Water level drawdown is a hot moment for methane ebullition in a small eutrophic reservoir, Lacamas Lake, Washington

Background/Question/Methods

Methane (CH₄) is a potent greenhouse gas responsible for approximately 20% of the global greenhouse effect.  Recent work suggests that lakes and reservoirs are a globally significant source of methane, but these estimates are limited by lack of data and by methodological challenges. Direct bubble flux from reservoirs to the atmosphere (ebullition) is likely to be a very important pathway for CH₄ transfer, accounting for up to 90% of atmospheric CH₄ fluxes in the few systems where it has been measured.  Despite its likely importance, this CH₄ flux pathway is poorly constrained and its controls are poorly understood, particularly in temperate zone reservoirs.  The main goals of this project were: (1) to obtain an estimate of summertime methane ebullition rates in a temperate reservoir, and (2) to quantify the effects of a controlled dam spill on methane ebullition. To accomplish these goals, we constructed and deployed four bubble traps spanning vegetated and non-vegetated, littoral and profundal habitats.  Traps were deployed between 9 Aug and 27 October 2011 and were sampled at least weekly.  Gas samples were stored in exetainers and analyzed for CH₄on a Hewlett Packard gas chromatograph. 

Results/Conclusions

Between 8 and 15 September 2011, water level dropped by approximately 1.5 m and was associated with a dramatic increase in methane ebullition rates.  Lake-wide average flux was 4.21 mg CH₄ m^-2 d^-1 pre-spill (SD +/- 161), 154.04 mg CH₄ m^-2 d^-1 during-spill (SD +/- 3815), and 13.04 mg CH₄ m^-2 d^-1 post-spill (SD +/-297), representing a 36-fold increase in methane ebullition during dam spill.  Although the pattern of increased bubble flux with dam spill was consistent across all sites monitored, there was substantial variation amongst sites, with highest bubble fluxes during dam spill observed in the inlet channel (325.46 mg CH₄ m^-2 d^-1), somewhat lower fluxes observed in a vegetated site (75.24 mg CH₄ m^-2 d^-1), and the lowest flux observed from the deepest part of the lake (22.01 mg CH₄ m^-2 d^-1).  A comparison of mean summertime CH₄ ebullition and average CH₄ ebullition rates during spill suggests that upwards of 50% of total summertime CH₄ ebullition occurred during a one week spill event.  While post-spill fluxes dropped off significantly, they were still significantly elevated relative to pre-spill fluxes.  These results highlight the potential role of reservoir management in controlling atmospheric flux of an important greenhouse gas.