 probe (S157-P, Qubit Biology Inc.,
Canada). The pCO2 probe is coated with a polytetrafluoroethylene
membrane sleeve that is permeable to CO2 gases but not to water. The
system was calibrated with standard CO2 gases ranging from 0 to to
10,000 μatm. Recalibration of the system was done before each round of
fieldwork.
Fluxes of CO2 and CH4 were measured simultaneously with surface water
sample collection. Local ambient air samples were also collected at
the same time. At each site, four floating chambers were deployed from
the shallow water near the river bank to the deep water in the
mid-channel to measure carbon gas fluxes. These chambers were of the
same size and shape and streamlined with a flexible plastic foil
collar to minimize the effects of chamber-induced turbulence when
measuring fluxes. Also, each chamber was covered with aluminum foil to
reflect sunlight and minimize internal heating. Chambers were allowed
to drift and each chamber measurement lasted for 60 to 80 min. After
mixing the chamber content 3 times, 50 mL of gas were extracted from
inside the chambers to air-tight gas sampling bags at the 0, 5, 10,
20, 40, 60, 80 min time intervals. This multi-chamber method and
prolonged deployment not only increased the probability of capturing
ebullition but also incorporated spatiotemporal variability in both
diffusion and ebullition within and among streams and rivers. CO2 and
CH4 concentrations of gas samples were determined as described above.
In addition to gas sampling, surface water samples were collected for
physicochemical analyses at each sampling sites. Sediment samples were
collected simultaneously alongside every gas and surface water
sampling for physicochemical and microbial analyses unless there were
only gravels/cobbles in the riverbed, and a total of 80 sediment
samples were obtained. Air temperature, atmospheric pressure, and wind
speed were measured in situ with a portable anemometer (Testo 480,
Germany). Water temperature, pH, DO, ORP, and conductivity were
measured in situ with portable field probes (Hach HQ40d).
Precipitation and solar radiation were obtained from National
Meteorological Information Center (http://data.cma.cn/). Flow
velocity, discharge, width, and mean depth of streams and rivers were
provided by hydrological stations at each sampling site.
Flux calculation
Total gas fluxes were calculated according to the following equation:
Ft = (nt – n0)/(A x t)
where, nt and no are the number of moles of carbon gases in the
chamber at time t and time zero (mol), respectively; A is the surface
area of water covered by the chamber (m2); and t is the measurement
duration time of drifting (min). Diffusive and ebullitive fluxes for
CH4 were separated using the Campeau et al. approach. Briefly, the Ft
for CO2 is assumed to be exclusively diffusive (that is, the CO2
ebullition is negligible). Only the linear section of the pCO2
increase curve during the first 10 min of chamber deployment was used
for FCO2 calculation to eliminate possible biases due to gas
accumulation in the chambers which will affect the flux rates. The
theoretical diffusive k for CH4 was calculated based upon kCO2 as
follows:
kCH4/kCO2 = (ScCH4/ScCO2)^-n
where, kCH4 and kCO2 are the gas transfer velocity of CH4 and CO2; Sc
is the Schmidt number and n is assigned a value of 1/2 for wind speed
> 3.6 m s^-1 or 2/3 for wind speed < 3.6 m s^-1. We then
calculated the theoretical diffusive flux for CH4 according to:
Fd = k (Cwater – Ceq)
where k is the gas transfer velocity (m·d^-1), Cwater is the water gas
concentration (mol·m^-3), and Ceq is the gas concentration in water in
equilibrium with the local atmosphere corrected for temperature
induced changes in solubility according to the Henry’s law (mol m^-3).
Thus, the difference between the total and diffusive fluxes is
attributed to ebullition.
