A new field approach for the collection of samples for aquatic 14
Introduction
40The importance of lakes, rivers and streams as globally significant sources of 41 greenhouse gases is being increasingly recognised (Borges et al., 2015). CO 2 evasion from 42 these inland waters has been estimated at ~2.1 Pg C yr -1 (Raymond et al., 2013), which is 43 broadly the same magnitude as the total terrestrial land carbon sink. Moreover, there is still 44 considerable uncertainty concerning the sources of aquatic CO 2 (Raymond et al., 2013),
45which must be addressed if we are to more accurately predict the response of the global 46 carbon cycle to change. This is particularly relevant to the Arctic regions where there is 47 evidence that "old" carbon is being released into aquatic systems in response to warming (e.g.
48Vonk et al., 2015).
-3-The isotopic composition of aquatic carbon can be used to identify the sources of (Billett et al., 2015), which is best determined through measurement of the 62 14 C content of aquatic CO 2 .
63The "gas stripping" method is frequently used to determine the isotopic composition 64 of dissolved inorganic carbon (DIC) and aquatic CO 2 (the latter via temperature and pH 65 dependent equilibria (Stumm and Morgan, 1981; Butler, 1982) and experimentally-derived 66 isotopic fractionation factors (Zhang et al., 1995)
137Immediately before use, the AWC was "collapsed" to remove atmospheric air and 138 then flushed with CO 2 -free air and evacuated using a 1 L "jumbo" syringe (Thames Restek 139 UK Ltd); the gas port of the syringe had a quick coupling attached via a ~3 cm length of
140Tygon tubing providing a leak-tight connection to the AWC. CO 2 -free air was generated on-141 site using the 1 L syringe by passing atmospheric air twice through a cartridge ( gas were syringe-injected into the same MSC, as calculated using:Where V is the volume of headspace gas (ml) at a CO 2 concentration of C (ppm) 168 required to provide the minimum 3 ml of CO 2 for a sample. injection into a molecular sieve cartridge.
193A second field test aimed to determine the optimum water:headspace ratio for hazardous to process and dispose (Gulliver et al., 2010). In addition, the logistics of 281 transporting large water volumes from remote locations to analytical laboratories remain,
282regardless.
283This new sampling method employs headspace equilibration (Kling et al., 1991; Hope 284 et al., 1995; Jahangir et al., 2012), which is widely used to provide samples for the 285 determination of dissolved greenhouse gas concentrations and δ 13 C of CO 2 in inland waters.
286In our case, we envisaged the method being used to collect aquatic CO 2 from relatively acidic