Supporting Information
Campeau et al., 2014
This supporting information includes additional methods detailing the sampling
schedule with an illustration of the hydrological condition in which the systems where
sampled (Figure S1) (Section S1). We also present the detailed calculations of the surface
water pCO2 and pCH4 (Section S2), gas fluxes and gas exchange coefficient (k600) derived
from floating chamber measurements (Section S3) and the non-strictly diffusive
component of the CH4 fluxes captured in the floating chamber (Section S4). This
document also includes a detailed section about the digital elevation model (DEM)
interpolation, used to estimate the rivers and streams area in the region (Table S1) (Section
S5). The methods and equations presented in this document can be use by other
researchers to quantify similar component of rivers and streams regional scale area and C
gas fluxes.
1
S.1 Sampling schedule and hydrological conditions
Rivers and streams were sampled during 1 to 9 different synoptic surveys. The
periods during which those surveys took place were chosen in order to capture some of
the major stream hydrological features. A hydrograph illustrating the hydrological
conditions (water level and temperature) for three streams that we continuously
monitored in the Abitibi region is presented in Figure S1. The different sampling periods
are indicated with gray areas on the x-axis.
Figure S1: Hydrograph illustrating the hydrological conditions, water level at gauging
station and water temperature, for 3 different monitored streams in the Abitibi region,
between May 2010 to August 2011. Gray sections on the x axes indicate sampling
periods. No data was recorded between January and March 2011.
2
S.2 Measurements of surface water pCO2 and pCH4
Surface water pCO2 and pCH4 were measured using the headspace equilibrium
method. 60 ml polypropylene syringe were used to collect 30ml of stream water from
approximately 10cm bellow the surface and added this volume to 30ml of ambient air to
create a 1:1 ratio of ambient air: stream water. For pCO2, triplicate syringes were
vigorously shaken for 1 minute, in order to equilibrate the gases in water and air. The
resulting headspace was directly injected into an infrared gas analyzer (PP Systems,
EGM-4). The original surface water pCO2 was then calculated based on the headspace
ratio and the in situ measured ambient air pCO2 using the following equation:
Ambient pCO2 = (HpCO2–ApCO2)*(A:W)/ Vm +(HpCO2 * AkH)/SkH
(S1)
where HpCO2 is the equilibrated pCO2 in the headspace (µatm), ApCO2 is atmospheric
pCO2 (µatm), A:W is the air to water volume ratio in the syringe, Vm is the gas molar
volume at analysis temperature, AkH is Henry’s constant at analysis temperature, and SkH
is Henry’s constant at the sample temperature. The atmospheric concentration for CO2
(ApCO2) was determined locally and for every site using the EMG-4.
A similar procedure was used for collection of surface water pCH4. The resulting
30 ml headspace, however, was rather injected into 30-mL glass vials equipped with
rubber stoppers (20mm of diameters red bromobuptyl) filled with saturated saline solution
(David Bastviken, personal communication), and kept inverted until analysis. In the lab,
the gas in the headspace of the vials was injected into a Shimadzu GC-8A Gas
chromatograph with FID (flame ionization detector), to determine its CH 4 concentration.
3
The original surface water pCH4 was then calculated according to the headspace ratio
(eq.S1) and assuming a constant ambient air pCH4 of 1.77 µatm [Denman et al., 2007].
S.3 Determination of pCO2, kCO2 and CO2 fluxes
The circular plastic chamber, with a surface area of 0.09 m2 and volume of 16 L,
was covered with aluminum foil to reduce solar heating and equipped with an internal
thermometer to monitor temperature changes that may affect the exchange rates during the
measurements. The floating chamber was connected to an infrared gas analyzer (PPsystem, model EGM-4) via enclosed re-circulating system providing a continuous reading
of the CO2 concentrations in the chamber. The rates of change in pCO2 in the chamber
were used to estimate ƒCO2 (mmol m-2 d-1) with the following equation:
æ s *V ö
fCO 2 = ç
*t
è Vm * S ÷ø
(S2)
where (s) is the rate of change of the gas in the chamber (µatm min-1), (V) is the volume of
the chamber in liters (L), S is the surface area of the chamber (m2), (Vm) (molar volume) is
the molar volume of one mole of gas (L mol-1), and (t) is a conversion factor from minutes
to day (1d = 1440 min).
ƒCO2 was further used to estimate kCO2 by inverting the equation describing Fick’s
law, following Vachon et al. [2010] as follows:
kCO 2 =
fCO 2
kH ( pCO2water - pCO2air )
(S3)
4
where kCO2 is the gas transfer velocity specific for CO2 (in m d-1); ƒCO2 is the measured
CO2 flux between the surface water and the atmosphere in the floating chamber, kH is
Henry’s constant adjusted for salinity and temperature, and ∆pCO2 is the difference
between CO2 partial pressure in the surface waters and the atmosphere (µatm).
To simplify the exploration of regional patterns of gas exchange across the fluvial
network, we standardized kCO2 to a Schmith number of 600 to derive a k600, with
following the equation from Jähne et al. [1987]:
k600 =
kCO 2
(ScCO 2 / 600)- n
(S4)
where ScCO2 is the CO2 Schmidt number at the stream water temperature [Wanninkhof,
1992], and n was fixed at 2/3, which corresponds to surface roughness resulted at wind
speed of <3.7m s-1 accordingly to [Guérin et al., 2007].
5
S.4 Quantifying diffusive and non-diffusive CH4 fluxes
Due to the low solubility of CH4 in freshwater, ƒCH4 to the atmosphere not only
occurs via diffusion but also through none-diffusive pathways, such as ebullition, which
may contribute significantly to the total fluvial CH4 emissions [Bastviken et al., 2004;
Baulch et al., 2011]. In this regard, the kCH4 derived from the ƒCH4 measured in the floating
chambers may reflect both diffusive fluxes as well as potentially non-diffusive ƒCH4, and
for this reason the resulting empirical kCH4 are often much higher than those predicted by
Fick’s law relative to k600 [Prairie and del Giorgio, 2013]. In order to quantify the
contribution of non-diffusive ƒCH4 to the overall ƒCH4, we first calculated the theoretical
diffusive kCH4 on the basis of our empirically-determined k600, using equation A4 adjusted
for CH4.
kCH 4 =
k600
(600 / ScCH 4 )- n
(S5)
This yield a strictly diffusive kCH4 estimate, which we used to back-calculate a theoretical
diffusive CH4 flux (mmol m-2 d-1) with the following equation;
ƒDCH4= kCH4 * kH (pCH4water-pCH4air)
(S6)
where ƒDCH4 is the diffusive flux of CH4 at the interface water-air according to Fick’s law;
kCH4 is the gas transfer velocity of CH4 derived from the k600; kH is the Henry’s coefficient
(adjusted for ambient salinity and temperature); and the pCH4water and the pCH4air are the
partial pressure of CH4 in water and air (1.77 µatm), respectively. We then used the
difference between ƒDCH4 and the ƒCH4 measured from the floating chambers as an estimate
of the potential non-diffusive ƒCH4 [Prairie and del Giorgio, 2013].
6
S.5 Determination of total river and stream areal coverage
Our regional-scale estimate of total fluvial area was based on the tributary
classification system of Strahler stream orders, and on that basis calculated the total area
covered by each of the 6 different stream orders present in the fluvial network of both
regions. We performed a digital elevation model (DEM) interpolation to calculate the
length (m) of each stream order in the two regions. The DEM is based on the digitized
topography and generates a segment raster map representing a classified linear coverage
of stream and rivers according to their Strahler stream orders with a resolution of 100m 2.
To run this interpolation, we set a threshold value for water flow accumulation in the
landscape, which aimed to match as closely as possible to the length of the rivers and
streams available on digitized maps. However, the total cumulative length of streams and
rivers (from order 1-6), estimated with the DEM interpolation, was slightly higher when
compared to the available digitized data. We assumed that this difference represented the
small headwater streams (stream order 0), and further assumed that the DEM
interpolation was more accurate. In fact, the smallest streams sampled in our survey,
narrower than 0.5m, did not appear at all in the digitized maps, which supports our
assumption that digitized streams and rivers underestimated the true areal extend of the
fluvial network.
We finally combined the estimated stream length (m) per stream order with the
average channel width (m) corresponding to each stream order from the data collected on
the field for our 46 sampled sites. This yielded an estimate of 190 km2 occupied by the
fluvial network, representing approximately 0.4% of the total landscape area considered
in this study (44,182 km2) (Table AM1). This represents a drainage density of 1.06 km
7
km-2, and suggests a potential underestimation of the true streams areal coverage
[Benstead and Leigh, 2012]. The areal coverage of each Strahler stream orders [1 to 6]
was roughly evenly distributed across both regions, averaging 31.8 km2 (Table S1) The
largest areal contribution, however, was by stream order 1 (38.7 km2), which occupied
about 20% of the total fluvial network area (Table 3 in article). The smallest headwater
streams (stream order 0), which do not appear on digitized maps, were combined with the
stream order 1, but accounted only and 6% of the total fluvial network and 29% of the
surface of stream order 1.
The use of the DEM interpolation to estimate the rivers and stream coverage can
be problematic in regions of low topography such as Abitibi and James Bay. When
comparing our estimate to other studies, however, we find that our estimate are well within
the range of published values with fluvial coverage estimate, ranging from 0.9-0.11% in
Sweden [Humborg et al., 2010; Jonsson et al., 2007], 0.5% in the US [Butman and
Raymond, 2011], and 0.3 to 0.6% at the global scale [Downing et al., 2012]. It should be
mentioned that there is relatively high uncertainty associated to our estimates of river and
stream area (Table S1). Flowing water systems are extremely dynamic, especially in terms
of width that fluctuates among seasons, and also within streams in the same stream order
category. Methods and tools to properly assess river areal coverage is an essential
component in the determination of their biogeochemical function, and need to be further
advanced.
8
Table S1: Summary of the results from the DEM interpolation used to estimate fluvial network surface in Abitibi and James Bay
(44,182 km2). The table presents the distribution and characteristics for each Strahler stream orders in the fluvial network, the number
and average total stream length (TSL) of the different sites visited in each stream order category. Stream order zero represents the
category of stream that appeared in the digital elevation model (DEM) but did not figure on digitized maps. The uncertainty of the
estimated regional surface for each stream order was based on the changes in width of the streams during the different seasons and the
natural variability of width between streams of the same order.
Order
n
0
1
2
3
4
5
6
Total
5
14
9
6
7
2
3
46
Average
Average
Cumulative
Stream areal
Regional
TSL for
channel width regional stream
coverage
stream surface
sampled sites for sampled sites
length
2
in the region
(km )
(km)
(m)
(km)
(%)
0.33
1.74
7.06
39.44
156.85
337.73
3466.11
0.8
1.83
2.37
7.16
18.42
41.59
120.0
13 718
15 018
11 412
4 323
1 643
827
242
47 183
10.97 (±8.2)
27.48 (±20.3)
27.07 (±17.1)
30.96 (±29.1)
30.26 (±28.0)
34.41 (±14.5)
28.99 (±7.2)
190 (±124)
0.02%
0.06%
0.06%
0.07%
0.07%
0.08%
0.07%
0.43%
9
References
Bastviken, D., J. Cole, M. Pace, and L. Tranvik (2004), Methane emissions from
lakes: Dependence of lake characteristics, two regional assessments, and a global
estimate, Global Biogeochemical Cycles, 18, GB4009.
Baulch, H. M., P. J. Dillon, R. Maranger, and S. L. Schiff (2011), Diffusive and
ebullitive transport of methane and nitrous oxide from streams: Are bubble-mediated
fluxes important?, Journal of Geophysical Research, 116, G04028.
Benstead, J. P., and D. S. Leigh (2012), An expanded role for river networks, Nature
Geoscience, 5, 678-679.
Butman, D., and P. A. Raymond (2011), Significant efflux of carbon dioxide from
streams and rivers in the United States, Nature Geoscience, 4(12), 839-842.
Denman, K.L., G. Brasseur, A. Chidthaisong, P. Ciais, P.M. Cox, R.E. Dickinson, D.
Hauglustaine, C. Heinze, E. Holland, D. Jacob, U. Lohmann, S Ramachandran, P.L.
da Silva Dias, S.C. Wofsy and X. Zhang, (2007), Couplings Between Changes in the
Climate System and Biogeochemistry. In: Climate Change 2007: The Physical
Science Basis. Contribution of Working Group I to the Fourth Assessment Report of
the Intergovernmental Panel on Climate Change [Solomon, S., D. Quin, M. Manning,
Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Millet (eds.)] Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
Downing, J. A., J. J. Cole, C. A. Duarte, J. J. Middelburg, J. M. Melack, Y. T. Prairie,
P. Kortelainen, R. G. Striegl, W. H. McDowell, and L. J. Tranvik (2012), Global
abundance and size distribution of streams and rivers Inland Waters, 2(4), 229-236.
Guérin, F., G. Abril, D. Serça, C. Delon, S. Richard, R. Delmas, A. Tremblay, and L.
Varfalvy (2007), Gas transfer velocities of CO2 and CH4 in a tropical reservoir and its
river downstream, Journal of Marine Systems, 66(1–4), 161-172.
Humborg, C., C.-M. Mörth, M. Sundbom, H. Borg, T. Blenckner, R. Giesler, and V.
Ittekkot (2010), CO2 supersaturation along the aquatic conduit in Swedish watersheds
as constrained by terrestrial respiration, aquatic respiration and weathering, Global
Change Biology, 16, 1966-1978.
Jähne, B., G. Heinz, and W. Dietrich (1987), Measurement of the Diffusion
Coefficients of Sparingly Soluble Gases in Water, Journal of Geophysical Research,
92, 10767-10776.
Jonsson, A., G. Algesten, A. K. Bergström, K. Bishop, S. Sobek, L. J. Tranvik, and
M. Jansson (2007), Integrating aquatic carbon fluxes in a boreal catchment carbon
budget, Journal of Hydrology, 334, 141-150.
Prairie, Y. T., and P. A. del Giorgio (2013), A new pathway of freshwater methane
emissions and the putative importance of microbubbles, Inland Waters, 3(3), 311-320.
10
Wanninkhof, R. (1992), Relationship Between Wind Speed and Gas Exchange Over
the Ocean, Journal of Geophysical Research, 97, 7373-7382.
11