Assessing the effects of chamber placement, manual sampling and headspace mixing on CH4 fluxes in a laboratory experiment

A laboratory experiment was conducted with two types of closed static chambers to estimate the effects of chamber placement, manual headspace sampling and headspace mixing on methane (CH₄) fluxes. Chamber fluxes were compared to a known reference flux in a chamber calibration system. The measurements were conducted with three types of soils (coarse dry, fine dry and fine wet quarts sand) at five flux levels ranging from 60 to 2000 ??g CH₄ m^?2 h^?1. We found that the placement of a non-vented chamber disturbed the initial CH₄ concentration development within the chamber headspace for 10 to 30 s. Excluding this short period from the flux calculation resulted in a lower flux estimate (mean±SE) of 126?±?26 ??g CH₄ m^?2 h^?1 compared to 134?±?26 ??g CH₄ m^?2 h^?1 if data from time zero of the enclosure were included. We also found that in non-mixed chambers (no fan mixing) the gas sampling by syringes or gas bottles disturbed the development of CH₄ concentration during the enclosure. Furthermore, flux estimates in non-mixed chambers were significantly underestimated (on average 36%) compared to the measured reference fluxes. However, the use of fans to constantly mix the chamber headspace during enclosure significantly improved the goodness-of-fit of the regression analysis used to calculate the flux and further eliminated the disturbance of the manual sampling on the concentration development. We recommend that chambers should be vented during the placement of the chamber, and that fans are used as an integrated part of static chambers while headspace mixing with syringes should be avoided.